Structural Changes & Relative Perfusion Measurements in...
158
Structural Changes & Relative Perfusion Measurements in Crohn’s Disease PhD dissertation Rune T. Wilkens Health Aarhus University 2016
Transcript of Structural Changes & Relative Perfusion Measurements in...
Structural Changes & Relative Perfusion
Measurements in Crohn’s Disease
PhD dissertation
Rune T. Wilkens
Health Aarhus University
2016
Structural Changes & Relative Perfusion
Measurements in Crohn’s Disease PhD dissertation
Rune T. Wilkens
Health
Aarhus University
Department of Hepatology and Gastroenterology
Aarhus University Hospital
1
Supervisors Klaus Krogh, Professor, MD, DMSc, PhD Department of Hepatology and Gastroenterology Aarhus University Hospital, Aarhus, Denmark Henning Glerup, MD, PhD Diagnostic Centre, University Research Clinic for Innovative patient pathways Silkeborg Regional Hospital, Silkeborg, Denmark Agnete Hedemann Nielsen, MD Diagnostic Centre, University Research Clinic for Innovative patient pathways Silkeborg Regional Hospital, Silkeborg, Denmark Anders Tøttrup, Associate Professor, MD, DMSc, PhD Department of Surgery Aarhus University Hospital, Aarhus, Denmark Assessment committee Odd Helge Gilja, Professor, MD, DMSc, PhD Department of Medicine, Section of Gastroenterology Haukelund University Hospital, Bergen, Norway Jens Kjeldsen, Professor, MD, PhD Department of Gastroenterology Odense University Hospital, Odense, Denmark Frank V Mortensen, Professor, MD, DMSc (chairman) Department of Surgery Aarhus University Hospital, Aarhus, Denmark Correspondence Rune T Wilkens Diagnostic Centre, University Research Clinic for Innovative patient pathways Silkeborg Regional Hospital Falkevej 1-3, Silkeborg, Denmark [email protected]
2
Preface The present dissertation comprises the work of my PhD studies, performed at Silkeborg Regional Hospital,
where all the scans have been conducted and in close collaboration with Aarhus University Hospital, where all
the surgical procedures and the impedance planimetry procedures are performed.
It all began at a casual meeting in my apartment back in 2009 where Birgit Larsen and Henning Glerup inspired
me to take on the challenges with intestinal ultrasonography – which, according to their prediction, some day
might become the future in Danish gastroenterology. Henning Glerup later became my daily go to person and
supervisor. We have had endless discussions, visions and a common belief on innovation in gastroenterology. I
thank you both for introducing me into this world, and especially HG for being an inspiration as a supervisor,
doctor and as a person helping me through difficult times.
Klaus Krogh, whom I met at medical school and initially introduced me to research, was my obvious choice as
main supervisor. I’m grateful for all your hard work and timely manuscript editing in the critical phase, for your
advices (which I did not always follow) and your insight into this research world.
Agnete H Nielsen, I thank you for your constant believe in me, your motivation, inspiration and support.
Thanks to Anders Tøttrup, the IBD surgeon with a mind as a gastroenterologist but skills and passion like no
other surgeon.
Thank you to all of my great colleagues in Silkeborg, that helped me with recruitment and all the scans,
especially Valeriya P Hovgaard for MRE interpretation, John Hansen, Birgitte H Kristensen, Bonnie M Kirkegaard,
Susanne S Petersen and the rest of the nurses, secretaries and radiographers at Diagnostic Centre.
Thank you David A Peters, for building the software program for the MR analysis and for teaching me some of
this difficult technology. A special thanks also to Rikke H Hageman-Madsen for sticking to the project in crisis
and after seeking new challenges in Vejle, I’ve enjoyed our collaboration and all of our discussions.
The surgical department for helping me with recruitment, I thank Charlotte B Nørager, Daniel Kjær and Bodil
Sørensen, the fabulous nurses Margit Majgaard, Gitte K Sørensen, Chistina O Rasmussen.
The gastroenterologists from Silkeborg and Aarhus, especially Jens F Dahlerup, Lisbet A Christensen, Jørgen S
Agnholt, Christian L Hvas, Annett Canon, Ole K Bonderup, Sabine Becker, the study nurses Lisbet Gerdes and
Toto N Markussen and secretary Marianne Vonsild.
3
My skilled and smart Chinese/Danish collaborations Donghua Liao and Hans Gregersen, I thank you for picking
your brains and all of your contributions.
Patients for participating in the study, thank you all.
A great thanks to ALL of my phd/research colleagues and friends for supporting me through this long process,
Lotte Fynne, Anne-Mette Haase, Sara Hebøll, Anne Grosen, Stine Hald, Tine N Gregersen, David Haldrup and
Cecillie B Siggaard, René D Østgaard (and all the rest) – also for distracting me at our basement office and
reminding me that research should be fun.
I’m delighted for all the personal and legal help, that was provided by Susanne Kudsk prior to and throughout
my studies. For the kindness and helpfulness by all my ultrasound co-workers, Lars Bolvig, Jeannette Slot and
Charlotte Strandberg.
I’m totally indebted by the kindness, hospitality, true friendship, inspiration and endless discussions I’ve had
with Stephanie R Wilson & Kerri L. Novak from Calgary, Canada. The outmost obligations to you and your
families for giving me adventures of my life. Also BIG regards to all my friends and collaborators from (University
of) Calgary, Luneburg (Christian Maaser and Frauke Petersen), Düren (Horst Kinkel) and Bergen (Kim Nylund).
Last but not least, I’m thankful for the love, understanding and support from my family and friends, that I’ve
been neglecting for so long. A special thanks to my beloved fiancé Trine for all your support, acceptance and
fruitful discussions.
Thank you
Rune Wilkens
May 2016, Aarhus
4
Financial support This PhD study was financially supported by an unrestricted grant from AbbVie Denmark, Central Region Denmark Research Fund of Health Science, Regional Hospital Central Jutland, Research Fund, Crohn’s & Colitis Foundation Denmark, & Becket Fonden. AbbVie had no influence on study design, data analysis or interpretation, writing or submission of manuscripts. Conflict of interest RW has received travel grants from AbbVie Denmark and Takeda Denmark RW has received lecture grants, paid to his institution, from AbbVie Denmark and MSD Denmark The PhD thesis is based on the three papers listed below, referred to in the thesis by Roman numerals
I. Dynamic Contrast Enhanced Magnetic Resonance Enterography and Dynamic Contrast Enhanced Ultrasonography in Crohn’s Disease: An observational comparison study Wilkens R, Peters DA, Nielsen AH, Hovgaard VP, Glerup H, Krogh K (Manuscript submitted)
II. Validity of Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced MR
Enterography in the assessment of Crohn’s Disease Wilkens R, Hagemann-Madsen RH, Peters DA, Nielsen AH, Nørager CB, Glerup H, Krogh K (Ready for submission)
III. Which Cross-Sectional Imaging Parameters Predict Stiffness of Strictures in Crohn’s Disease
Wilkens R, Liao DH, Gregersen H, Nielsen AH, Peters DA, Nørager CB, Glerup H, Krogh K (Manuscript in preparation)
5
Abbreviations AIF arterial input function AIU arbitrary intensity units AUC area under curve CD Crohn’s disease CDAI Crohn’s disease activity index CDEIS Crohn’s disease endoscopic index of severity CDI colour Doppler imaging CEUS contrast enhanced ultrasonography CRP C-reactive protein CSA cross sectional area dB decibel DCE-MRE dynamic contrast enhanced magnetic resonance enterography DCEUS dynamic contrast enhanced ultrasonography E Young’s modulus ESR erythrocyte sedimentation rate f-calpro faecal calprotectin fps frames per second HBI Harvey-Bradshaw index ICC intra-class correlation coefficient) LoA limits of agreement MaRIA Magnetic Resonance index of activity MEGS magnetic resonance enterography global score MR Magnetic Resonance MRCA magnetic resonance contrast agent MRE Magnetic Resonance Enterography MTT mean transit time MTTl mean transit time local NS not significant PROM patient reported outcome measurements PE peak enhancement QIBA Quantitative Imaging Biomarkers Alliance QoF quality of fit ROI region of interest RT rise time SES-CD Simple endoscopic score in Crohn’s disease TIC time intensity curve TOA time of arrival TTP time to peak WiAUC wash-in area under curve WiR wash-in rate WiPI wash-in perfusion index WiT wash-in time WoR wash-out rate WoT wash-out time WoAUC wash-out area under curve
6
Table of Contents
English summary ...................................................................................................................................................... 8
Danish summary (Dansk resumé) ............................................................................................................................ 9
Background ............................................................................................................................................................ 10
Crohn’s Disease ................................................................................................................................................. 10
Chronic inflammatory bowel disease ............................................................................................................ 10
Disease manifestations .................................................................................................................................. 10
Disease course ............................................................................................................................................... 11
Defining disease severity ............................................................................................................................... 11
Inflammatory burden: Disease activity vs. fibrosis / chronicity .................................................................... 11
Diagnostic modalities and indices. .................................................................................................................... 12
Clinical assessment ........................................................................................................................................ 12
Biochemical evaluation.................................................................................................................................. 12
Endoscopic assessment ................................................................................................................................. 12
Imaging .......................................................................................................................................................... 13
Capsule Endoscopy ........................................................................................................................................ 15
Histology ........................................................................................................................................................ 15
Aims ....................................................................................................................................................................... 17
Hypotheses ............................................................................................................................................................ 17
Choice of methods ................................................................................................................................................. 18
Bowel Ultrasonography ..................................................................................................................................... 18
Dynamic Contrast Enhanced Ultrasonography ................................................................................................. 18
Magnetic Resonance Enterography .................................................................................................................. 24
Dynamic Contrast Enhanced Magnetic Resonance Enterography .................................................................... 25
Clinical assessment. ........................................................................................................................................... 26
Biochemical assessment .................................................................................................................................... 26
Faecal Calprotectin ............................................................................................................................................ 27
Histopathology .................................................................................................................................................. 27
Impedance planimetry ...................................................................................................................................... 28
Methods ................................................................................................................................................................ 29
Ultrasonography ................................................................................................................................................ 29
Contrast Enhanced Ultrasonography ................................................................................................................ 29
7
Magnetic Resonance Enterography .................................................................................................................. 31
Dynamic Contrast Enhanced Magnetic Resonance Enterography .................................................................... 31
Clinical assessment ............................................................................................................................................ 32
Blood and stool Samples ................................................................................................................................... 32
Histopathology .................................................................................................................................................. 32
Impedance Planimetry ...................................................................................................................................... 34
In Vitro Ultrasonography ................................................................................................................................... 34
Data handling and Statistics .............................................................................................................................. 35
Ethical considerations ........................................................................................................................................ 36
Summary of results ................................................................................................................................................ 37
Paper I ................................................................................................................................................................ 37
Paper II ............................................................................................................................................................... 39
Paper III .............................................................................................................................................................. 43
Discussion .............................................................................................................................................................. 45
Technical considerations in relative perfusion measurements ......................................................................... 45
Contrast Enhanced Ultrasonography ............................................................................................................ 45
Dynamic Contrast Enhanced MR Enterography ............................................................................................ 46
Evaluation of Reference standards.................................................................................................................... 47
Clinical Disease Activity Indices ..................................................................................................................... 47
Biochemical ................................................................................................................................................... 48
Histopathological references ......................................................................................................................... 48
Impedance Planimetry................................................................................................................................... 49
General study limitations .................................................................................................................................. 50
Statistical ....................................................................................................................................................... 50
Overall study outcomes ..................................................................................................................................... 51
Paper I: ........................................................................................................................................................... 51
Paper II ........................................................................................................................................................... 51
Paper III .......................................................................................................................................................... 52
Conclusion ............................................................................................................................................................. 53
Further perspectives .............................................................................................................................................. 54
References ............................................................................................................................................................. 55
Appendices ............................................................................................................................................................ 68
8
English summary Background: Measuring disease activity and severity in Crohn’s disease is a complex task, that relies on
objective measurements to guide both diagnosis and choice of treatment. Cross sectional imaging is becoming a
mandatory adjunction to gold standard ileocolonoscopy for mapping of disease. Increasing evidence is arising
for the applicability for dynamic contrast enhanced imaging techniques using both MR enterography and
particularly ultrasonography. The underlying hypothesis is the correlation between neoangiogenesis, perfusion
and acute inflammation. Most prior studies have investigated relative perfusion measurements utilizing either
mediocre methodology description or weak defined outcomes. We hypothesized, that ultrasonography and MR
enterography were equally good at predicting pathology and relative perfusion, that correlated with histology
and distensibility changes in affected bowel segments.
Aims: We aimed at investigating validity of structural changes and relative perfusion from ultrasonography and
MR enterography measurements in transmural Crohn’s disease.
Material and methods: We recruited patients with known Crohn’s disease with moderate to severe clinical
activity (study I) and patients referred to elective surgery for small bowel inflammation or complications (study II
& III). All patients were investigated with dynamic contrast enhanced ultrasonography and dynamic contrast
enhanced MR enterography. In study II we graded the histopathological changes of bowel specimens, and
specimens were investigated for distensibility stiffness in study III. These results were analysed for association to
structural imaging findings and time intensity curve parameters of the relative perfusion.
Results: In general, we found good relationship between the structural imaging findings; bowel wall thickness,
ulcers, and prestenotic dilatation with both histopathology and stiffness. Only moderate association was found
between the two dynamic contrast enhanced modalities, and only for some parameters. None of the relative
perfusion measurements were associated with histology and only initial slope of enhancement on MR
enterography associated with bowel wall stiffness.
Conclusion: Crohn’s disease patients can be reliably examined by both ultrasonography and MR enterography
for structural pathological changes, that reflect histological outcomes and stiffness. However, relative perfusion
measurements cannot be used interchangeably and are not predicting histological findings or bowel stiffness.
More methodological research is needed before dynamic contrast enhanced perfusion measurements can be
implemented as a separate part of standard of care.
9
Danish summary (Dansk resumé)
Baggrund At måle sygdomsaktivitet og sværhedsgrad af Crohn’s sygdom er en vanskelige opgave, der afhænger
af objektive mål til at guide både diagnostik og behandlingsmetode. Som supplement til guld-standard ileo-
koloskopien er billeddiagnostiske undersøgelser efterhånden ved at blive obligatoriske til vurdering af
sygdomsudbredningen. Der er efterhånden stigende evidens for anvendeligheden af dynamiske
kontrastforstærkede billeddiagnostiske tekniker så som MR enterography og ultralyd. Den underliggende
hypotese er sammenhængen mellem karnydannelse, blodgennemstrømning og akut inflammation. De fleste
tidligere studier har undersøgt måling af relativ blodgennemstrømning, men anvendt enten middelmådig
metode beskrivelse eller svage definitioner af de anvendte endepunkter. Vores hypotese var, at ultralyd og MR
enterografi var ligeværdige til at forudsige patologi og relative blodgennemstrømning, som modsvarer
histologiske fund samt distensibilitet af afficerede tarmsegmenter.
Formål: Vores mål var at undersøge validiteten af strukturelle forandringer og relativ blodgennemstrømning
målt med ultralyd og MR enterography i tarmvæggen hos patienter med Crohn’s sygdom.
Materiale og metoder: Vi inkluderede patienter med kendt Crohn’s sygdom med moderat til svær klinisk
sygdomsaktivitet (studie I) og patienter henvist til elektiv kirurgi på grund af betændelse i eller komplikationer
relateret til tyndtarmen (studie II & III). Alle patienter blev undersøgt med kontrast ultralyd og dynamisk
kontrastforstærket MR enterography. I studie II graduerede vi de histopatologiske forandringer i tarmresektater
og i studie III undersøgte vi resektaterne for distensibilitets graden af stivhed. Disse resultater analyserede vi for
association til både de strukturelle billeddiagnostiske forandringer samt tids intensitets kurve parametre for den
relative blodgennemstrømning.
Resultater: Generelt fandt vi god sammenhæng mellem de strukturelle forandringer på de billeddiagnostiske
undersøgelser, som tarmvægs tykkelse, sår og præstenotisk dilatation, med både histologiske fund og graden af
stivheden. Mellem de 2 dynamiske kontrastforstærkede undersøgelses modaliteter fandt vi kun en moderat
sammenhæng, og det kun for enkelte parametre. Ingen af de relative blodgennemstrømningsparametre var
associeret til histologien og kun den initiale hældning fra MR skanningen associerede med tarmens vægstivhed.
Konklusion: Crohn patienter kan undersøges pålideligt med ultralyd og MR enterografi for strukturelle pato-
logiske forandringer. Relativ blodgennemstrømning kan ej sammenlignes mellem modaliteterne og kan ikke for-
udsige de histologiske forandringer eller tarmens grad af stivhed. Betydelig forskning kræves fortsat før måling
af tarmens blodgennemstrømningen kan implementeres som selvstændig parameter i sygdomsvurderingen.
10
Background
Crohn’s Disease
Chronic inflammatory bowel disease
The two major chronic inflammatory bowel diseases are; Crohn’s Disease (CD) and Ulcerative Colitis. This thesis
will mainly focus on diagnosis, classification and assessment of CD. Where ulcerative colitis mainly affects the
rectum and the colon, CD may affect the entire intestinal tract from the mouth to the anus1. Additionally,
extraintestinal manifestations are common1. The pathogenesis of CD is not fully understood, but may involve
complex interactions between pathogenic bacteria, the intestinal epithelium, diminished intestinal mucosal
barrier and increased intestinal permeability1. The disease may be caused by environmental factors2 in
genetically susceptible individuals. A recent genetic association study even raises the probability, that
inflammatory bowel diseases should be further subdivided. Based on the individual persons DNA, CD could
potentially be better classified as CD with colonic involvement and CD with ileal involvement3.
Disease manifestations
Crohn’s Disease is a transmural disease characterized by intermittent periods of active disease and periods of
remission4. Traditionally, a rough disease classification from the Montreal working group5,6 divides patients into
age at onset (< 16 years (A1), 16-40 years (A2), and > 40 years (A3)), disease location (small bowel (L1), colon
(L2), small bowel and colon (L3), isolated upper disease (+L4)). Disease behaviour is further grouped as
inflammatory (B1), stricturing (B2) and penetrating (B3) disease with the possibility of a perianal disease
modifier (+p). Mucosal involvement is often the first sign of activity. Aphthous ulcers can progress into small,
large and confluent ulcerations involving large proportions of the bowel. However, CD is characterized by a
discontinuous involvement of the bowel, often containing skip lesions with intermittent normal bowel wall.
Diarrhoea, possibly with bloody stools, is often seen with active mucosal inflammation especially in the
colorectum. Stricturing disease with narrowing of the lumen and prestenotic dilatation can be seen in both
acute inflammation and, more commonly, in chronic disease1,7. Symptoms commonly associated with stricturing
disease are post prandial abdominal cramping, excessive bowel sounds (borborygmia), nausea and vomiting.
Patients often avoid certain types of food, that trigger discomfort and symptoms often resolve if food intake is
avoided8. Penetrating disease is characterized by luminal contents penetrating through the bowel wall, causing
phlegmons and abscesses in the mesentery or fistulas into other parts of the bowel or other epithelial lined
organs7. Apart from the localized involvement, there is a systemic autoimmune inflammatory response giving
11
rise to general symptoms like fatigue, weight loss, work disability9,10 and extra intestinal manifestations mainly
affecting eyes, joints, and skin11.
Disease course
Although symptoms might diminish over time4, CD is generally considered a progressive disease12 going from
inflammation towards stricturing disease and penetrating disease. Thus, up to 70 % of patients will need surgery
within a 10 year period after diagnosis13. It is hypothesized that early diagnosis and aggressive medical
treatment may modify disease outcome14,15 as seen in rheumatology16. Furthermore, smoking doubles the risk
of surgery17 and smoking cessation is associated with a more benign disease course18.
Defining disease severity
Crohn’s Disease is a complex autoimmune disease with several diverse phenotypes and multiple aspects of
impact on the individual patients’ quality of life. Even though, evaluation of treatment efficacy has tended to be
a “one size fits all” approach. Clinical studies focus on corticosteroid-free remission14, mucosal healing, deep
remission19 and, recently, even transmural healing20. However, grading of CD severity is undergoing constant
debate and the “treat-to-target” approach, adopted from rheumatology, is increasingly promoted21,22. Driven by
the U.S. Food and Drug Administration recommendations for future trial endpoints, consensus is moving
towards the use of patient reported outcome measurements (PROMs) combined with objective measurements
of inflammation21. Recently, a group of experts proposed an even broader perspective on this complex
disease23, dividing severity into
Impact of the disease on the patient
Inflammatory burden
Disease course
In the following we will focus on methods for assessment of inflammatory burden.
Inflammatory burden: Disease activity vs. fibrosis / chronicity
Inflammatory burden may be summarized as the extent of disease combined with the actual amount of
inflammation or wall thickness. Some authors have proposed various scores to measure elements of this
entity24,25. Others are claiming to define activity26–28, without acknowledging the high concomitant presence of
both histologically active and chronic / fibrotic disease29. Existing treatment for CD is not effective for treating
fibrotic strictures30. Thus, a modality is needed to measure burden of disease, and additionally to differentiate
between disease mainly consisting of treatable active inflammation, and chronic fibrotic disease where surgery
12
is needed. Since CD is exhibiting continues progression, grading indices should preferably give the amount of
activity or chronicity as a continuum rather than a dichotomous statement of active versus inactive disease.
Diagnostic modalities and indices.
Clinical assessment
Patient symptoms or clinical assessment are broadly used in research and everyday clinic to guide decision
making. The first and most widely used clinical index is the Crohn’s Disease activity index (CDAI)31. This index
consists of a seven days’ evaluation of number of stools per day, daily symptom severity, and general wellbeing
combined with assessment of extraintestinal manifestations, haemoglobin level, weight changes, use of
antispasmodic drugs and palpable abdominal mass. CDAI < 150 is regarded as clinical remission, 150-220 as
mild, 220-450 as moderate and > 450 as severe disease7. A simplified version was adopted in 1980 and called
the Harvey-Bradshaw Index (HBI)32. The advantage (and disadvantage) of this index is the momentary
assessment of the score on a daily basis. The indices have been the main outcome parameter in evaluation of
drug efficacy for decades33. However, treating beyond symptoms may improve long term patient outcomes15,19.
The STRIDE initiative suggested “resolution of pain” as the PROM of choice. However, the same group voted
that this cannot stand alone without more objective parameters like endoscopy21.
Biochemical evaluation
Systemic markers of inflammatory activity in blood samples have traditionally been haemoglobin, erythrocyte
sedimentation rate (ESR) and c-reactive protein (CRP)34. Stools can be analysed for calprotectin, a major protein
in neutrophilic granulocytes35 and a surrogate marker of intestinal inflammation in the small and large
intestine36,37. Faecal calprotectin (f-calpro) is a marker for predicting relapse after surgery38 and correlates better
with endoscopic indices than CRP and clinical assessment37,39. However, f-calpro does not provide information
on disease localization in the gut and it can be positive in the absence of other signs of inflammation and even
negative in patients with severe activity40.
Endoscopic assessment
Ileo-colonoscopy is regarded the gold standard for assessment of luminal CD. The best evaluated and validated
scores for grading activity are CD endoscopic index of severity (CDEIS)41 and Simple endoscopic score in CD (SES-
CD)28. CDEIS describes the rectum, sigmoid and left colon, transverse colon, right colon including cecum, and
terminal ileum with assessment of deep ulcers, superficial ulcers, percent of luminal surface involved by disease,
percent of luminal surface involved by ulcers, and ulcerated stricture or no- ulcerated stricture41. The simple
13
score assesses the same segments semi quantitatively by scoring presence of ulcers from (0-3), extent of
ulcerated surface (0-3), extent of affected surface (0-3) and presence and type of narrowings (0-3)28. The scores
have been developed to evaluate treatment response in clinical trials42. A third, but unfortunately non-
validated, scoring system for assessment of the neo-terminal ileum, the Rutgeerts score43, is also commonly
used in clinical trials. The European Crohn’s and Colitis Organisation42 suggested the definition of mucosal
healing as CDEIS or SES-CD = 0. Unfortunately, no consensus exists and different cut-off values are being
proposed in the literature44–46. Although endoscopy is regarded gold standard, endoscopic scores have some
limitations. They are complex, cumbersome and typically requires post procedural time to evaluate42. The scores
have a bias towards classifying colonic disease as more severe than ileal disease, since only 26 out of 75 patients
had their ileum scored during the development phase41. Reproducibility is good for both scores only as long as
scoring is performed by experienced gastroenterologists47,48 and further standardization of their interpretation
is needed49.
Imaging
Traditionally, plain film small bowel follow through was the modality of choice for investigating the small
intestine for ulcerated mucosa, strictures and fistulas with excellent spatial resolution50. Today, this modality is
regarded obsolete due to poor visualization of transmural disease and because of radiation burden51. CT
enterography or enteroclysis have relatively high spatial resolution and high accuracy in detecting inflammation
and complications in CD52. This ionizing cross-sectional modality is widely accessible but the use should be
minimized due to the high radiation burden that exceeds the small bowel follow through by a factor five51. For
repeated examinations, especially in young adults or teenagers with CD, repeated CT can lead to a high
accumulated radiation dose53,54.
Magnetic Resonance Enterography
Magnetic resonance enterography (MRE) is an accurate technique for the diagnosis and determination of extent
and activity of CD55. T1-weighted gadolinium-enhanced imaging is usually added to conventional anatomical
imaging as part of standard MRI small bowel protocols and shows considerable promise as a marker of disease
activity56. Three MRE based activity scoring systems have been developed based on endoscopy or histology as
gold standard57. The Magnetic Resonance Index of Activity (MaRIA)58 is an index validated against endoscopy
and with capability to follow treatment efficacy59. The index encompasses bowel wall thickness, relative contrast
enhancement, bowel wall oedema and intestinal ulcers58 with a 90 % sensitivity and 94 % specificity for ulcer
healing and mucosal healing, respectively. The Magnetic Resonance Enterography Global Score (MEGS) was
14
designed and initially validated against biopsy histopathology60 and later f-calpro61 and a combined clinical
reference standard62. The MEGS can distinguish responders from non-responders with a follow-up scan62. The
score is based on a multifactor semiquantitative score and a simplified score, described in more detail in
Methods of choice and Methods. The Nancy score includes diffusion-weighted imaging63, which is out of the
scope of this thesis. The Lemann score is also proposed for assessing digestive disease damage and is applicable
by both CT and MR24.
Dynamic Contrast Enhanced Magnetic Resonance Enterography
The chemical structure of the paramagnetic MR contrast agent (MRCA) includes a centrally placed gadolinium
ion, an octadentate ligand, and a coordinated water molecule. It changes the T1 relaxation time of extracellular
water molecules. MRCA is freely redistributed from the vascular to the interstitial space and cannot truly reflect
blood flow64–66. Dynamic contrast enhanced MRE (DCE-MRE) is commonly used to determine perfusion of the
brain as the vascular-brain-barrier keeps MRCA within the blood vessels. This is unfortunately not the case in
extra-cerebral MR contrast studies, where contrast extravasation makes assessment of blood volume difficult67.
Pixel-by-pixel or manually drawn regions of interest (ROIs) are used to build signal amplitude time intensity
curves (TICs), reflecting dynamic changes in T1 relaxation time. Calculating TIC parameters is easily done.
However, tissue extravasation of MRCA and difficulties in consistent imaging acquisition68 limit interpretation as
they do not reflect biophysiology69. Nevertheless, changes in signal intensity may be useful as relative
quantitative surrogate markers of tissue perfusion70,71.
Ultrasonography
B-mode ultrasound of the bowel with colour Doppler imaging (CDI), either with or without use of oral contrast
solution, has similar sensitivity and specificity as MRI and CT for assessment of inflammatory activity, extent and
location52. Furthermore, CDI correlates well with endoscopy72. Stenotic and penetrating disease can also be
detected by US55,73. A score has been proposed by Calabrese et al. for the evaluation of disease burden74.
However, the score is cumbersome and time consuming and does not differentiate between activity and
chronicity. Blood flow on US has traditionally been assessed with conventional CDI75,76. Although CDI is highly
sensitive to the detection of blood flow, it is better able to detect relatively fast-moving blood in larger blood
vessels than the lower perfusion level of the gut. Thus, CDI is inadequate for the quantitative assessment of
perfusion77 and can be falsely negative78.
15
Dynamic Contrast Enhanced Ultrasonography
As with MRI and CT, intravenous contrast can be used to evaluate the bowel vasculature on US. Unlike the other
modalities, however, US has excellent temporal resolution and all contrast-enhanced ultrasonography (CEUS)
examinations are dynamic. Therefore, CEUS can be used to perform relative and absolute quantifications in
addition to real-time qualitative assessments of enhancement pattern and intensity79. Although many have tried
to establish a scoring system with various quality of reference standards78,80–83, none have so far been validated.
One of the main problems with the modality is the post-acquisition analysis. The procedure is far from
standardized and there is a lack of a solid methodology to analyse data from TICs with a high quality of fit and
reliability. This modality will be discussed in further details in Choice of methods.
Capsule Endoscopy
Three different scores exist for capsule endoscopy of the small bowel in CD42. The presence of more than three
ulcers in the small intestine has been regarded as diagnostic for CD84. The Capsule Endoscopy CDAI is a validated
score assessing small bowel inflammation, extend and narrowings, with strong correlation between observers
for the total score85. Finally, the Lewis score86 assess the small bowel in tertiles for villus oedema, number of
ulcers, extend and distribution of lesions. The great advantage is the low ulcer miss rate of 1 % yielding a high
sensitivity86. However, the consequence is a low positive predictive value at about 50 %45. Further, the modality
can by compromised by poor cleansing quality of the small bowel and there is a risk of capsule retention in
stenotic CD.87
Histology
Inflammation can be defined as local reaction to injury. In the inflammatory cascade, the contributions made by
hyperaemia, exudation of fluid, infiltrates carrying inflammatory cells, and cell proliferation vary with the type of
injury88. In classic histopathology active inflammation is regarded as epithelial damage in association with
neutrophils7,89,90. If the cause of injury does not disappear, homeostasis is disturbed and drives the active
inflammation in the direction of chronicity with activated myofibroblasts and intestinal fibroses91 rather than
successful intestinal healing where myofibroblasts undergo degradation92. For the diagnosis of CD7, colonoscopy
with multiple colonic and ileal biopsies (minimum of two per site) is established as first line procedure. Biopsies
are preferably taken from both involved and non-involved mucosa. European guidelines7 state that if
granulomas are detected, only one additional feature is needed for establishing a firm diagnosis (preferably
focal inflammation like architectural abnormalities). Tanaka et al. evaluated 70 histological features in colon
16
biopsies from Japanese subjects and validated their index on Canadian subjects. They have suggested elements
for scoring IBD vs. non-IBD and for differentiating CD against ulcerative colitis93.
Gross examination of a resected specimen, typically reveals firm “creeping fat”, shown as sub-serosal fat
wrapping. The bowel wall is usually thickened and often narrowed over short or long distances. The mucosal
surface can be seen with aphthous or deep longitudinal ulcers, sometimes leaving a cobblestone apperance94.
Inflammatory changes of the serosa, including Crohn’s rosaries, can be seen in CD, which distinguishes it from
UC95. Abundant inflammation can be seen creating deep fissures, sinuses and fistulas94 which is more common
in ileum or ileocolonic disease90. Graham et al.96 showed that a strictured bowel segment contains more
collagen than a normal bowel segment within the same patient. The same group96 also found a relative increase
in type III collagen, produced by fibroblasts97, and type V collagen, produced mainly by adjacent smooth muscle
cells96,98.
17
Aims
1) To compare the following characteristics of CD obtained with US and MRE (study I):
a. Structural changes (e.g. thickness and length of disease)
b. Relative TIC perfusion (obtained with DCE-MRE and CEUS)
2) To compare the following findings on US and MRE with histopathology from CD stenoses (study II):
a. Structural changes (e.g. thickness and ulcers)
b. Relative TIC perfusion parameters (obtained by DCE-MRE and CEUS)
3) To compare the following findings on US and MRE with wall stiffness in CD stenoses (study III):
a. Structural changes (e.g. prestenotic dilatation and bowel wall pattern)
b. Relative TIC perfusion parameters (obtained with DCE-MRE and CEUS).
4) To investigate the repeatability of ROIs for CEUS analysis of small intestinal perfusion in CD (study I).
Hypotheses
1) Comparison of characteristics of CD obtained with US and MRE: a. There is no difference in structural changes found with US and MRE (e.g. bowel wall thickness
and length of disease) b. Relative TIC perfusion parameters are associated between CEUS and DCE-MRE for the initial
part of the TIC 2) US and MRE and histopathology from CD stenoses:
a. US and MRE are equally valid for identification of structural bowel wall changes like thickness
and ulcers b. Relative TIC perfusion parameters obtained by CEUS and DCE-MRE are equally valid and
correlate with active inflammation and inversely with chronic fibrotic changes 3) Correlations between US and MRE and wall stiffness in CD stenoses:
a. US and MRE can predict bowel wall stiffness from structural findings (presence of stricture and
prestenotic dilatation) b. Relative TIC perfusion parameters obtained by CEUS and DCE-MRE are inversely correlated with
wall stiffness. 4) The selection of ROI affects the repeatability of CEUS analysis of small intestinal perfusion in CD.
18
Choice of methods
As described above, the best modalities for non-invasive, radiation-free assessment of transmural localized ileal
CD are US and MRE54. As we wanted to explore the use of methods for assessment of relative perfusion, we
decided on CEUS and DCE-MRE. Considerations about experimental protocols and data analysis are given in the
following sections.
Bowel Ultrasonography
In experienced hands, ultrasound of the bowel may be possible even in the presence of intestinal gas. Usually,
an abdominal probe is used initially to get an overview while a medium to high frequency probe is used for
detailed examination of the site of interest99. The choice of probe depends on the investigator and the patient
composition. If the area of interest is deeply located, a low frequency probe will usually be needed. In our study,
most of the scans were performed with a linear 4-9 MHz matrix probe, which provided good details and
penetration, see Methods. With a high frequency probe it is usually possible to detect 5 to 9 sonographic layers
of the bowel wall that correlate relatively well with histology100,101. Nylund et al.101 has also shown that Crohn’s
rosaries appears as tiny echo poor bumps on the outer intestinal border on an ex-vivo ultrasound scan. Bowel
wall thickness is regarded as one of the most important factors for discriminating between normal and diseased
bowel with cut-offs between 2-4(5) mm for terminal ileum and colon52,102. However, a comprehensive study on
healthy volunteers revealed that in 95 % the bowel wall thickness from jejunum to sigmoid colon was between
0.5 – 1.8 mm103. In a meta-analysis, sensitivity for diagnosing IBD was found to be 91.0 % with a cut off at 3 mm
and 94 % for 4 mm. Specificity was also highest for the 4 mm cut-off. However, the meta-analysis was not
designed with the purpose to define the best cut-off and a lower cut-off value would intuitively give a higher
sensitivity but lower specificity. We decided to use 3 mm rather than 4 mm as a cut-off for recruitment of
patients into our studies, since we regarded patients with ≤ 3 mm as normal.
Dynamic Contrast Enhanced Ultrasonography
The US Probe The trade-off between resolution and penetration, which plagues all ultrasound exams, is even more
pronounced when performing CEUS. Although high frequency probes provide better image resolution for bowel
segments close to the abdominal wall, there is less backscatter signal intensity, more bubble destruction and
more signal attenuation104,105. Therefore, scanning with high frequency probes is ineffective in patients with high
BMI or deep bowel loops. Applying different probes and systems for comparison still needs to be evaluated for
repeatable performance. As for now selection of different probes and probe frequencies will not give
19
comparable results between patients or even between exams106. Therefore, we chose the 9L4 probe for all CEUS
examinations throughout the studies as it is a medium frequency probe allowing for sufficient amount of details
without highly sacrificing penetration.
Image Acquisition The imaging was performed by continuous acquisition and recording of data in a fixed location. This allows for
objective measurement of relative bowel perfusion as the intensity of backscatter signal correlates well with the
bubble concentration in a specific ROI107. Beam attenuation occurs with increasing volume of bubbles
necessitating appropriate dosing for low concentration of the contrast agent108. On the other hand, the
recommended contrast dose for bowel is higher than that for liver109. Since we generally use 1.2 ml for liver and
the equipment has a sensitive bubble detection, we chose 2.4 ml SonoVue® (Bracco Imaging, Milan, Italy) per
injection. In analysis of non-raw data, a high dynamic range of over 50 is required for good software de-log-
compression and thus accurate results108,110–113. Reduction in dynamic range/compression will produce higher
signal intensity, increase the risk of oversaturation and, thereby, limit reproducibility106,110,112. A constant low
mechanical index is required to preserve the bubble population105. Frame rates can easily be kept low with very
little effect on TIC114. We therefor decided on the maximum of dynamic range of 80, mechanical index < 0.1 and
framerate at 10 frames per second.
Choice of contrast administration At the time of study planning and approval, two main quantification techniques were described. Bolus injection
and infusion-burst-replenishment79. The advantages of the former are the simple setup, less use of contrast and
that it is unnecessary to have a pump and waiting for a steady state level of contrast to arrive. The advantages
of the latter are the ability to perform multiple repeated measurements and direct computation of relative
blood volume, mean transit time and thus a perfusion index. Due to simplification and practicality, we chose the
bolus injection, which was the most common use in prior publications assessing bowel wall perfusion115–117.
Deconvolution is a process that takes the arterial input function (AIF) into account, thus allowing to scale the
relative perfusion measurements accordingly118. In a phantom model, the reproducibility was comparable
between deconvolution and non-deconvolution, whereas the reproducibility was superior for deconvolution in a
mouse model. However, the even greater benefit is the possibility to obtain absolute measurements of blood
perfusion119. The central volume theorem states that
𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑓𝑓𝐵𝐵𝐵𝐵𝑓𝑓 = 𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑣𝑣𝐵𝐵𝐵𝐵𝑣𝑣𝑣𝑣𝑣𝑣𝑀𝑀𝑣𝑣𝑀𝑀𝑀𝑀 𝑡𝑡𝑡𝑡𝑀𝑀𝑀𝑀𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡𝑣𝑣𝑣𝑣
(1)
Where blood flow is the main parameter of interest. In brief, in the deconvolution theory the equation is
20
𝐶𝐶(𝑡𝑡) = 𝜌𝜌𝑘𝑘𝐻𝐻∙ (𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑓𝑓𝐵𝐵𝐵𝐵𝑓𝑓) ∙ (𝐶𝐶𝑀𝑀(𝑡𝑡) ⊗𝑅𝑅(𝑡𝑡)) (2)
Where C and Ca are the concentration of contrast at the function of time t in the tissue and in the artery
respectfully. The letter ρ denotes the tissue density and kH the difference in haematocrit between arteries and
capillaries and R(t) is the residue function. Finally ⨂ is the tensor product. This is under several assumptions,
including ρ = 1 and kH = 1 and an instantaneous bolus injection. Through an algebraic approach using matrix
calculations, the AIF can be calculated and the blood flow measured. The details of this119,120 is outside the scope
of this thesis. Since the modality was not tested for CEUS in humans and the assumptions and calculations were
overwhelming without dedicated commercialized product at hand, we decided not to take the AIF into account.
Performing the Exam: There are many techniques which improve the acquisition and thereby the post-acquisition analysis. In order to
support in-plane motion compensation, alignment of the long axis of the probe in the plane of movement of the
bowel, caused by respiration, can be crucial. Therefore, in the majority of cases, a sagittal plane orientation is
preferred to compensate for expected diaphragmatic excursion with breathing. Also, a longitudinal orientation
of the transducer face to the bowel wall was preferred, as it provides more desirable ROIs. However, we tried
both directions on the bowel wall, since the transverse approach may be favourable in some occasions. In
general, longer acquisition time is preferable79,121, as more parameters can be measured due to a more
comprehensive TIC, especially the area under curve (AUC). We were not able to record longer than 90 seconds
from end of flush injection.
Post-acquisition: Dedicated software for detailed analysis of quantification data is required and offered by most ultrasound
manufactures. The majority of systems have the limitation of being vendor specific, allowing for analysis of their
own data only. Therefore, data from different systems cannot be compared. Platform neutral software allows
for analysis of data from all systems utilizing data in a proper format. The ability of such a system to neutralize
differences in quantification parameters across platforms is desirable but not yet validated in clinical trials. CEUS
equipment generally acquires raw native data in a linear format at the time of acquisition. However, images are
displayed in a log-compressed format. The transfer and export of data is generally in DICOM format, log-
compressed, including a picture or movie file, plus a header, containing the acquisition details79,113. Data analysis
should be based on linearized data only, which can be accomplished with optional special built-in software
programs. Alternately, the DICOM header information can be used to re-linearize acquisition data for analysis by
a specific de-log-compression data conversion with dedicated analytical software110–113. Other forms of data
21
such as AVI, MPEG and other video formats without acquisition information are less useful for analysis and
should be avoided. These video formats allow only a pixel-by-pixel analysis of log-compressed data and not an
intensity measurement from linear format. Since the build in option for Siemens was less useful for contrast
quantification, our main option to re-linearize data from DICOM format112 was the commercialized software,
VueBox® (Bracco Suisse SA, Geneva, Switzerland).
A TIC is constructed by calculating the change in mean ROI intensity over time, figure 1. Therefore, ROI size and
placement do play a crucial role in representing the overall mean intensity and tissue perfusion, figure 1122. A
small ROI size, positioned on the most enhanced segment of the bowel, does not show the true state of disease
activity in the bowel segment and a small ROI size is also more susceptible to motion related fluctuations in
intensity, figure 1. In addition, small ROI size may not have a high reproducibility rate due to variation in
intensity when ROI location is changed. A large ROI size may show a better approximation of overall disease
activity in the inflamed region by providing a better mean intensity, quality fitted curve, and measurement of
tissue perfusion. However, large ROI size may cover less perfused layers or segments of the bowel and
consequently result in a falsely lower peak enhancement. A useful feature for ROI placement optimization is an
in-plane motion compensation feature, which enables automatic detection of movement in the B-mode picture
and adjusts the ROI placement accordingly. In-plane motion compensation helps reducing the severe fluctuation
in TIC112, but it can be inaccurate when only small movements occur, and should thus not be used123. Single
frames can then be deleted due to severe displacement of ROI secondary to respiration or peristalsis. The
resultant curve will still be suitable even with very few frames114. Also, colour mapping of selected parameters
such as PE, RT and AUC is beneficial in ROI placement, figure 2. In study I, we therefore tested different ROI sizes
and numbers to determine the best repeatable ROI strategy. Some analytical software programs offer two types
of data presentation, a linear graph measured in arbitrary intensity units (AIU) and measurements or a graph in
decibels (dB). The linear graph shows the most realistic presentation of raw data, and consequently provides the
most reliable method for performing measurements. However, viewing the actual TIC in log-format
approximates more closely the visual interpretation of the enhancement79. Additionally, the dB values followed
a Gaussian distribution, which the AIU did not, see statistical analyses. TIC is analysed based on time
parameters, intensity parameters and a combination of them. These parameters should be calculated and
reported from the fitted curve on linear data. Due to the nature of the log-formula, curve fitting on log-data is
heavily susceptible to minor changes in the baseline and late wash out period. Therefore, log-fitted curves
should not be used79,121,124,125. Also, various mathematical equations are used for curve fitting in each graph
setting, especially for the linear data121.
22
Figure 2 Heat map
Colour mapping image on the left showing Peak Enhancement
with in the large circular Region of Interest. On the right side
the corresponding b-mode image is shown. In this case Peak
Enhancement is constant throughout the bowel wall layers and
thus a homogenous enhancement pattern.
Figure 1 TIC effect of size of region of interest (ROI).
Four different ROIs drawn are illustrated in the upper
right contrast enhanced image and the
corresponding time intensity curves are shown on
the graph. The largest ROI in blue shows the lowest
peak enhancement (PE) and the least signal
fluctuations. Smaller ROIs will have higher PEs but
show considerable signal fluctuation. Therefore, care
has to be taken when using very small ROIs
Figure 3 Measurable time intensity curve
parameters on contrast enhanced ultrasound.
See text for further details. Time parameters,
measured in seconds (s), are displayed in grey.
Intensity Parameters, measured in arbitrary
intensity units (AIU) on Linear scale and are
displayed in blue Combined parameters,
measured in AIU/s and Sum of AIUs are displayed
in green and red respectively.
23
The details and specifics of each model are beyond the scope of this thesis. We used the non-disclosed formula
from VueBox. The TIC parameters, that can be obtained from a bolus injection curve are shown in Figure 3 and
described in the following126.
Time Parameters: Time of arrival (TOA) or Time zero offset (t0) is the time from injection to arrival of the microbubble contrast
agent bolus in the ROI. It is a parameter heavily dependent on many factors like injection site/distance from the
heart, cardiac output and speed of contrast and flush injection127. Therefore, it should not be used as a relevant
clinical indicator. However, it is useful in calculating other time parameters. It is worth noting that some
programs start the upslope of the fitted curve from the first recorded frame and not TOA which should be taken
into consideration. Thus, manual elimination of the frames prior to TOA is necessary. To have consistent
measurement of TOA, timer and recording should be started at the same time. TOA can be useful in case of
reinjection and when breath hold timing is required.
Wash-in Time (WiT) and Rise Time (RT) are similar parameters, defined individually, based on quantification
software and thus, a clear definition of measured values is required in clinical studies. WiT is the interval from
TOA (start of the upslope) to the PE. RT is the time for which intensity varies from 5 % to 95 % or from 10 % to
90 % to PE (depending on software).
Time to Peak (TTP) is the time from injection to PE. The parameter is defined by TTP = TOA + WiT and should
therefore only be used for calculating the WiT. WiT and RT are often used interchangeably with TTP, which can
be misleading in data interpretation and analysis.
Wash-out Time (WoT) or Fall Time is the time from PE until total washout, which not always occurs within the
recording time.
Mean Transit Time (MTT) or Mean Transit Time local (MTTl) corresponds to the centre of gravity of the fitted
curve calculated using the programs specific perfusion model function, corrected for TOA. MTT can be
estimated by the time point where the AUC is divided into two halves, which for optimal estimation requires
total washout. Some programs, therefore, use the Full Width Half Max for the same estimation79.
Full Width Half Max: Is measured from the time between 50 % PE on the wash in slope to 50 % PE on wash out
slope. This parameter is often confused with or used for calculation of MTT/MTTl79,121.
Intensity Parameters Peak Enhancement (PE) refers to the highest enhancement level on the TIC and can be measured in AIU and dB.
Like other parameters, PE also should be accurately measured and calculated from linear data.
Intensity and combined parameters depend on the US system and many confounding factors discussed earlier.
24
Combined Parameters Area Under the Curve (AUC) is calculated from the linear curve only and shows the sum of all instant intensity
values on the fitted curve. It is believed that AUC reflects a reasonable approximation of disease activity.128,129
Since AUC is the sum of all instant intensity values, the same limitations which apply to measurement of PE will
apply to AUC when it comes to standardizing the calculation and its correlation with disease activity among
multiple CEUS centres. AUC can even be diverted into Wash-in AUC (WiAUC) and Wash out AUC (WoAUC).
AUC = WiAUC + WoAUC
Wash-in Rate (WiR) is the tangent of the wash in slope. Either perfectly calculated at the steepest part or
approximated by calculating difference in intensity values between two points on the curve divided by the time
between these two points, e.g. intensity 75%-25%
𝑊𝑊𝑊𝑊𝑅𝑅 =Y75%Wi− Y25%Wi
X75%Wi− X25%Wi
Wash-out Rate (WoR) is the tangent of the wash out slope. Either perfectly calculated at the steepest part or
approximated by calculating difference in intensity values between 2 points on the wash out curve (see WiR).
Wash-in Perfusion Index (WiPI) can be described as
𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊 = 𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝐶𝐶𝑅𝑅𝑅𝑅
Since there is no consensus in the literature about what parameters to use, we tested most of the TIC
parameters against our outcome variables. TOA and TTP was consistently avoided and Fall Time deemed the
least important factor and it was thus avoided to minimize mass significance. During the analysis phase and
repeatability phase we found the MTTl very heterogeneous on the heat maps and less reproducible between
calculations. It was therefore not utilized for further correlation.
Magnetic Resonance Enterography
The use of oral contrast agents in MRE is essential to evaluate finer details in the bowel wall and minimize false
positive diagnostics. The choice of either enterography and enteroclysis is continuously debated. Enterography
is favoured by patients, since it is more tolerable and without the use of fluoroscopy to ensure correct
placement of the nasojejunal catheter130. Although enteroclysis is superior because of the distension of the
small intestine, especially the jejunum, the diagnostic accuracy and reproducibility are equal for the two
methods131. We have chosen enterography for better patient acceptance in our institution, which we also
applied in our studies. Like on ultrasonography, structural findings can be reported from MRE. MEGS and MaRIA
were the basis of our structural elements to record from the scans.
25
𝑀𝑀𝑀𝑀𝑀𝑀𝑊𝑊𝑀𝑀 (𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑡𝑡) = 1.5 ∙ 𝑓𝑓𝑀𝑀𝐵𝐵𝐵𝐵 𝑡𝑡ℎ𝑊𝑊𝑖𝑖𝑖𝑖𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 (𝑠𝑠𝑠𝑠) + 0.02 ∙ 𝑅𝑅𝐶𝐶𝑅𝑅 + 5 ∙ 𝐵𝐵𝑠𝑠𝐵𝐵𝑠𝑠𝑠𝑠𝑀𝑀 + 10 ∙ 𝑢𝑢𝐵𝐵𝑖𝑖𝑠𝑠𝑀𝑀𝑀𝑀𝑡𝑡𝑊𝑊𝐵𝐵𝑠𝑠
Where RCE is Relative contrast enhancement at 70 seconds.
Dynamic Contrast Enhanced Magnetic Resonance Enterography A variety of contrast-enhanced acquisition techniques and analyses are available for MRE. They range in
complexity from a single post-contrast dataset for radiologist-subjective assessment to rapid serial acquisitions
with full pharmacokinetic modelling68. However, all methods have limitations or may not be feasible. If we want
to analyse underlying pathophysiologic processes with DCE-MRE in an absolute quantitative manner,
parameters like the volume transfer constant Ktrans (that represents perfusion), extravascular volume fraction Ve
and Kep (the gadolinium washout rate) can be calculated using the two-compartment Tofts (extended) model or
variations of this model132.
𝐾𝐾𝑣𝑣𝑒𝑒 = 𝐾𝐾𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
𝑉𝑉𝑒𝑒 (3)
Reliable arterial input function (AIF) and temporal tissue sampling of ≤ 5 s / image are needed to fulfil the
requirements of this calculation68. The intrinsic pre-contrast T1 values of the area of interest and blood are also
needed. This is commonly performed using the variable flip angle technique133. The technique is unfortunately
far from perfect, since there is a variability of up to ±40 % of the flip angle in vivo. As a 15 % flip angle deviation
accounts for about 28 % error margin, this can lead to large uncertainties. In high temporal and resolution
images, signal-to-noise ratio further leads to measurement errors when comparing individual images to
Table 1 Magnetic Resonance Enterography Global Score (MEGS) Score 0 1 2 3
Mural thickness < 3 mm > 3 - 5 mm > 5 - 7 mm > 7 mm Mural T2 signal Equivalent to normal
bowel wall Minor increase in signal: bowel appears dark grey on fat-saturated images
Moderate increase in signal: bowel wall appears light grey on fat-saturated images
Marked increase in signal: bowel wall contains areas of white high signal approaching that of luminal content
Peri-mural T2 signal (mesenteric oedema)
Equivalent to normal mesentery
Increase in mesenteric signal but no fluid
Small fluid rim (≤ 2mm) Larger fluid rim (> 2 mm)
T1 Enhancement Equivalent to normal bowel wall
Minor enhancement: Bowel wall signal greater than normal small bowel but significantly less than nearby vascular structures
Moderate enhancement: bowel wall signal increased but somewhat less than nearby vascular structures
Marked enhancement: bowel wall signal approaches that of nearby vascular structures
Mural enhancement pattern N/A or homogenous Mucosal Layered Haustral loss (colon only) None <1/3 segment 1/3 to 2/3 segment >2/3 segment Multiplication factor per segment 0-5 cm x 1 5-15 cm x 1.5 >15 cm x 2 Additional score for extramural features 0 5 Lymph nodes (≥ 1 cm measured in shortest diameter)
Absent Present
Comb sign (linear densities on the mesenteric side of affected bowel segments)
Absent Present
Abscess Absent Present Fistula Absent Present
26
baseline. A recent study concludes that a simple signal difference strategy is more robust to the varying flip
angles observed in vivo than the underlying assumptions needed for Tofts model allowing for the technique to
be used in follow-up studies133. Sourbron et al. even state that Tofts model is useful in weakly vascularized or
highly perfused tissues only134. Absolute quantification of blood flow analyses in the bowel using DCE-MRE has
only been described in few studies71,135 and more simple methods are more often used.
We applied the variable flip angle technique (using the following degrees of the flip angle: 5, 10, 15, 20, 25) and
tested the reliability in the psoas muscle, expected to have a constant T1. However, we found the method
unreliable and decided to use the absolute signal difference instead. We chose the flip angle of the dynamic
sequence to be 24, based on prior recommendations136. We also considered applying the AIF for DCE-MRE.
However, there was a large variability even between aorta, iliac artery and superior mesenteric artery. Even
worse, the proportion of changes within arteries were vast between individuals, hampering the possibility to
choose the optimal artery for comparison.
To extract DCE-MRE values, a custom made program (RoiTool for Mac) was built in MATLAB® (MathWorks®,
Natick, MA). This program allows the examination of MRE DICOM files for the dynamic sequence and the
drawing of ROIs within all slices and frames. Since only the initial curves were regarded in our study, we did not
fit a specific curve type137.
Clinical assessment. We decided to assess all patients with the two most accepted clinical scores; CDAI and HBI. In study I, we chose
the defined cut-off for moderate disease (CDAI ≥ 220 or HBI ≥ 8) as inclusion criteria23. This was based on the
assumption that moderate to severe active disease would have a higher perfusion and because we aimed at
including patients with a more chronic or fibrotic state in study II, thereby covering a wider spectrum of CD.
Furthermore, by including mild disease, the bowel would potentially have more peristalsis and the bowel wall
could be thinner, giving rise to a lower success rate.
Biochemical assessment At our department routine biochemical assessment of CD includes ESR, haematocrit, haemoglobin, CRP,
leucocytes, vitamin D, albumin, and f-calpro (see below). Haematocrit is part of the CDAI score while CRP and
leucocytes are markers of systemic inflammation. Low vitamin D is also regarded as an indirect marker of
inflammation138, whereas albumin is a negative acute phase reactant and also related to malnutrition and post-
surgical complications139,140.
27
Faecal Calprotectin
F-calpro is a routine marker at our institution, and has been shown to help predicting recurrence after surgery38
and correlates with endoscopy and MRE141,142.
Histopathology Bowel wall thickness,
luminal diameter and
histological components of
inflammation or fibrosis
vary significantly along the
longitudinal axis of the intestine and maybe even within the
circumference143. We therefore decided to fixate the bowel
without cutting it open, allowing us to examine the full
circumference and the full transmural disease in one slice. The
disadvantage about this approach is the slightly increased
difficulties in cleansing of the bowel and proper fixation in
formaldehyde. We could overcome the limitations by dividing
exceptional long resected segments (e.g. > 50 cm) into two or
three segments, see Methods, and by using intraluminal wicks for
better central formaldehyde uptake and fixation. There are
several validated histological indices for ulcerative colitis144,145, but
still no optimal method for the grading of transmural inflammation and/or fibrosis in CD. Borley et al.143
described a semiquantitative score for both chronic and acute histological changes in the bowel wall, allowing
investigation of the heterogeneity of histology throughout the involved segment. The score is neither validated
nor produced with the purpose of assessing pure activity in CD. Other scores exist but are too complex146. We
further aimed at an area based score of tissue components like collagen, muscle, neutrophils etc. However, the
computer system intended for this purpose did not allow us to make these calculations within our time frame,
and we ended up with the semiquantitative approach for activity by Borley et al.143 and chronicity/fibrosis by
Chiorean et al.147, and Baumgart et al.148, Table 2, 3 & 4. Haematoxylin-eosin is the standard histopathology
staining, whereas Masson’s trichrome enhances the collagen (blue) and muscle tissue (red/purple).
Immunohistochemical staining on Ventana OptiView with anti-Desmin (Dako® clone D33, Glostrup, Denmark)
was performed to characterize smooth muscle cells and myofibroblasts149. Anti-CD34 (Confirm® clone
Table 2 Acute inflammatory features. Score Mucosal Ulceration score Oedema score Depth of neutrophil penetration 0 None None None 1 Aphthous ulceration only, < 7 mm diameter Mild Mucosa 2 Linear or rake ulceration Moderate Submucosa 3 Confluent or large ulceration Severe Muscularis propria 4 Serosa extramural fat
Table 3 Fibrostenotic features Score Mucosal Ulceration score 0 No or minimal fibrosis limited to submucosa
(< 25 % thickness) 1 Mild stricture (> 15 mm) with nondilated
lumen. Submucosal fibrosis and muscular hyperplasia > 25 % with preserved layers
2 Massive transmural fibrosis; effacement of normal layers; severe stricture.
Table 4 Collagen deposition Score Depth of excessive collagen deposition 0 No increased collagen 1 Increased in mucosa 2 Increased in muscularis mucosa 3 Increased in submucosa 4 Increased in muscularis propria 5 Increased in all layers incl. serosa
28
QBEnd/10, Roche®, Ventana Medical Systems, Inc., Tucson, AZ) stains vessels and was chosen for quantification
of micro vessel density within the bowel wall150.
Impedance planimetry
Stenoses in CD are usually located in the small intestine making in vivo mechanical testing difficult. Although
balloon dilatation is commonly utilized in strictures located at the terminal ileum or the neo-terminal ileum151
there is currently only a limited knowledge about the mechanical properties of CD, incl. strictures148,152,153.
Impedance planimetry is a valid method for computation of luminal cross sectional area (CSA) and pressure in
hollow organs like the gastrointestinal tract. During the last 25 years a large number of papers based on
impedance planimetry have described biomechanical properties of most regions of the GI tract in humans and
animals. Previously, the use of impedance planimetry was hampered by the cumbersome calibration procedure.
However, a recent device, EndoFLIP® probe (Crospon™, Galway, Ireland), has been introduced for impedance
planimetric description of mechanical properties of GI sphincters154–157. Based on the experience from our
institution, we chose impedance planimetric assessment with the EndoFLIP for description of mechanical
properties of resected small intestinal stenoses from patients with CD.
The length of stenoses in CD varies significantly. Therefore, we wanted pressure and not volume as the
independent parameter for distension. For that purpose, we had to bypass the EndoFLIP’s pumping mechanism,
as described in the methodology section. A limitation with the EndoFLIP probe is that it only detects the luminal
CSA. For advanced biomechanical evaluation the wall thickness has to be involved. Therefore, we combined
impedance planimetry with ultrasound allowing us to determine both the inner and outer CSA. Since change in
pH can influence muscle activity, we installed the specimen in a Krebs solution also allowing us for better US
visualisation. The point of the narrowest part of a stenosis was marked by acupuncture needles, since these are
very thin and only slightly affects the tissue with the additional benefit of being highly visible on
ultrasonography in a water bath. Thereby, we ensured that still images were taken at the same locations
regardless of pressure level.
29
Methods Ultrasonography Participants were investigated after four hours fast. An
Acuson S3000 ultrasound machine with a 4-9 MHz linear
matrix transducer and a 1-6 MHz curvilinear transducer
was used (Siemens Medical Solutions, Malvern, PA). CDI
was set with a transmit frequency of 6.75 MHz, gain 1 dB,
pulsed repetition frequency 1099, low wall filter of 2, and
a scale of 6 cm/s.
Bowel wall thickness was measured in continues numbers
at the thickest site of the bowel wall without including
valvula conniventes. Length of disease was measured
within the lumen, drawn in free hand as illustrated in
Figure 4. As a standard bowel US examination, we further
examined bowel wall pattern (layered; partly disturbed;
disturbed), CDI after the Limberg classification76,
inflammatory fat (none; barely visible; < 2 cm; > 2 cm),
presence of ulcers (none; possible/small (< 1 mm);
large/transmural), lymph nodes (present or not), and
diameter (less than 1 cm or larger), presence of abscesses,
fistulas and strictures with prestenotic dilatation, Figure 5
& 6.
Contrast Enhanced Ultrasonography The consistent pulse centre frequency was set to 4 MHz,
the dynamic range at 80, the mechanical index was 0.06-
0.08 fixed throughout the examination, the frame rate was
10 frames per second (fps), the focal zone was placed
beneath the bowel loop123; contrast bolus injection was 2.4
ml SonoVue followed by a 5 ml saline flush over 2 seconds
in an 18 Gauge iv. catheter in the cubital vein. Scans were
recorded for 90 seconds. Scan plane was kept constant and
Figure 4 B-mode ultrasonography image showing terminal ileum and a free-hand length measurement.
Figure 5 B-mode ultrasonography image showing bowel wall thickening and a severe prestenotic dilatation.
Figure 6 B-mode ultrasonography image showing deep ulcer and Limberg 4 color Doppler signals.
30
patients were instructed in gentle breathing. More than 5 minutes after first injection, the scan was repeated at
the same spot, but in a different scan plane to cover the segment in both the transverse and longitudinal axes.
The chosen bowel segments were either terminal ileum, neo-terminal ileum or proximal ileum. The location for
CEUS was determined as the most inflamed area of the segment according to a prior classification78,158. Cineloop
files were exported in DICOM format, re-linearized, and quantified on VueBox 5.1 (Bracco Suisse SA, Geneva,
Switzerland) as earlier described82,112. If possible, four ROIs were drawn using the following criteria: the largest
possible ROI (ROI1) which reliably included the full bowel wall without measuring the lumen or surrounding
fatty tissue. Subsequently three smaller ROIs were chosen and had to be within the area of the large ROI while
still > 0.1 cm2. Within these (ROI2, ROI3 or ROI4) we identified the “highest peak enhancement ROI” and a mean
value of the three smaller ROIs ((ROI2 + ROI3 + ROI4)/3). Shapes and placement of ROIs were optimized to
obtain a quality of fit (QoF) >90 % or as high as possible. Built-in motion compensation was applied whenever
beneficial, Figure 7. Analyses were then compared for repeatability between the largest ROI, the maximum peak
ROI and mean of the three latter ROIs, exhibiting a QoF >85%, entitled mean ROI. The average (log-converted)
values of the best reproducible method were subsequently chosen for comparison with DCE-MRE results. Data
post processing was badge-analysed by the same investigator more than six months after the US and clinical
scoring of the last patient to ensure effective blinding of data.
Figure 7 VueBox post-processing. Dual contrast enhanced ultrasonography image with contrast mode on the left side and b-mode on the right site. Turquoise delineation marks the border for analysis and serves as the area for motion compensation. Four ROIs are drawn, ROI 1 (green) as the largest possible ROI. First injection (top left) and corresponding time intensity curves (bottom left). Second injection dual image and ROIs (top right) and corresponding time intensity curves (bottom right). NB Y-axis scale differs. Table on bottom right shows quality of fit.
31
Magnetic Resonance Enterography Patients were instructed to fast for four
hours and drink one litre of oral contrast
one hour before the scan. Oral contrast
comprised a suspension of 125 ml
mannitol 15 % (Fresenius Kabi, Bad
Homburg, Germany) in 875 ml of tap
water, 30 ml psyllium HUSK® fibre, and ice
cubes. Peristalsis was suppressed by
intravenous injection of 20 mg Hyoscine
butylbromid (Buscopan®; Boehringer
Ingelheim, Ingelheim, Germany) prior to
non-dynamic sequences and repeated
before contrast injection. Images were
acquired using a 1.5 T MR unit (Avanto;
Siemens, Erlangen, Germany) with
patients in the prone position. The
intravenous contrast agent used was gadoterate meglumine (Dotarem®; Guerbet, Villepinte, France), 0.2 mg /
kg bodyweight at 5 ml/s followed by a 24 ml saline flush. Patients were instructed to hyperventilate prior to a
long breath hold followed by gentle breathing. MR scanning protocols can be seen in Appendix III.
In the first study, we did an interobserver correlation analysis for continues measures of maximum wall
thickness; continues measurements for length of disease involvement; type of enhancement pattern
(homogenous, layered, mucosal or absent); degree of oedema (based on bright mural T2 fat sat signal);
perimural T2 signal; presence or absence of fistulas, abscesses, lymph nodes, comb sign and ascites159. These
data were used to calculate the MEGS score, Figure 8. In article 2, only one reader examined the patient for the
same parameters as described above. Additionally, ulcers were classified and contrast signal intensity was
obtained to calculate the MaRIA score.
Dynamic Contrast Enhanced Magnetic Resonance Enterography The DCE-MRE analysis was performed within the same bowel segment examined by CEUS, using a RoiTool. ROI
selection was of moderate size encompassing the highest enhancement area within the bowel wall. No motion
compensation was applicable. Since the surrounding fatty tissue generally has a high signal intensity, all frames
Figure 8 MR Enterography T1 weighted sequence post contrast enhancement. Neoterminal ileum is clearly thickened and with homogenous enhancement. Additional overlapping images are from different slices
32
were assessed to make sure that the ROI stayed within the bowel wall. ROIs were deleted in frames where
significant out-of-plane motion occurred. Based on these ROIs, a MRI physicist fitted a cubic spline curve,
defined the baseline length and the peak, and calculated the TIC parameters from intensity measurements. All
data were then exported in comma separated files for further import into Stata.
Clinical assessment Patients were asked about their smoking habits and symptoms within the last period of flair. Symptoms
assessed were; pain, nausea, vomiting, diarrhoea, bloody stools, weight loss, and fatigue. Symptoms were
graded semi quantitatively into none, mild, moderate and severe. The duration of the present period of flair was
reported. CDAI and HBI (appendix I & II) were recorded at the time of inclusion.
Blood and stool Samples At the day of CEUS, patients had routine blood samples taken. These included ESR, haematocrit, haemoglobin,
CRP, leucocytes, vitamin D, and albumin. At first visit, patients were instructed in taking a faecal sample and
send it to the central laboratory of Randers Regional Hospital, certified for analysis of f-calpro.
Histopathology The bowel segment was rinsed immediately after surgical excision. To preserve the luminal diameter, we did not
open the bowel segment longitudinally, but occasionally long segments were divided into two or three
segments, see Figure 9 & 10. Fixation was made in neutral buffered formaldehyde for 24-48 hours using
intraluminal wicks to ensure sufficient preservation. After fixation, all segments were transversally cut into slices
of 3-6 mm thickness, Figure 11 & 12. Representative slices (e.g. with greatest wall thickness and smallest lumen)
as whole mounted slices where then paraffin-embedded for sectioning and Haematoxylin-eosin staining. Based
on histological assessment of all slices, the slice with greatest wall thickness, smallest lumen, highest
inflammatory sub score, and highest fibrotic content was selected for digitalization. If the above mentioned
parameters did not all occur in one slice, two slices were selected. From these two sites, four new slices were
made and batch stained with 1) Haematoxylin-eosin, 2) Masson’s trichrome, 3) anti-Desmin, respectively 4) anti-
CD34. Slices were digitalized with NanoZoomer 2.0 HT and NanoZoomer 2.0-RS (Hamamatsu, Hamamatsu City,
Japan). Maximum wall thickness was measured digitally in the dedicated software program NDP.View2
(Hamamatsu, Hamamatsu City, Japan)160. In each patient, both slices were scored using the acute inflammatory
index described by Borley et al.143 and fibrosis by Chiorean et al.147 and collagen by Baumgart et al.148, see table
2, 3 & 4.
33
Figure 9 Drawing of the intestine showing where the intestine is divided and the punctum maximum of disease.
Figure 10 Gross specimen
Figure 11 Full specimen cut in 3-6 mm slices
Figure 12 Close up of specimen
34
Impedance Planimetry The bowel wall was fixated at the in vivo length in a
water tank containing a Krebs solution (118 mmol/L
NaCl; 4.7 mmol/L KCl; 25 mmol/L NaHCO3; 1.0 mmol/L
NaH2PO4; 1.2 mmol/L MgSO4; 2.5 mmol/L CaCl2-H2O;
11 mmol/L Glucose; 0.11 mmol/L ascorbic acid161) at 37
°C, using an electric warming tray (Bartscher,
Salzkotten, Germany). To determine the optimal
position of the EndoFLIP probe, an initial short
distension was done with a low volume to identify the
narrowest part of the stenosis. Half a litre of the non-
disclosed EndoFLIP liquid was connected directly to the
EndoFLIP balloon and pressure was reset with the water level at the height of the probe. Pressure was
subsequently increased with 10 cm H2O increments until the maximum of 100 cm H2O. Pressure was
simultaneously measured by the EndoFLIP machine and ultrasound b-mode still images were taken at each
pressure level.
In Vitro Ultrasonography Acupuncture needles were inserted through the fatty tissue or the outer 0.5 mm of the bowel wall/serosa at the
narrowest part of the specimen as well as 1 cm and 2 cm proximal and distal from this point. A HI VISION Preirus
US machine with a EUP-L73S probe (Hitachi Medical Corporation, Tokyo, Japan) was used to obtain b-mode still
images at each site of needle placement and annotated with 0, +1, +2 or -1, -2 and the current pressure level.
Post processing
Dicom files were imported into OsiriX 5.7.1 64 bit (Pixmeo, Bernex, Switzerland) and the lumen were outlined
using an oval-shaped ROI, Figure 14. The outer border was likewise determined with an oval-shaped ROI. ROIs
from all locations at all pressure levels, containing information about area and circumference were exported in
csv-format and subsequently
imported into MATLAB for further
analysis of mean bowel wall
thickness.
Figure 13 Experimental set-up. Placed on a table heater, the water container with oxygenated Krebs solution and the specimen stretched to original length using magnetic clamps
Figure 14 Regions of interests to outline inner and outer bowel wall, drawn in OsiriX
35
Data handling and Statistics Case report forms were used for obtaining clinical data for study I and II. Data was entered into EpiData v. 2.0
(The EpiData Association, Odense, Denmark) for Study I and directly exported to Stata for further analysis. In
Study II and III data was entered into REDCapTM 162 and then exported to our statistical software. Analysed data
from VueBox (CEUS ROI data), MATLAB (DCE-MRE data, EndoFLIP, and CSA data), OsiriX (CSA ROI data) were
first exported into .csv files before further import. Statistical analyses and graphs were performed using Stata
for Mac 13.1 (Stata Corp LP, College Station, TX). In all studies, Gaussian distribution were checked using
quantile-quantile plots and histograms. In general, CEUS intensity data (e.g. peak enhancement, area under
curve, etc.) and CEUS/DCE-MRE time related data (rise time, fall time) were non-Gaussian, Figure 15.
Intensity data were logarithmic converted into dB as per standard ultrasonography conversion of linear echo
power data dB = 10 · log10 (AIU), where AIU is arbitrary intensity units, Figure 16. Time related data was log-
converted before further analysis. Biochemical data that did not follow a Gaussian distribution were
subsequently log-converted (CRP and f-calprotectin). Association analyses were performed with Spearman’s rho
between CEUS and DCE-MRE. Few parameters did however follow a normal distribution, but for the sake of
consistency, we only performed the Spearman’s correlation. Comparing two modalities examining the same
entity, using the same scale, were presented as 95 % limits of agreement (LoA) rather than Pearson’s correlation
analysis163–165.
We used the simple LoA calculation with the ”concord” command in Stata for individual measurements between
modalities (e.g. bowel wall thickness). When assessing multiple parameters for the comparison of ROIs for
Figure 16 Quantile quantile plot and histogram of Peak enhancement after log transformation. Sufficient Gaussian distribution.
Figure 15 Quantile quantile plot and histogram of Peak enhancement. Clearly left skewed.
36
repeatability in CEUS, a mixed effect model with individual residuals pr. ROI was used165. For reproducibility
measurements between readers of MR or between MR and US, Cohens kappa166 was used of categorized
variables and intra-class correlation coefficients167 (ICC) for continues numbers like bowel wall thickness and
length of disease and MEGS score. Comparing difference in mean wall thickness between modalities were
performed with a paired t-test. Sensitivity and specificity were computed using the “diagt” command and
compared between modalities using McNemar. We have not adjusted for multiple testing with a specific
method like the Bonferroni correction, thus p < 0.05 is regarded statistically significant. However, the magnitude
of the conclusions drawn are taking the p-value size into account.
Ethical considerations All studies were approved by the Danish National Authorities (2011-005846-36 & 2011-005886-19), the Regional
Committee on Health Research Ethics (1-10-72-339-12 & 1-10-72-340-12), and the Danish Data Protection
Agency (1-16-02-206-12 & 1-16-02-207-12). The studies complied with the Declaration of Helsinki and all
participants gave written informed consent prior to entering in the trials.
37
Figure 17 Bland-Altman plot of bowel wall thickness. On the left side difference between two MRE readers. On the right side difference between MRE and US
Summary of results
Paper I Dynamic Contrast Enhanced Magnetic Resonance Enterography and Dynamic Contrast Enhanced Ultrasonography in Crohn’s Disease: An observational comparison study
The study I patient population consisted of 25 patients with moderate to severe CD, all with detectable bowel wall thickness > 3 mm. The characteristics of the patient population can be seen in Table 5
Table 5 Patient Demographics
Parameter No. of Patients Included patients 25 Female 13 (52) Age, years 37 [19-66] Body mass index (kg/m2) 24.5 ± 4.4 Location of disease (Montreal)
Terminal Ileum (L1) 19 (76) Colon (L2) 1 (4) Ileocolonic (L3) 4 (16) Isolated upper disease (L4) 1 (4) Additional upper disease (+L4) 3 (12)
Crohn Disease Activity Index 298 ± 85 Harvey Bradshaw Index 9.9 ± 3.5 Faecal Calprotectin (μg/g)* 356 [63-3600] C-Reactive Protein (mg/l)* 5.9 [0.7-34.4]
Note – Numbers in parenthesis are percentages. Numbers in brackets are ranges. Unless indicated otherwise, data are means ± standard deviations. * Median values and ranges.
Pathoanatomical data
The thickest bowel wall segments had a mean of 7.9 ±
2.3 mm, (range 4 to 12 mm) when assessed with US
and 8.1 ± 2.8 mm (range 4 to 14.5 mm) when assessed
by two MRE readers. The mean difference between
modalities was 0.22 mm (95 % LoA -4.3 to 3.9) and the
corresponding ICC was 0.71 (95 % CI: 0.44-0.86, P <
0.001), Figure 17. The median length of the inflamed
segment was 15 cm (range 3 to 57 cm) on US and 12
cm (range 1 to 70 cm) on MRE. The corresponding ICC
was 0.89 (95 % CI: 0.76-0.95, P < 0.001).
38
Best Region of Interest repeatability
We tested the repeatability of CEUS with dual contrast injection and multiple ROI placements. Mean difference
between injections and the range of LoA were assessed. As seen in Table 6, There were in most cases a
statistically significant higher value obtained in the second injection compared to the first injection. However, in
most cases, this became insignificant if strict criteria for QoF were applied, Figure 18. Thus no difference were
seen between injections for Mean ROIs, except for area under curve. Although the Maximum Peak ROIs in
general had smaller mean difference between injections, the LoA were always narrowest for the Mean ROIs and
broadest for the Large ROIs. Based on these results we chose the Mean ROIs as the ROI selection of choice for
association analyses with MRE parameters. By definition, the Maximum Peak ROIs had the greatest difference
from the Large ROIs and in all but one case, Mean ROIs were also significantly larger than the Large ROIs.
Table 6 Repeatability of time intensity curve parameters, dynamic CEUS
CEUS parameter Region of Interest (ROI)
Mean difference between injection 1 and injection 2 P value Limits of agreement
Difference from Large ROI P value
Peak Enhancement Large ROI 1.36 dB (0.77 to 1.96) P < .001 [-4.0 to 6.8] dB reference NA Good QoF -0.14 dB (-0.66 to 0.38) P = .588 [-4.2 to 3.9] dB reference NA Maximum Peak ROI 0.63 dB (0.05 to 1.20) P = .032 [-4.4 to 5.7] dB 1.34 dB (0.93 to 1.75) P < .0001 Good QoF -0.49 dB (-0.89 to -0.08) P = .018 [-3.7 to 2.7] dB 1.78 dB (1.45 to 2.10) P < .0001 Mean ROI 0.73 dB (0.17 to 1.28) P = .010 [-3.8 to 5.3] dB 0.90 dB (0.50 to 1.30) P < .0001 Good QoF 0.24 dB (-0.13 to 0.61) P = .198 [-2.3 to 2.8] dB 1.18 dB (0.87 to 1.50) P < .0001 Area under curve Large ROI 1.46 dB (0.78 to 2.13) P < .0001 [-4.6 to 7.5] dB reference NA Good QoF 0.46 dB (-0.03 to 0.95) P = .068 [-3.4 to 4.3] dB Reference NA Maximum Peak ROI 0.18 dB (-0.34 to 0.71) P = .489 [-4.3 to 4.7] dB 0.88 dB (0.46 to 1.30) P < .0001 Good QoF 0.16 dB (-0.32 to 0.63) P = .515 [-3.5 to 3.8] dB 1.31 dB (0.98 to 1.64) P < .0001 Mean ROI 0.64 dB (0.15 to 1.13) P = .010 [-3.3 to 4.6] dB 0.32 dB (-0.09 to 0.73) P = .122 Good QoF 0.75 dB (0.32 to 1.17) P < .001 [-2.2 to 3.7] dB 0.79 dB (0.48 to 1.11) P < .0001 Wash-in Rate Large ROI 1.41 dB/s (0.74 to 2.09) P < .0001 [-4.7 to 7.6] dB/s Reference NA Good QoF -0.59 dB/s (-1.09 to -0.08) P = .023 [-4.6 to 3.4] dB/s Reference NA Maximum Peak ROI 0.90 dB/s (0.20 to 1.59) P = .011 [-5.2 to 7.0] dB/s 1.54 dB/s (1.06 to 2.02) P < .0001 Good QoF -0.68 dB/s (-1.13 to -0.24) P = .003 [-4.2 to 2.8] dB/s 1.94 dB/s (1.61 to 2.27) P < .0001 Mean ROI 0.61 dB/s (-0.00 to 1.23) P = .051 [-4.4 to 5.6] dB/s 1.17 dB/s (0.72 to 1.62) P < .0001 Good QoF -0.16 dB/s (-0.53 to 0.21) P = .393 [-2.8 to 2.4] dB/s 1.35 dB/S (1.04 to 1.66) P < .0001 Wash-in Perfusion index Large ROI 1.34 dB/s (0.75 to 1.93) P < .0001 [-4.0 to 6.7] dB/s Reference NA Good QoF -0.13 dB/s (-0.64 to 0.38) P = .616 [-4.2 to 3.9] dB/s Reference NA Maximum Peak ROI 0.57 dB/s (0.01 to 1.14) P = .045 [-4.4 to 5.5] dB/s 1.31 dB/s (0.91 to 1.72) P < .0001 Good QoF -0.50 dB/s (-0.90 to -0.10) P = .016 [-3.7 to 2.7] dB/s 1.74 dB/s (1.42 to 2.06) P < .0001 Mean ROI 0.71 dB/s (0.17 to 1.26) P = .011 [-3.8 to 5.2] dB/s 0.87 dB/s (0.48 to 1.27) P < .0001 Good QoF 0.25 dB/s (-0.12 to 0.61) P = .191 [-2.3 to 2.8] dB/s 1.16 dB/s (0.85 to 1.47) P < .0001
Note – Numbers in parenthesis are 95 % confidence intervals. Numbers in brackets are 95 % Limits of agreement. ROI = region of interest, QoF = quality of fit, CEUS = Contrast enhanced ultrasonography.
39
Associations between perfusion data from modalities
Total area under curve, including wash-in
and wash-out for CEUS and wash-in and
plateau-phase at 70 seconds for DCE-
MRE, had a low and insignificant
correlation between the two methods (r =
0.16, P = 0.494). Wash-in area under
curve also showed poor correlation (r =
0.18, P = 0.443). Rise time and time to
peak showed no correlation between
modalities (r = 0.11, P = 0.659 and r =
0.02, P = 0.930 respectively) either. Slope
and maximum slope for DCE-MRE and
wash-in rate for CEUS correlated moderately
well (r = 0.60, P = 0.005, and r = 0.62, P =
0.004), Figure 19.
Peak intensity and Wash-in perfusion index
determined by the two modalities were
moderately and moderate to weakly
associated (r = 0.59, P = 0.006 and r = 0.47, P
= 0.036 respectively). No significant
correlation was found between peak
enhancement of CEUS and by DCE-MRE (r =
0.41, P = 0.076).
Figure 18 Bland-Altman plot of Peak enhancement between 2 injections. On the left side only scans with good quality of fit. On the right side all scans are shown
Figure 19 Scatter plot and best fitted line forcorrelation between Maximum Wash-in Rate between DCE-MRE and CEUS
40
Paper II Validity of Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced MR Enterography in
the assessment of Crohn’s Disease
In study II, we recruited 25 patients with known CD that were referred to elective surgery for small bowel
inflammation, stricture or penetrating disease. All patients had bowel wall thickness > 3 mm detectable on US.
Patient characteristics are seen in Table 7
Table 7 Demographics
Parameter No. of Patients Included patients 25 Female 17 (68) Age, years 37 [19-66] Body mass index (kg/m2) 24.8 [20.2-32] Location of disease
L1, Terminal Ileum 17 (68) L2, Colon 0 (0) L3, Ileocolonic 6 (24) L4, Upper disease 2 (8)
Crohn Disease Activity Index 202 ± 94 Harvey Bradshaw Index 7.5 ± 4.4 Faecal Calprotectin (μg/g)* 340 [30-3600] C-Reactive Protein (mg/l)* 4.3 [0.5-29.4] Haemoglobin (mmol/l) 8.6 ± 0.9 Albumin (g/l) 34.9 ± 4.8 Time between US and Surgery, days* 7 [1-26]
Note – Numbers in parenthesis are percentages. Numbers in brackets are ranges. Unless indicated otherwise, data are means ± standard deviations. * Median values and ranges. US = Ultrasonography
Pathoanatomical data
Histopathological measured bowel wall thickness was 8.7 ± 1.6 mm, measured with US bowel wall thickness was
9.1 ± 2.1 mm, average difference of 0.4 (-0.3 to 1.0) mm thicker, p = 0.238 (95 % LoA -2.7 to 3.5 mm). MRE
measured 10.0 ± 2.6 mm, with an average difference of 1.4 (0.4 to 2.3) mm compared to histopathology, p =
0.006 (95 % LoA -3.0 to 5.7 mm). Ulcers defined as linear, rake, confluent, or large were present in 19 patients
on histopathology. Sensitivity and specificity for ulcer detection by US and MRE are shown in Table 8. Patients
without ulcers had a thinner bowel wall of 7.0 mm (5.9 to 8.2) on histology, 7.0 mm (5.0 to 9.0) on US and 8.3
mm (4.0 to 12.7) on MRE compared with histological proven ulcers revealing wall thickness of 9.2 mm (8.0 to
9.4, p = 0.002) on histology, 9.7 mm (8.9 to 10.6, p = 0.003) on US and 10.5 mm (9.6 to 11.4, p = 0.080) on MRE.
41
Table 8. Accuracy for ulcer detection on cross sectional imaging.
Parameter Sensitivity Specificity Accuracy PPV NPV P Ulcers on US 18/19 (94.7 %)
[74.0 - 99.9 %] 4/6 (66.7 %) [22.3 – 95.7 %]
80.7 % [59.4 – 100 %]
90 % [68.3 – 98.8 %]
80 % [28.4 – 99.5 %]
P = 0.005
Ulcers on MRE 15/18 (83.3 %) [58.6 – 96.4 %]
2/6 (33.3 %) [4.3- 77.7 %]
58.3 % [35.9 – 80.8 %]
78.9 % [54.4 – 93.9 %]
40 % [5.3 – 85.3 %]
P = 0.366
Difference P = 0.625 P = 0.5 P = 0.219 US = Ultrasonography; MRE = Magnetic Resonance Enterography; PPV = positive predictive value; NPV = negative predictive value; Nominator is True positive or true negative and denominator is total ulcers on histology, Brackets are 95 % confidence interval.
Table 9 Correlation between relative perfusion and histological, biochemical and clinical markers of activity
DCEUS parameter Acute inflammation Fibrosis f-calprotectin CRP CDAI HBI P
Peak enhancement NS NS 0.48 NS NS NS 0.021§* Wash-in Wash-out area under curve
NS NS 0.49 NS NS NS 0.017§*
Wash-in area under curve NS NS 0.46 NS NS NS 0.029§* Wash-out area under curve
NS NS 0.51 NS NS NS 0.013§*
Wash-in perfusion index NS NS 0.48 NS NS NS 0.021§* Wash-in Rate NS NS 0.44 NS NS NS 0.034§* Rise time NS NS -0.02 NS NS NS 0.932§ Fall time NS NS 0.13 NS NS NS 0.547§ Mean transit time local NS NS 0.39 NS NS NS 0.068§ DCE-MRE parameters Rise time NS NS NS NS 0.11 NS 0.629# Peak enhancement NS NS NS NS -0.46 NS 0.026#* Area under curve wash-in NS NS NS NS -0.19 NS 0.406# Area under curve 25 s NS NS NS NS -0.49 NS 0.019#* Area under curve 70 s NS NS NS NS -0.56 -0.43 0.006#** / 0.040†* Initial slope NS NS NS NS -0.41 NS 0.052# Maximum slope NS NS NS NS -0.30 NS 0.157# Wash-in perfusion index NS NS NS NS -0.36 NS 0.089#
§P value for Calprotectin, #P value for CDAI, †P value for HBI, * P < 0.05, ** P < 0.01
DCEUS = Dynamic Contrast Enhanced Ultrasonography, DCE-MRE = Dynamic Contrast Enhance Magnetic Resonance Enterography, CRP = C-reactive protein, HBI = Harvey Bradshaw Index, CDAI = Crohn’s Disease Activity Index, NS = Not significant
42
Relative perfusion
A total of 187 ROIs were drawn on CEUS with a median
QoF of 95 % (range 65.2 to 99.6 %). Predefined good QoF
> 90 was obtained in 90.9 % of ROIs. There was no
statistical significant correlation between any CEUS or
DCE-MRE based perfusion parameters and the acute
inflammation index by Borley143 or the fibrosis index by
Chiorean147, p > 0.38. For secondary endpoints, there
were weak to moderate correlations between CEUS
intensity parameters and f-calpro. However, no statistical
significant correlation was present between CEUS and
CDAI, HBI or CRP. DCE-MRE showed a moderate to weak
was statistical significant inverse correlation to CDAI for
peak enhancement and area under curve at 25 and 70 seconds,
Table 9
Post-hoc analyses of MR enterography global score, MR index of activity and bowel wall thickness
In a post-hoc analysis, the histology index of activity
correlated moderately well with MEGS by Makanyanga et
al.61, (r = 0.53, p = 0.008) and with bowel wall thickness
measured by US, (r = 0.61, p = 0.001) but not with MaRIA
score (r = 0.17, p=0.434). However, MaRIA and MEGS
both correlated weakly with the semiquantitative score of
collagen depth (r = 0.47 p = 0.022 and r = 0.45, p = 0.027,
respectively). Bowel wall thickness showed a moderate
correlation with collagen depth (r = 0.53, p = 0.006).
Figure 14 Scatter plot and best fitted line for invers correlation between DCE-MRE AUC and CDAI. Red lines indicate levels of moderate and mild clinical disease severity.
Figure 13 Scatter plot and best fitted line for correlation between AUC and f-calprotectin
43
Paper III Can Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced MR Enterography
predict the stiffness of a Crohn´s Disease stricture?
Of the 25 patients included in the original study
undergoing elective small bowel surgery, EndoFLIP data
were available for 18.
Stiffness of the stricture and Ultrasonography or MR
enterography
Young’s modulus (E) for the stricture correlated
moderately well with the grading of prestenotic
dilatation on US, (r = 0.60, P = 0.009), but not on MRE,
(r = -0.10, P = 0.704), Figure 15. Strictures classified as
“certain” on US were stiffer, E: 19.0 kPa (14.0 to 25.8)
than those classified as “no” or “uncertain” stricture, E:
12.6 kPa (10.5 to 15.1), P = 0.026. No association was
found between certainty of strictures on MRE and E:
15.4 kPa ((12.8 to 18.6) for certain strictures and 16.6
kPa (8.9 to 31.0) for uncertain/no stricture, P = 0.726),
Figure 16. Patients with large ulcers on US had a trend
towards also having a stiffer bowel wall, E: 17.5 kPa
(13.6 to 22.6) than those with no ulcers, E: 12.1 kPa (9.6
to 15.3), P = 0.082. Normal (layered) versus disturbed
bowel wall pattern on US was not associated with
stiffness E: 17.2 kPa (12.0 to 24.7) and E: 14.2 kPa (12.0
to 18.8) respectively, P = 0.291. In contrast, increasing
levels of bowel oedema on MRE were associated with E
of the stricture (r = 0.53, P = 0.025). The two widely used
MRE scoring indices performed somewhat differently.
Whereas no correlation was found between E and MaRIA (r = 0.22, P = 0.376), there was a moderate association
between severity of MEGS and E (r = 0.66, P = 0.003).
Figure 16 Associations between stiffness and certainty of stricture on ultrasonography (left) and MR enterography (right). Boxes are inter-quartile ranges and medians. * P < 0.05
Figure 15. Stiffness and degree of prestenotic dilatation. Associations between stiffness of the stricture and the degree of prestenotic dilatation on ultrasonography (left) and MRE (right). Boxes are inter-quartile ranges and medians. * P < 0.05
44
Stiffness of the stricture and Dynamic
Contrast Enhanced MR Enterography or
Ultrasonography
The enhancement pattern of DCE-MRE
showed a weak but statistically significant
association to E of the stricture (r = 0.48, P
= 0.042). A moderate correlation was found
between E and DCE-MRE Initial slope of
enhancement (r = 0.72, P = 0.001), Figure
17. No significant correlation was found
between E and any of the CEUS
parameters. For all correlation parameters,
see Table 10.
Table 10 Correlation between Young’s modulus and imaging
Parameter Spearman’s rho P value CEUS Peak Enhancement 0.061 P = .8153 Area Under Curve -0.054 P = .8372 Wash-in Rate 0.179 P = .4920 Wash-in Perfusion index 0.061 P = .8153 Rise Time -0.368 P = .1466 Fall Time -0.245 P = .3430 DCE-MRE Peak Enhancement 0.113 P = .6666 Area Under Curve 70 s -0.208 P = .4223 Initial Slope 0.718 P = .0012 Initial Slope Max 0.409 P = .1028 Wash-in Perfusion Index -0.115 P = .6598 Rise Time -0.487 P = .0477 Wash-out Slope 60 s 0.069 P = .7935 Wash-out Slope 120 s -0.142 P = .5863
CEUS = Contrast enhanced ultrasonography, DCE-MRE = Dynamic Contrast Enhanced MR Enterography.
Figure 17 Stiffness of the stricture is associated with the initial slope of Dynamic Contrast Enhanced MR Enterography (DCE-MRE) (right) but not with wash-in rate of Contrast Enhanced Ultrasonography (CEUS) (left).
45
Discussion
Technical considerations in relative perfusion measurements Contrast Enhanced Ultrasonography Bolus injection is by far the most applied technique in perfusion measurements of the bowel78,82,109,116,121,168–170.
It is the preferred method of choice over the infusion administration and burst-replenishment techniques. Both
administration techniques will allow for measurement of relative perfusion only. Therefore, arterial input
function needs to be taken into account usually by the process called deconvolution118. Absolute perfusion
measurements allow reliable comparison between patients. Recently, Jirik et al.171 presented a new combined
approach for deconvolution and the bolus injection method. By this technique, the time intensity curve is
captured under circumstances comparable with our methodology. However, the authors describe the benefit of
bursting the bubbles be the means of a high mechanical index scan after 60 seconds and measure the time until
steady state is regained. This allows for calculation of the MTT and has the benefits of the infusion-burst-
replenishment without the time and contrast consumption of the infusion until steady state. The main issue
with this modality, is the low and varying concentration of contrast after 60 seconds which makes it particularly
prone to a low signal to noise ratio. The model assumes a constant or almost horizontal intensity, and small
fluctuations can change the fitted curve. Further, two injections have to be performed, and thus another level of
uncertainty is added, since the contrast volume and thus injection time differs. Iliac artery is chosen as the
artery of interest as it is often localized close to the terminal ileum. However, the iliac artery (acting as a global
AIF) may be perfused differently than the superior mesenteric artery (acting as a regional AIF) and again
differently from a local intestinal artery (local AIF)67. The authors mention, that the greatest issue is likely to be
the scaling procedure. At the time of planning our study, we did not have access to the program, that could
calculate and compensate for all these assumptions. To our knowledge, the reproducibility of procedure has so
far not been tested in patients with IBD. We hoped to overcome some of the limitations by using the abdominal
muscle as a reference. Since patients are in the resting state, low standardized flow should occur in this muscle.
However, at the time of badge analyzation, we soon realized, that this approach was very uncertain. Since the
perfusion measured in the very nearfield is very low and the transducer is not optimized for detecting flow in
the near field, the overall signal intensity is almost negligible and the horizontal orientation of muscle fibres
generated significant tissue artefacts172. The consequence is an extremely broad range of measurements
depending on a minimum change in reference perfusion173. We therefore discarded this option.
46
The relative perfusion method, that we applied, has a considerable amount of challenges asides from
lacking the arterial input function. Injection was performed manually, without a pump, which can influence the
comparability between patients, since a slight difference between injections could occur. Intensity is measured
in arbitrary units based on the individual machine setting. This prevents comparison with other machines, and
even small changes within the same machine (e.g. mechanical index or transmit central frequency) will result in
large variation of results174. Tissue motion was corrected by the build in motion compensation functionality.
However, this was not always working well, and some fluctuations occurred due to breathing artefacts.
However, as we achieved an overall very good curve fit the error may be of limited influence. Motility could not
be compensated for, and large curve deviations happened. This is also one of the main factors, that influence
the ability to investigate the normal or near-normal bowel wall with full motility174.
We tested the repeatability for our quantification method, and found that the mean value of three ROIs was
most reproducible. The repeatability was further improved by restricting good QoF for the curves used. This
seems meaningful, since a poor QoF may be caused by movement artefacts, peristalsis or heterogeneity of the
tissue examined, just like a ROI containing the air-filled lumen, sub-serosal tissue or large pulsating vessels.
Our sample size was too small to investigate the effects of medication taken before surgery, but medication
affect perfusion82.
Dynamic Contrast Enhanced MR Enterography Contrast enhancement has been used in most MRE studies. The degree of enhancement can be investigated in
several different ways ranging from a qualitative interpretation of enhancement compared with nearby normal
bowels, over semi quantitative or quantitative methods using ROIs. Pattern of enhancement is an element of
some semiquantitative scoring systems61 and it can even change over time175. The number of acquisitions after
contrast injection can vary from one to several hundreds. We chose to investigate the more advanced version of
intestinal contrast enhancement with MR than the simple baseline and post contrast acquisition. This is a well
described method, which is commonly used in cerebral perfusion analyses where the MRCA does not cross the
blood brain barrier67. However, even here, there are several challenges with correctly measurements of blood
flow. Should the AIF measured originate from a global AIF, a regional AIF or a local AIF, with the latter
minimizing bolus dispersion compared to the tissue of interest? Is the best measurement obtained from inside
the vessel, or just outside the vessel owing to the fact of partial volume effect?67
We examined the possibility of using AIF from the abdominal aorta, common iliac vessel of superior mesenteric
artery. Additionally, we examined the psoas muscle as a tissue reference instead of the AIF. As for the AIF, we
47
applied the variable flip angle technique and achieved unreliable results for all arterial measurements. For the
tissue reference, the correlation was even worse, since the psoas muscle often was localized in a different slice
and the MR field inhomogeneity and noise caused too large discrepancy compared with the bowel ROI.
Another limitation in our DCE-MRE study, was that we only used one ROI size and one investigator, thus not
testing the intra-rater and inter-rater variability for relative perfusion measurements, which is known to be
large176. The main reason for this, was the very time-consuming process for drawing ROIs that were only located
inside the bowel wall. In most cases, there were large displacement of structures and the ROIs had to be
manually moved in most of the 120 frames.
In order to achieve the most reliable initial curve, patients were instructed to hyperventilate prior to the
dynamic scan allowing for a longer breath hold. Injecting contrast before breath hold was instituted, gave a
smoother curve around the peak and the time after the peak, but sacrificed a long baseline, also leading to
some uncertainty.
Finally, the variation of flip angles within the observational field may cause large errors and varying noise levels
for absolute measurements. Wang et al. suggested that the signal difference strategy may be equally applicable
for tissues in studies investigating individual patient follow-up as the full pharmacokinetic modelling133. Overall,
the DCE-MRE methodology is based on many assumptions that are hard to govern. In a review by Makanyanga
et al.68 the association between enhancement assessment and disease activity was consequently very
inconsistent, as was the reference standard. The only absolutely quantitative method that used transmural
specimen histology as a reference did not show any correlation with disease activity150.
Evaluation of Reference standards Clinical Disease Activity Indices CDAI is widely used in clinical practice and scientific studies7. For study I, we chose to incorporate CDAI above
220, corresponding to moderate clinical disease activity, as inclusion criterion to ensure significant disease.
However, CDAI is commonly regarded as poorly correlated to objective findings of inflammation19. We also had
to exclude some patients without sufficient clinical disease activity, that had clear evidence of mural
inflammation on US. Correspondingly, there was no relevant association between CDAI and relative perfusion
parameters. Very surprisingly, some parameters correlated inversely with DCE-MRE. We chose to include
patients with HBI above 7, regardless of the CDAI level. This was a pragmatic choice allowing us to recruit slightly
more patients. HBI is prone to the same shortcomings as CDAI, and the main advantage over CDAI is the lack of
a seven-day bowel movement and symptom diary. However, our best guarantor for excluding non-existing
48
objective inflammation was our cut-off value at 3 mm bowel wall thickness. A few patients had moderate to
severe disease activity without any objective findings. They may likely have had other reasons for symptomatic
disease (e.g. constipation or bile acid malabsorption) but also the US could have been falsely negative causing
selection bias.
Biochemical There are several benefits by the use of laboratory markers; they are cheap, reproducible among various
laboratories, rapid and minimally invasive139. We chose to incorporate two biochemical markers as our
secondary association outcome. With focus on the two most acknowledged37,177; CRP and f-calpro. CRP is a
systemic acute phase protein with a short half-life of 19 hours. CRP correlates with objective disease activity to
some extend19,139 and a higher CRP before initiating treatment may respond better to infliximab than patients
with a lower CRP level178. Although potentially helpful to guide clinicians, it is still just a systemic marker and
elevation can be caused by many inflammatory conditions in addition to CD. Sensitivity and specificity has thus
been reported to be in the range of 50 – 100 % and 65 – 100 % respectively177. F-calpro is a sensitive and stable
marker of neutrophils in the gut. With reported diagnostic values of 76 - 100 % sensitivity and 83 – 95 %
specificity177, and correlates well with SES-CD19,37. It is often regarded a good marker for disease activity with the
ability to predict relapse after surgery38 and depending on the cohort and cut-off value for mucosal
inflammation very high negative predictive values are seen36. We found a moderate correlation between CEUS
AUC and f-calpro in our study and, as only two patients had a negative f-calpro, sensitivity was 92 %. The
limitation of f-calpro is the lack of ability to determine site of inflammation, presence of complications and the
cut-off of choice (50, 100, 150, 250 or 300 μg/g) is still debatable.
Histopathological references Histopathological references are often regarded as gold standard for transmural disease classification, and thus
our main reference of choice. Features of inflammation and chronicity are well described90,179. However, in CD
no validated or recommended scoring system exists for active inflammation in transmural specimens. Several
authors have suggested individual scoring systems, often for use in their own specific trials29,93,101,143,147,148,180–186.
Most of the scores are semiquantitative and without weighting of individual factors or the intra-factor
increment score. Some authors used quantitative approaches to quantify fibrosis185 or specific leucocytes182 (e.g.
macrophages, granulocytes etc.), while others have applied immunohistochemical staining allowing for
manually calculating micro vessel density71,184. We aimed at histomorphometric analysis of collagen, smooth
muscle and micro vessel density, but ran into technical difficulties, time and money constraints leaving us with
the only option to perform the semiquantitative scoring by Borley et al.143 and Chiorean et al.147. Although these
49
scoring systems have been used in several publications29,56,71,142,187, we find them somewhat inadequate. Often,
score 1 and 3 can be clearly excluded, however, the description of score 2 is not appropriate either. The score by
Borley et al., was mainly developed for showing changes of inflammation over distance within an affected bowel
segment143. These limitations are likely to influence our correlation analyses with relative perfusion
measurements. Although Romanini et al.184 finds a 100 % sensitivity and 100 % specificity for active CD with a
cut-off of 265 vessels per 10 high-power fields (40x) in colonic biopsies and Knod et al.188 also finds increasing
micro vessel density with inflammation, there are still conflicting results regarding the correlation analyses for
neoangiogenesis, perfusion and active inflammation on histology in CD71,189–191. Micro vessel density or
angiogenesis may even occur when local hypo-perfusion or ischemia exist190,191. Angiogenesis, nevertheless is
coexisting also with chronic inflammation, although it remains unknown if angiogenesis is causing inflammation
or the result of it192.
Impedance Planimetry To overcome some of the limitations in the histological outcome-correlation, we also added biomechanical
properties utilizing the impedance planimetry modality. Our overall strength is that we apply circular
intraluminal pressure with CSA measurements reflecting human pathological conditions of strictures rather than
the longitudinal stress and strain analysis described earlier148,193,194. Keeping the bowel at the most natural
environment at oxygenated 37 °C Krebs buffer solution at all times minimized muscular contractions due to pH
shifts161. Although we tried to mimic natural conditions there were several challenges in our method.
The water column was connected directly to the EndoFLIP probe, bypassing the build in volume neutral system.
Although the pressure measurement corresponded to the height of the water column in most cases, some
experiments had to be stopped around 70 cm H2O, since measurements became unreliable after this point and
some of the experiments failed all together. Consequently, our total sample size was relatively small. Although
the probe was fixated in both ends, minor movements could occur during increased pressure levels. This only
affected the EndoFLIP CSA measurements and not the US and could be corrected for at the time of analysis in
MATLAB. The total setup took between two and three hours to complete. Although water was kept at 37 °C at
all times, the long time period might have had some minor unwarranted influences on the specimen during the
examination and potentially related to fixation. The acupuncture needles were chosen to minimize tissue
damage and ensure scanning at the same location at all times. However, the needles were so thin, that mainly
the non-fixated handle of the needle were detectable by the US. Hence there could have been very minor
inconsistencies in the location of US still images. Further, the scan was made in the transverse plane of the
50
bowel but minor deviation from this could occur making the lumen more ovoid than perfectly circular. CSA
measurements were performed in a DICOM reader software program utilizing circular/ovoid ROIs. Although US
scans in a water solution generally exhibits good quality scans, the EndoFLIP balloon were in some instances
hampering the view of the posterior delineation, making the ROIs slightly imprecisely. Although we utilized
curve-fitting in MATLAB, the visual impression of QoF was not perfect. This is likely to be caused by the above
mentioned methodological challenges. We planned to measure up to five locations within the bowel specimen.
However, at the end we only compared the narrowest part with US and MRE findings.
General study limitations Statistical
Even though CEUS has been investigated in several prior studies, the relative nature of the methodologies used
and the diversity in utilized systems, settings, probes and even post acquisition analyses hindered us in a
sufficient power calculation for our studies. Lack of significant associations may thus be affected by our small
sample size, causing a type II error. On the other hand, we did multiple analyses in our study without
subsequently correcting the significance level (e.g. by Bonferroni correction). Some statisticians are against
these corrections, since they may be too conservative and thus may discard true significant results (type I
error)195. The general advice from our statistician, was not to correct for multiple testing, but to conclude the
strength of the results based on the level of p-value.
Another statistical limitation is our use of Spearman’s correlation rather than Pearson’s correlation. Although we
were hoping for a direct linear correlation, not all of our measurements were following a Gaussian distribution,
and thus the assumptions behind Pearson’s correlation analyses was not fulfilled. We therefor chose to use
Spearman’s correlation throughout all correlation analyses to ensure consistence of analysis between
modalities. When comparing two measurements using the same unit of measure, the use of simple correlation
analysis is inappropriate to measure modality agreement164. We therefore applied limits of agreement as the
appropriate measure. In our analysis of limits of agreement between CEUS ROIs, we applied the more advanced
mixed effect model as suggested by the Quantitative Imaging Biomarkers Alliance of the Radiology Society of
Northern America165. Another strength in our analyses are that we convert the CEUS intensity units to dB, since
arbitrary intensity units do not follow a Gaussian distribution. This fact in unfortunately not recognized by most
of the quantitative CEUS community.
51
Overall study outcomes Paper I: Structural changes between US and MRE are identical
We found an overall good agreement between pathoanatomical data description by US and MRE. This is not
surprising, since many meta-analyses are reporting equal diagnostic accuracy between these modalities52,55.
However, in our study, we investigated the continues numbers for bowel wall thickness and length of
involvement, and not only presence of disease; Yes/No. US is often regarded as user dependent, but so is the
interpretation of MRE. This is illustrated in paper I, where the LoA between US and MRE is virtually the same as
those for two MRE readers investigating the exact same images.
Relative TIC perfusion parameters are correlated between CEUS and DCE-MRE for the initial part of the TIC When investigating the initial part of the TIC, we mainly found a moderate association between wash-in rate of
enhancement and wash-in perfusion index between CEUS and DCE-MRE. We investigated the same patients
using both modalities, thus minimizing the influence of the AIF, which should be the same for the same patient.
However, the contrast agents used and the administration technique combined with the bolus size may likely be
the most important factors limiting associations. Further, we did not perform repeatability measures in DCE-
MRE to test the robustness of this method. However, to our knowledge, this is the first study to compare
relative perfusion measurements with high temporal resolution between these two modalities.
ROI selection affects the repeatability in CEUS analysis We show, that size and selection of ROIs clearly influence the outcome measurements and consequently should
be standardized. Neither applying the greatest possible ROI nor a ROI within the highest peak gave the best LoA
and we suggest the use of three ROIs with optimized curve fit with a QoF of minimum > 85 % and a minimum
size of 0.1 cm2. Three ROIs have earlier been suggested129,196. However, the benefit of this has not previously
been tested against a single ROI approach.
Paper II Structural changes on US and MRE can correctly identify histopathological changes In this study, we again tested the structural changes observed on US and MRE. This time, we applied the
transmural histopathological specimen as gold standard for the ability to grade findings on imaging. We found,
good agreement between wall thickness on both US and MRE, however with US performing slightly better. This
is not surprising, since the spatial resolution of US is superior to that of MRE. Again, to our knowledge, this has
not been investigated by others. Martínez et al.197 investigated difference between bowel wall thickness
measured by US and MRE and also found that MRE was measuring a thicker bowel wall than US. Unfortunately,
52
they did not use a reference standard in their study. Ulcers generally had a good accuracy, but we did not test
the intra- or inter-observer variability of our findings. Reproducibility has been high in previous studies56,131.
However, in one study131 there was a prevalence of 45 % and in another study56 of 0 %. Since our study
population was planned for surgery, we had a very high ulcer prevalence of 76 %.
Relative TIC perfusion parameters obtained by CEUS and DCE-MRE are correlated with histopathological changes Surprisingly, we were not able to show an association between TIC summary statistics and histological indices.
This may be due to the limitations earlier mentioned, perhaps in combination with type II error, lack of AIF and
limitations in both the technical performance of DCE-MRE and CEUS combined with poor histological scoring
systems. However, controversy exists about associations with histology content using individually modified
scoring systems71,115,198. We found an association between CEUS and f-calpro suggesting that there is indeed a
relation between perfusion and active inflammation. Bowel wall thickness is the best marker, so far, to correlate
with both histological activity and chronicity. However, this does not help in distinguishing between the
spectrum of coexisting fibrosis and active inflammation.
Paper III Structural changes on US and MRE can predict bowel wall stiffness
The certain identification of a stricture on US is associated with a significantly higher stiffness of the stricture
than no or uncertain strictures. This was not the case for MRE. Further, grading prestenotic dilatation also
associated with stricture stiffness on US but not the applied grading on MRE. The large difference between
these two modalities in this regard is probably the difference in preparation. Before MRE, patients had to drink
1 l of solution while patients were fasting before US. In contrast, oedema on MRE and the MEGS associated
moderately well with stricture stiffness. Since the latter is a compound score of many elements, including the
oedema score, this might be a type I error. That oedema correlates with stiffness is somewhat surprising, since
oedema normally is associated with active inflammation rather than fibrotic disease58,60. A Limitation to our
study, is the lack of a valid histological correlation with our stiffness measurements.
Relative TIC perfusion parameters obtained by CEUS and DCE-MRE can predict bowel wall stiffness. In our relative perfusion analyses, we showed a moderate association between stiffness of the stricture and the
Initial slope of DCE-MRE enhancement only. Other DCE-MRE parameters were either weakly or insignificantly
associated with stricture stiffness. This finding is surprising and in contrast to our hypothesis. A possible
explanation could be that active inflammation and fibrosis are present in the same strictures.
53
Conclusion
In patients with CD and inflammation of the small intestine
1) there is good agreement between structural findings on US and MRE, especially with regard to bowel
wall thickness and length of involvement 2) there is only poor to moderate association between relative TIC perfusion parameters obtained by CEUS
and DCE-MRE 3) structural findings US and MRE show good agreement with histopathology for wall thickness and good
sensitivity for ulcers 4) there is no association between relative TIC perfusion parameters and semiquantitative scores for
inflammation and fibrosis 5) there is moderate association between prestenotic dilatation and “certain strictures” determined by US
stiffness of the stenosis 6) the initial slope of enhancement on DCE-MRE is directly associated with stiffness of the stenosis 7) ROI selection influences TIC parameters, but are repeatable if strict criteria are applied.
54
Further perspectives
Objective measurements in CD are needed for classification of disease at time of diagnosis and for follow-up on
disease activity. Since endoscopy has some limitations, non-ionizing cross sectional imaging, like US and MRE,
are obvious modalities. In the present thesis evidence is provided for associations between structural changes
found on both modalities and histology. With the fast improvement of the technologies, spatial resolution,
temporal resolution, image clarity, portability and accessibility will improve and as price is likely to continue
decreasing, it must be expected that both US and MRE will gain even more use in the years to come. In contrast
to the basic structural parameters, measuring and computing absolute and quantifiable perfusion in a complex
organ like the bowel is still difficult and there is a lack of standardization of protocols. Applied techniques,
equipment, procedures and outcomes in published studies are highly diverse. Although systematic reviews have
been made199, the routine use of perfusion measurements in CD -regardless of modality used is still in its
infancy. Questions like where in the bowel to place the ROI, at what transducer pressure200 and how to correctly
perform deconvolution are just about to be investigated.
Many of these shortcomings may be limited by investigating change in relative perfusion in the same individual
over time201. Although a patient may experience weight gain or weight loss, more or less stress influencing
cardiac output, many influencing factors will be kept constant. The rate of change in perfusion may also occur
earlier than structural changes like wall thickness, indicating successful treatment202.
We recommend a multi institutional and multidisciplinary collaborative approach for dynamic perfusion
measurement using CEUS with translational approach from high-end phantoms to healthy volunteers and finally
pathology in a cross organ approach. Unfortunately, Volume Blood Flow was chosen over DCE US by the
Quantitative Imaging Biomarkers Alliance (QIBA) committee as their next project of standardization203.
Finally, this thesis further shows the complexity of CD with remarkably continuum of diversity in perfusion,
histopathological changes and bowel wall stiffness, that cannot be simplified as “active vs. chronic” “stricturing
vs. inflammatory” “symptomatic vs. non-symptomatic”.
Hopefully, future cognitive computer interpretation allowing for big data analyses may help us better diagnose
and understand the complexity of this chronic and invalidating disease with the final goal of improved
classification, prognoses and treatment.
55
References
1. Baumgart, D. C. & Sandborn, W. J. Crohn’s disease. Lancet (London, England) 380, 1590–605 (2012).
2. Chassaing, B. et al. Dietary emulsifiers impact the mouse gut microbiota promoting colitis and metabolic syndrome. Nature 519, 92–6 (2015).
3. Cleynen, I. et al. Inherited determinants of Crohn’s disease and ulcerative colitis phenotypes: a genetic association study. Lancet (London, England) 387, 156–67 (2016).
4. Solberg, I. C. et al. Clinical Course in Crohn’s Disease: Results of a Norwegian Population-Based Ten-Year Follow-Up Study. Clin. Gastroenterol. Hepatol. 5, 1430–1438 (2007).
5. Silverberg, M. S. et al. Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: report of a Working Party of the 2005 Montreal World Congress of Gastroenterology. Can. J. Gastroenterol. 19 Suppl A, 5A–36A (2005).
6. Satsangi, J., Silverberg, M. S., Vermeire, S. & Colombel, J.-F. The Montreal classification of inflammatory bowel disease: controversies, consensus, and implications. Gut 55, 749–53 (2006).
7. Van Assche, G. et al. The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Definitions and diagnosis. J. Crohns. Colitis 4, 7–27 (2010).
8. Tanaka, M., Kawakami, A., Iwao, Y., Fukushima, T. & Yamamoto-Mitani, N. Coping Strategies for Possible Flare-Ups and Their Perceived Effectiveness in Patients With Inflammatory Bowel Disease. Gastroenterol. Nurs. 39, 42–7
9. Høivik, M. L. et al. Work disability in inflammatory bowel disease patients 10 years after disease onset: results from the IBSEN Study. Gut 62, 368–75 (2013).
10. Peyrin-Biroulet, L. et al. Development of the first disability index for inflammatory bowel disease based on the international classification of functioning, disability and health. Gut 61, 241–7 (2012).
11. Bernstein, C. N., Blanchard, J. F., Rawsthorne, P. & Yu, N. The prevalence of extraintestinal diseases in inflammatory bowel disease: a population-based study. Am. J. Gastroenterol. 96, 1116–22 (2001).
12. Pariente, B. et al. Development of the Crohn’s disease digestive damage score, the Lémann score. Inflamm. Bowel Dis. 17, 1415–22 (2011).
13. Bernell, O., Lapidus, a & Hellers, G. Risk factors for surgery and postoperative recurrence in Crohn’s disease. Ann. Surg. 231, 38–45 (2000).
14. Khanna, R. et al. Early combined immunosuppression for the management of Crohn’s disease (REACT): a cluster randomised controlled trial. Lancet (London, England) 386, 1825–34 (2015).
15. Sandborn, W. J. et al. Treating beyond symptoms with a view to improving patient outcomes in inflammatory bowel diseases. J. Crohns. Colitis 8, 927–935 (2014).
16. Smolen, J. S. et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2013 update. Ann. Rheum. Dis. 73, 492–509 (2014).
17. Kennelly, R. P., Subramaniam, T., Egan, L. J. & Joyce, M. R. Smoking and Crohn’s disease: Active
56
modification of an independent risk factor (Education alone is not enough). J. Crohns. Colitis 7, 631–5 (2013).
18. Cosnes, J., Beaugerie, L., Carbonnel, F. & Gendre, J. P. Smoking cessation and the course of Crohn’s disease: an intervention study. Gastroenterology 120, 1093–9 (2001).
19. Peyrin-Biroulet, L. et al. Clinical disease activity, C-reactive protein normalisation and mucosal healing in Crohn’s disease in the SONIC trial. Gut 63, 88–95 (2014).
20. Castiglione, F. et al. Transmural healing evaluated by bowel sonography in patients with Crohn’s disease on maintenance treatment with biologics. Inflamm. Bowel Dis. 19, 1928–34 (2013).
21. Peyrin-Biroulet, L. et al. Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE): Determining Therapeutic Goals for Treat-to-Target. Am. J. Gastroenterol. 110, 1324–38 (2015).
22. Bouguen, G. et al. Treat to target: a proposed new paradigm for the management of Crohn’s disease. Clin. Gastroenterol. Hepatol. 13, 1042–50.e2 (2015).
23. Peyrin-Biroulet, L. et al. Defining Disease Severity in Inflammatory Bowel Diseases: Current and Future Directions. Clin. Gastroenterol. Hepatol. 14, 348–354.e17 (2016).
24. Pariente, B. et al. Development of the Lémann Index to Assess Digestive Tract Damage in Patients With Crohn’s Disease. Gastroenterology 148, 52–63.e3 (2015).
25. Zorzi, F. et al. A Sonographic Lesion Index for Crohn’s Disease Helps Monitor Changes in Transmural Bowel Damage During Therapy. Clin. Gastroenterol. Hepatol. 12, 2071–2077 (2014).
26. Rimola, J. et al. Magnetic resonance imaging for evaluation of Crohn’s disease: validation of parameters of severity and quantitative index of activity. Inflamm. Bowel Dis. 17, 1759–68 (2011).
27. Mary, J. Y. & Modigliani, R. Development and validation of an endoscopic index of the severity for Crohn’s disease: a prospective multicentre study. Groupe d'Etudes Therapeutiques des Affections Inflammatoires du Tube Digestif (GETAID). Gut 30, 983–989 (1989).
28. Daperno, M. et al. Development and validation of a new, simplified endoscopic activity score for Crohn’s disease: the SES-CD. Gastrointest. Endosc. 60, 505–512 (2004).
29. Zappa, M. et al. Which magnetic resonance imaging findings accurately evaluate inflammation in small bowel Crohn’s disease? A retrospective comparison with surgical pathologic analysis. Inflamm. Bowel Dis. 17, 984–93 (2011).
30. Latella, G., Sferra, R., Speca, S., Vetuschi, A. & Gaudio, E. Can we prevent, reduce or reverse intestinal fibrosis in IBD? Eur. Rev. Med. Pharmacol. Sci. 17, 1283–1304 (2013).
31. Best, W. R., Becktel, J. M., Singleton, J. W. & Kern Jr, F. Development of a Crohn’s disease activity index. National Cooperative Crohn's Disease Study. Gastroenterology 70, 439 (1976).
32. Harvey, R. F. & Bradshaw, J. M. A simple index of Crohn’s-disease activity. Lancet (London, England) 1, 514 (1980).
33. Hindryckx, P., Baert, F., Hart, A., Armuzzi, A. & Ecco, O. Clinical trials in luminal Crohn ’ s disease : A historical perspective. J. Crohn’s Colitis 8, 1339–1350 (2014).
57
34. Vilela, E. G. et al. Evaluation of inflammatory activity in Crohn’s disease and ulcerative colitis. World J. Gastroenterol. 18, 872–81 (2012).
35. Røseth, A. G., Fagerhol, M. K., Aadland, E. & Schjønsby, H. Assessment of the neutrophil dominating protein calprotectin in feces. A methodologic study. Scand. J. Gastroenterol. 27, 793–8 (1992).
36. Jensen, M. D., Kjeldsen, J. & Nathan, T. Fecal calprotectin is equally sensitive in Crohn’s disease affecting the small bowel and colon. Scand. J. Gastroenterol. 46, 694–700 (2011).
37. Schoepfer, A. M. et al. Fecal calprotectin correlates more closely with the Simple Endoscopic Score for Crohn’s disease (SES-CD) than CRP, blood leukocytes, and the CDAI. Am. J. Gastroenterol. 105, 162–9 (2010).
38. Wright, E. K. et al. Measurement of Fecal Calprotectin Improves Monitoring and Detection of Recurrence of Crohn’s Disease Following Surgery. Gastroenterology 148, 938–947.e1 (2015).
39. D’Haens, G. et al. Fecal calprotectin is a surrogate marker for endoscopic lesions in inflammatory bowel disease. Inflamm. Bowel Dis. 18, 2218–24 (2012).
40. Makanyanga, J. C. et al. Evaluation of Crohn’s disease activity: Initial validation of a magnetic resonance enterography global score (MEGS) against faecal calprotectin. Eur. Radiol. 24, 277–287 (2014).
41. Mary, J. Y. & Modigliani, R. Development and validation of an endoscopic index of the severity for Crohn’s disease: a prospective multicentre study. Gut 30, 983–989 (1989).
42. Annese, V. et al. European evidence based consensus for endoscopy in inflammatory bowel disease. J. Crohns. Colitis 7, 982–1018 (2013).
43. Rutgeerts, P. et al. Natural history of recurrent Crohn’s disease at the ileocolonic anastomosis after curative surgery. Gut 25, 665–672 (1984).
44. Walsh, A., Palmer, R. & Travis, S. Mucosal healing as a target of therapy for colonic inflammatory bowel disease and methods to score disease activity. Gastrointest. Endosc. Clin. N. Am. 24, 367–378 (2014).
45. Tontini, G. E., Bisschops, R. & Neumann, H. Endoscopic scoring systems for inflammatory bowel disease: pros and cons. Expert Rev. Gastroenterol. Hepatol. 8, 543–54 (2014).
46. Vuitton, L. et al. IOIBD technical review on endoscopic indices for Crohn’s disease clinical trials. Gut gutjnl–2015–309903 (2015). doi:10.1136/gutjnl-2015-309903
47. Feagan, B. G. et al. The role of centralized reading of endoscopy in a randomized controlled trial of mesalamine for ulcerative colitis. Gastroenterology 145, 149–157.e2 (2013).
48. Daperno, M. et al. Inter-observer agreement in endoscopic scoring systems: Preliminary report of an ongoing study from the Italian Group for Inflammatory Bowel Disease (IG-IBD). Dig. Liver Dis. 46, 969–973 (2014).
49. Khanna, R. et al. Reliability among central readers in the evaluation of endoscopic findings from patients with Crohn’s disease. Gut 1–7 (2015). doi:10.1136/gutjnl-2014-308973
50. Parente, F., Maconi, G. & Bianchi Porro, G. Bowel ultrasound in Crohn disease: current role and future applications. Scand. J. Gastroenterol. 37, 871–6 (2002).
58
51. Jaffe, T. a et al. Radiation doses from small-bowel follow-through and abdominopelvic MDCT in Crohn’s disease. AJR. Am. J. Roentgenol. 189, 1015–22 (2007).
52. Horsthuis, K., Bipat, S., Bennink, R. J. & Stoker, J. Inflammatory Bowel Disease Diagnosed with US, MR, Scintigraphy, and CT: Meta-analysis of Prospective Studies. Radiology 247, 64–79 (2008).
53. Sauer, C. G., Kugathasan, S., Martin, D. R. & Applegate, K. E. Medical radiation exposure in children with inflammatory bowel disease estimates high cumulative doses. Inflamm. Bowel Dis. 17, 2326–2332 (2011).
54. Zakeri, N. & Pollok, R. C. Diagnostic imaging and radiation exposure in inflammatory bowel disease. World J. Gastroenterol. 22, 2165–78 (2016).
55. Panés, J. et al. Systematic review: the use of ultrasonography, computed tomography and magnetic resonance imaging for the diagnosis, assessment of activity and abdominal complications of Crohn’s disease. Aliment. Pharmacol. Ther. 34, 125–45 (2011).
56. Tielbeek, J. a W. et al. Grading Crohn Disease Activity With MRI: Interobserver Variability of MRI Features, MRI Scoring of Severity, and Correlation With Crohn Disease Endoscopic Index of Severity. Am. J. Roentgenol. 201, 1220–1228 (2013).
57. Bruining, D. H. et al. CT and MR enterography in Crohn’s disease: current and future applications. Abdom. Imaging 40, 965–74 (2015).
58. Rimola, J. et al. Magnetic resonance imaging for evaluation of Crohn’s disease: validation of parameters of severity and quantitative index of activity. Inflamm. Bowel Dis. 17, 1759–68 (2011).
59. Ordás, I. et al. Accuracy of magnetic resonance enterography in assessing response to therapy and mucosal healing in patients with Crohn’s disease. Gastroenterology 146, 374–82.e1 (2014).
60. Steward, M. J. et al. Non-perforating small bowel Crohn’s disease assessed by MRI enterography: derivation and histopathological validation of an MR-based activity index. Eur. J. Radiol. 81, 2080–8 (2012).
61. Makanyanga, J. C. et al. Evaluation of Crohn’s disease activity: initial validation of a magnetic resonance enterography global score (MEGS) against faecal calprotectin. Eur. Radiol. 24, 277–87 (2014).
62. Prezzi, D. et al. Monitoring Crohn’s disease during anti-TNF-α therapy: validation of the magnetic resonance enterography global score (MEGS) against a combined clinical reference standard. Eur. Radiol. (2015). doi:10.1007/s00330-015-4036-1
63. Oussalah, A. et al. Diffusion-weighted magnetic resonance without bowel preparation for detecting colonic inflammation in inflammatory bowel disease. Gut 59, 1056–1065 (2010).
64. Semelka, R. C. & Helmberger, T. K. Contrast agents for MR imaging of the liver. Radiology 218, 27–38 (2001).
65. Balci, N. C. & Semelka, R. C. Contrast agents for MR imaging of the liver. Radiol. Clin. North Am. 43, 887–98, viii (2005).
66. Gandhi, S. N., Brown, M. A., Wong, J. G., Aguirre, D. A. & Sirlin, C. B. MR contrast agents for liver imaging: what, when, how. Radiographics 26, 1621–36 (2006).
67. Calamante, F. Arterial input function in perfusion MRI: a comprehensive review. Prog. Nucl. Magn.
59
Reson. Spectrosc. 74, 1–32 (2013).
68. Makanyanga, J., Punwani, S. & Taylor, S. a. Assessment of wall inflammation and fibrosis in Crohn’s disease: value of T1-weighted gadolinium-enhanced MR imaging. Abdom. Imaging 37, 933–43 (2012).
69. Stoker, J. MRI of the Gastrointestinal Tract. European radiology (Springer Berlin Heidelberg, 2010). doi:10.1007/978-3-540-85532-3
70. Rimola, J. et al. Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn’s disease. Gut 58, 1113–1120 (2009).
71. Taylor, S. A. et al. Mural Crohn Disease: Correlation of Dynamic Contrast-enhanced MR Imaging Findings with Angiogenesis and Inflammation at Histologic Examination—Pilot Study. Radiology 251, 369–379 (2009).
72. Moreno, N. et al. Usefulness of abdominal ultrasonography in the analysis of endoscopic activity in patients with Crohn’s disease: changes following treatment with immunomodulators and/or anti-TNF antibodies. J. Crohns. Colitis 8, 1079–87 (2014).
73. Panes, J. et al. Imaging techniques for assessment of inflammatory bowel disease: Joint ECCO and ESGAR evidence-based consensus guidelines. J. Crohns. Colitis 7, 556–85 (2013).
74. Calabrese, E. et al. Development of a numerical index quantitating small bowel damage as detected by ultrasonography in Crohn’s disease. J. Crohns. Colitis 6, 852–60 (2012).
75. Spalinger, J. et al. Doppler US in patients with crohn disease: vessel density in the diseased bowel reflects disease activity. Radiology 217, 787–91 (2000).
76. Limberg, B. [Diagnosis of chronic inflammatory bowel disease by ultrasonography]. Zeitschrift fur Gastroenterol. 37, 495–508 (1999).
77. Schlottmann, K., Kratzer, W. & Schölmerich, J. Doppler ultrasound and intravenous contrast agents in gastrointestinal tract disorders: current role and future implications. Eur. J. Gastroenterol. Hepatol. 17, 263–75 (2005).
78. Medellin-Kowalewski, A., Wilkens, R., Wilson, A., Ruan, J. & Wilson, S. R. Quantitative Contrast-Enhanced Ultrasound Parameters in Crohn Disease: Their Role in Disease Activity Determination With Ultrasound. AJR. Am. J. Roentgenol. 206, 64–73 (2016).
79. Greis, C. Quantitative evaluation of microvascular blood flow by contrast-enhanced ultrasound (CEUS). Clin. Hemorheol. Microcirc. 49, 137–49 (2011).
80. Quaia, E. et al. The value of time-intensity curves obtained after microbubble contrast agent injection to discriminate responders from non-responders to anti-inflammatory medication among patients with Crohn’s disease. Eur. Radiol. 23, 1650–9 (2013).
81. Nylund, K. et al. Quantitative Contrast-Enhanced Ultrasound Comparison Between Inflammatory and Fibrotic Lesions in Patients With Crohn’s Disease. Ultrasound Med. Biol. 39, 1197–1206 (2013).
82. Saevik, F., Nylund, K., Hausken, T., Odegaard, S. & Gilja, O. H. Bowel Perfusion Measured with Dynamic Contrast-enhanced Ultrasound Predicts Treatment Outcome in Patients with Crohn’s Disease. Inflamm. Bowel Dis. 20, 2029–37 (2014).
60
83. Girlich, C., Schacherer, D., Jung, E. M., Schreyer, A. & Büttner, R. Comparison between a clinical activity index (Harvey-Bradshaw-Index), laboratory inflammation markers and quantitative assessment of bowel wall vascularization by contrast-enhanced ultrasound in Crohn’s disease. Eur. J. Radiol. 81, 1105–9 (2012).
84. Mow, W. S. et al. Initial experience with wireless capsule enteroscopy in the diagnosis and management of inflammatory bowel disease. Diagnostic yield of wireless capsule enteroscopy in comparison with computed tomography enteroclysis. Clin Gastroenterol Hepatol 2, 31–40. (2004).
85. Gal, E., Geller, A., Fraser, G., Levi, Z. & Niv, Y. Assessment and validation of the new capsule endoscopy Crohn’s disease activity index (CECDAI). Dig. Dis. Sci. 53, 1933–1937 (2008).
86. Gralnek, I. M. et al. Development of a capsule endoscopy scoring index for small bowel mucosal inflammatory change. Aliment. Pharmacol. Ther. 27, 146–54 (2008).
87. Park, S., Keum, B., Hyun, J. & Seo, Y. A novel cleansing score system for capsule endoscopy. World J. Gastroenterol. 16, 875–880 (2010).
88. Oort, J. & Scheper, R. J. Histopathology of acute and chronic inflammation. Agents Actions. Suppl. 25–30 (1977).
89. Odze, R., Antonioli, D., Peppercorn, M. & Goldman, H. Effect of topical 5-aminosalicylic acid (5-ASA) therapy on rectal mucosal biopsy morphology in chronic ulcerative colitis. The American journal of surgical pathology 17, 869–875 (1993).
90. Magro, F. et al. European consensus on the histopathology of inflammatory bowel disease. J. Crohn’s Colitis 7, 827–851 (2013).
91. Latella, G., Di Gregorio, J., Flati, V., Rieder, F. & Lawrance, I. C. Mechanisms of initiation and progression of intestinal fibrosis in IBD. Scand. J. Gastroenterol. 50, 53–65 (2015).
92. Wynn, T. A. & Ramalingam, T. R. Mechanisms of fibrosis: therapeutic translation for fibrotic disease. Nat. Med. 18, 1028–1040 (2012).
93. Tanaka, M. et al. Morphologic criteria applicable to biopsy specimens for effective distinction of inflammatory bowel disease from other forms of colitis and of Crohn’s disease from ulcerative colitis. Scand. J. Gastroenterol. 34, 55–67 (1999).
94. Iacobuzio-Donahue, C. A. & Montgomery, E. A. Gastrointestinal and Liver Pathology: A Volume in the Series: Foundations in Diagnostic Pathology. (Elsevier Health Sciences, 2011).
95. Gramlich, T. & Petras, R. E. Pathology of inflammatory bowel disease. Semin. Pediatr. Surg. 16, 154–163 (2007).
96. Graham, M. F. et al. Collagen content and types in the intestinal strictures of Crohn’s disease. Gastroenterology 94, 257–65 (1988).
97. Stallmach, A., Schuppan, D., Riese, H. H., Matthes, H. & Riecken, E. O. Increased collagen type III synthesis by fibroblasts isolated from strictures of patients with Crohn’s disease. Gastroenterology 102, 1920–9 (1992).
98. Graham, M. F., Drucker, D. E., Diegelmann, R. F. & Elson, C. O. Collagen synthesis by human intestinal
61
smooth muscle cells in culture. Gastroenterology 92, 400–5 (1987).
99. Wilkens, R. & Strandberg, C. Transabdominal ultralydskanning af tarmen er et godt supplement ved diagnostik af tarmsygdomme. Ugeskr. Læger 176, 451–454 (2014).
100. Nylund, K. Sonography of the small intestine. World J. Gastroenterol. 15, 1319 (2009).
101. Nylund, K. et al. Crohn’s disease: Comparison of in vitro ultrasonographic images and histology. Scand. J. Gastroenterol. 43, 719–26 (2008).
102. Allgayer, H., Braden, B. & Dietrich, C. Transabdominal ultrasound in inflammatory bowel disease. Conven-tional and recently developed techniques–Update. Med. Ultrason. 13, 302–313 (2011).
103. Nylund, K., Hausken, T., Odegaard, S., Eide, G. E. & Gilja, O. H. Gastrointestinal Wall Thickness Measured with Transabdominal Ultrasonography and Its Relationship to Demographic Factors in Healthy Subjects. Ultraschall Med. 33, E225–E232 (2012).
104. Greis, C. Technology overview: sonoVue. Eur. Radiol. Suppl. 14, P11–P15 (2004).
105. Burns, P. N. & Wilson, S. R. Microbubble contrast for radiological imaging: 1. Principles. Ultrasound Q. 22, 5 (2006).
106. Gauthier, T. P., Chebil, M., Peronneau, P. & Lassau, N. In vitro evaluation of the impact of ultrasound scanner settings and contrast bolus volume on time-intensity curves. Ultrasonics 52, 12–9 (2012).
107. Payen, T. et al. Echo-power estimation from log-compressed video data in dynamic contrast-enhanced ultrasound imaging. Ultrasound Med. Biol. 39, 1826–37 (2013).
108. Lampaskis, M. & Averkiou, M. Investigation of the relationship of nonlinear backscattered ultrasound intensity with microbubble concentration at low MI. Ultrasound Med. Biol. 36, 306–12 (2010).
109. Piscaglia, F. et al. The EFSUMB Guidelines and Recommendations on the Clinical Practice of Contrast Enhanced Ultrasound (CEUS): Update 2011 on non-hepatic applications. Ultraschall Med. 33–59 (2011). doi:10.1055/s-0031-1281676
110. Gauthier, T. P., Averkiou, M. a & Leen, E. L. S. Perfusion quantification using dynamic contrast-enhanced ultrasound: the impact of dynamic range and gain on time-intensity curves. Ultrasonics 51, 102–6 (2011).
111. Arditi, M., Frinking, P. J. a, Zhou, X. & Rognin, N. G. A new formalism for the quantification of tissue perfusion by the destruction-replenishment method in contrast ultrasound imaging. IEEE transactions on ultrasonics, ferroelectrics, and frequency control 53, 1118–29 (2006).
112. Tranquart, F., Mercier, L., Frinking, P., Gaud, E. & Arditi, M. Perfusion quantification in contrast-enhanced ultrasound (CEUS)--ready for research projects and routine clinical use. Ultraschall Med. 33 Suppl 1, S31–38 (2012).
113. Rognin, N. G., Frinking, P., Costa, M. & Arditi, M. In-vivo perfusion quantification by contrast ultrasound: Validation of the use of linearized video data vs. raw RF data. in 2008 IEEE Ultrasonics Symposium 1690–1693 (IEEE, 2008). doi:10.1109/ULTSYM.2008.0413
114. Salvatore, V. et al. Quantification of enhancement of focal liver lesions during contrast-enhanced ultrasound (CEUS). Analysis of ten selected frames is more simple but as reliable as the analysis of the entire loop for most parameters. Eur. J. Radiol. 81, 709–13 (2012).
62
115. Girlich, C. et al. Comparison between Preoperative Quantitative Assessment of Bowel Wall Vascularization by Contrast-Enhanced Ultrasound and Operative Macroscopic Findings and Results of Histopathological Scoring in Crohn’s Disease. Ultraschall Med 6, 154–159 (2010).
116. Migaleddu, V. et al. Contrast-enhanced ultrasonographic evaluation of inflammatory activity in Crohn’s disease. Gastroenterology 137, 43–52 (2009).
117. Quaia, E. et al. The value of small bowel wall contrast enhancement after sulfur hexafluoride-filled microbubble injection to differentiate inflammatory from fibrotic strictures in patients with Crohn’s disease. Ultrasound Med. Biol. 38, 1324–32 (2012).
118. Gauthier, M. et al. Assessment of quantitative perfusion parameters by dynamic contrast-enhanced sonography using a deconvolution method: an in vitro and in vivo study. J. Ultrasound Med. 31, 595–608 (2012).
119. Gauthier, M. Impact of the arterial input function on microvascularization parameter measurements using dynamic contrast-enhanced ultrasonography. World J. Radiol. 4, 291 (2012).
120. Østergaard, L. Principles of cerebral perfusion imaging by bolus tracking. J. Magn. Reson. Imaging 22, 710–7 (2005).
121. Dietrich, C. F. et al. An EFSUMB Introduction into Dynamic Contrast-Enhanced Ultrasound (DCE-US) for Quantification of Tumour Perfusion. Ultraschall Med. 33, 344–51 (2012).
122. Ignee, A., Jedrejczyk, M., Schuessler, G., Jakubowski, W. & Dietrich, C. F. Quantitative contrast enhanced ultrasound of the liver for time intensity curves-Reliability and potential sources of errors. Eur. J. Radiol. 73, 153–8 (2010).
123. Hudson, J. M., Karshafian, R. & Burns, P. N. Quantification of flow using ultrasound and microbubbles: a disruption replenishment model based on physical principles. Ultrasound Med. Biol. 35, 2007–20 (2009).
124. Angelelli, P., Nylund, K., Gilja, O. H. & Hauser, H. Interactive visual analysis of contrast-enhanced ultrasound data based on small neighborhood statistics. Comput. Graph. 35, 218–226 (2011).
125. Suisse, B. Vuebox. Quantification Toolbox. Instructions for use. (2011).
126. Wilkens, R., Pournazari, P. & Wilson, S. R. in Contrast-enhanced ultrasound (ed. Weskott, H.-P.) 222–229 (Uni-Med, 2013).
127. Lee Pearce, M., Lewis, a. E. & Kaplan, M. R. The Factors Influencing the Circulation Time. Circulation 5, 583–588 (1952).
128. De Franco, A. et al. Ileal Crohn Disease: Mural Microvascularity Quantified with Contrast-enhanced US Correlates with Disesase Activity. Radiology 262, 680–688 (2012).
129. Quaia, E. et al. The diagnostic value of small bowel wall vascularity after sulfur hexafluoride-filled microbubble injection in patients with Crohn’s disease. Correlation with the therapeutic effectiveness of specific anti-inflammatory treatment. Eur. J. Radiol. 69, 438–444 (2009).
130. Negaard, A. et al. MRI of the small bowel with oral contrast or nasojejunal intubation in Crohn’s disease: randomized comparison of patient acceptance. Scand. J. Gastroenterol. 43, 44–51 (2008).
131. Negaard, A. et al. A prospective randomized comparison between two MRI studies of the small bowel in
63
Crohn’s disease, the oral contrast method and MR enteroclysis. Eur. Radiol. 17, 2294–2301 (2007).
132. Tofts, P. S. et al. Estimating kinetic parameters from dynamic contrast-enhanced t1-weighted MRI of a diffusable tracer: Standardized quantities and symbols. J. Magn. Reson. Imaging 10, 223–232 (1999).
133. Wang, P. et al. Effects of flip angle uncertainty and noise on the accuracy of DCE-MRI metrics: comparison between standard concentration-based and signal difference methods. Magn. Reson. Imaging 33, 166–73 (2015).
134. Sourbron, S. P. & Buckley, D. L. On the scope and interpretation of the Tofts models for DCE-MRI. Magn. Reson. Med. 66, 735–745 (2011).
135. Oto, A. et al. Quantitative analysis of dynamic contrast enhanced MRI for assessment of bowel inflammation in Crohn’s disease pilot study. Acad. Radiol. 16, 1223–30 (2009).
136. Goh, V. & Taylor, J. N. in MRI of the Gastrointestinal Tract (ed. Stoker, J.) 51–63 (Springer, 2010).
137. Horsthuis, K., Lavini, C., Bipat, S., Stokkers, P. & Stoker, J. Perianal Crohn Disease: Evaluation of Dynamic Contrast-enhanced MR Imaging as an Indicator of Disease Activity. Radiology 251, 380–387 (2009).
138. Del Pinto, R., Pietropaoli, D., Chandar, A. K., Ferri, C. & Cominelli, F. Association Between Inflammatory Bowel Disease and Vitamin D Deficiency: A Systematic Review and Meta-analysis. Inflamm. Bowel Dis. 21, 2708–17 (2015).
139. Vermeire, S., Assche, G. Van & Rutgeerts, P. Laboratory markers in IBD: useful, magic, or unnecessary toys? Gut 55, 426–431 (2006).
140. Dignass, a et al. The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Current management. J. Crohns. Colitis 4, 28–62 (2010).
141. Schoepfer, A. M. et al. Fecal calprotectin correlates more closely with the Simple Endoscopic Score for Crohn’s disease (SES-CD) than CRP, blood leukocytes, and the CDAI. Am. J. Gastroenterol. 105, 162–9 (2010).
142. Cerrillo, E. et al. Fecal Calprotectin in Ileal Crohn’s Disease: Relationship with Magnetic Resonance Enterography and a Pathology Score. Inflamm. Bowel Dis. 21, 1572–9 (2015).
143. Borley, N. R., Mortensen, N. J., Jewell, D. P. & Warren, B. F. The relationship between inflammatory and serosal connective tissue changes in ileal Crohn’s disease: evidence for a possible causative link. J. Pathol. 190, 196–202 (2000).
144. Mosli, M. H. et al. Development and validation of a histological index for UC. Gut Oct 16, 1–9 (2015).
145. Marchal-Bressenot, A. et al. Development and validation of the Nancy histological index for UC. Gut Oct 16, 1–9 (2015).
146. Iacucci, M. et al. Complete Mucosal Healing Defined by Endoscopic Mayo Subscore Still Demonstrates Abnormalities by Novel iSCAN Endoscopic and Refined Histological Gradings. Gastroenterology 146, S–152 (2014).
147. Chiorean, M. V et al. Correlation of CT enteroclysis with surgical pathology in Crohn’s disease. Am. J. Gastroenterol. 102, 2541–50 (2007).
64
148. Baumgart, D. C. et al. US-based Real-time Elastography for the Detection of Fibrotic Gut Tissue in Patients with Stricturing Crohn Disease. Radiology 275, 889–99 (2015).
149. Pucilowska, J. B. et al. IGF-I and procollagen alpha1(I) are coexpressed in a subset of mesenchymal cells in active Crohn’s disease. Am. J. Physiol. Gastrointest. Liver Physiol. 279, G1307–22 (2000).
150. Taylor, S. A. et al. Mural Crohn Disease : Correlation of Dynamic Contrast-enhanced MR Imaging Findings with Angiogenesis and Inflammation at Histologic Examination — Pilot Study 1 Purpose : Methods : Results : Conclusion : Imaging 251, 369–379 (2009).
151. Chen, M. & Shen, B. Endoscopic Therapy in Crohnʼs Disease. Inflamm. Bowel Dis. 21, 2222–2240 (2015).
152. Johnson, L. a et al. Matrix stiffness corresponding to strictured bowel induces a fibrogenic response in human colonic fibroblasts. Inflamm. Bowel Dis. 19, 891–903 (2013).
153. Tsamis, A., Pocivavsek, L. & Vorp, D. A. Elasticity and geometry: a computational model of the Heineke-Mikulicz strictureplasty. Biomech. Model. Mechanobiol. 13, 1185–98 (2014).
154. Russell, M. K. Functional assessment of nutrition status. Nutr. Clin. Pract. 30, 211–8 (2015).
155. Rao, S. S., Hayek, B. & Summers, R. W. Impedance planimetry: an integrated approach for assessing sensory, active, and passive biomechanical properties of the human esophagus. Am. J. Gastroenterol. 90, 431–8 (1995).
156. Fynne, L. et al. Distensibility of the anal canal in patients with systemic sclerosis: a study with the functional lumen imaging probe. Colorectal Dis. 15, e40–7 (2013).
157. Luft, F. et al. Functional luminal imaging probe: a new technique for dynamic evaluation of mechanical properties of the anal canal. Tech. Coloproctol. 16, 451–7 (2012).
158. Wilkens, R. et al. Dynamic Contrast Enhanced Magnetic Resonance Enterography and Dynamic Contrast Enhanced Ultrasonography in Crohn Disease: An observational comparison study. Ultrasound Int. Open (under Rev. (2016).
159. Murphy, D. J. et al. Subclassification of small bowel Crohn’s disease using magnetic resonance enterography: A review using evidence-based medicine methodology. Clin. Radiol. 70, 1336–1343 (2015).
160. Wilkens, R. et al. Validity of Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced MR Enterography in the assessment of Crohn’s Disease. Under Constr.
161. Bailey, L. E. & Ong, S. D. Krebs-Henseleit solution as a physiological buffer in perfused and superfused preparations. J. Pharmacol. Methods 1, 171–175 (1978).
162. Harris, P. A. et al. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J. Biomed. Inform. 42, 377–81 (2009).
163. Bland, J. & Altman, D. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet 346, 1085–1087 (1995).
164. Bland, J. M. & Altman, D. G. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1, 307–10 (1986).
65
165. Raunig, D. L. et al. Quantitative imaging biomarkers: a review of statistical methods for technical performance assessment. Stat. Methods Med. Res. 24, 27–67 (2015).
166. Sim, J. & Wright, C. C. The kappa statistic in reliability studies: use, interpretation, and sample size requirements. Phys. Ther. 85, 257–68 (2005).
167. Kirkwood, B. R. & Sterne, J. A. C. Essential Medical Statistics. (Blackwell Publishing, 2003).
168. Ripollés, T. et al. Contrast-enhanced ultrasound (CEUS) in Crohn’s disease: technique, image interpretation and clinical applications. Insights Imaging 2, 639–652 (2011).
169. Ripollés, T. et al. Crohn Disease: Correlation of Findings at Contrast-enhanced US with Severity at Endoscopy. Radiology 253, 241–248 (2009).
170. Girlich, C., Schacherer, D., Jung, E. M., Klebl, F. & Huber, E. Comparison between quantitative assessment of bowel wall vascularization by contrast-enhanced ultrasound and results of histopathological scoring in ulcerative colitis. Int. J. Colorectal Dis. 27, 193–8 (2012).
171. Jirik, R. et al. Ultrasound perfusion analysis combining bolus-tracking and burst-replenishment. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 60, 310–9 (2013).
172. Socaciu, M. et al. Non-Invasive Assessment of Inflammation and Treatment Response in Patients with Crohn’s Disease and Ulcerative Colitis using Contrast-Enhanced Ultrasonography Quantification. J. Gastrointestin. Liver Dis. 24, 457–465 (2015).
173. Zink, F. et al. Comparison of Two High-End Ultrasound Systems for Contrast-Enhanced Ultrasound Quantification of Mural Microvascularity in Crohn’s Disease. Ultraschall Med. 37, 74–81 (2016).
174. Tang, M.-X. et al. Quantitative contrast-enhanced ultrasound imaging: a review of sources of variability. Interface Focus 1, 520–39 (2011).
175. Rimola, J. et al. Characterization of Inflammation and Fibrosis in Crohn’s Disease Lesions by Magnetic Resonance Imaging. Am. J. Gastroenterol. 110, 432–441 (2015).
176. Sharman, A., Zealley, I. A., Greenhalgh, R., Bassett, P. & Taylor, S. A. MRI of small bowel Crohn’s disease: determining the reproducibility of bowel wall gadolinium enhancement measurements. Eur. Radiol. 19, 1960–7 (2009).
177. Mendoza, J. L. & Abreu, M. T. Biological markers in inflammatory bowel disease: practical consideration for clinicians. Gastroentérologie Clin. Biol. 33 Suppl 3, S158–73 (2009).
178. Louis, E. et al. A positive response to infliximab in Crohn disease: association with a higher systemic inflammation before treatment but not with -308 TNF gene polymorphism. Scand. J. Gastroenterol. 37, 818–24 (2002).
179. Langner, C. et al. The histopathological approach to inflammatory bowel disease: A practice guide. Virchows Arch. 464, 511–527 (2014).
180. Schaeffer, D. F. et al. Distinctive histopathologic phenotype in resection specimens from patients with Crohn’s disease receiving anti-TNF-α therapy. Hum. Pathol. 45, 1928–1935 (2014).
181. Bataille, F. et al. Histopathological parameters as predictors for the course of Crohn’s disease. Virchows Arch. 443, 501–7 (2003).
66
182. Sasaki, T. et al. Doppler ultrasound findings correlate with tissue vascularity and inflammation in surgical pathology specimens from patients with small intestinal Crohn’s disease. BMC Res. Notes 7, 363 (2014).
183. Adler, J. et al. Computed tomography enterography findings correlate with tissue inflammation, not fibrosis in resected small bowel Crohn’s disease. Inflamm. Bowel Dis. 18, 849–56 (2012).
184. Romanini, L. et al. Quantitative analysis of contrast-enhanced ultrasonography of the bowel wall can predict disease activity in inflammatory bowel disease. Eur. J. Radiol. 83, 1317–23 (2014).
185. Fraquelli, M. et al. The Role of Ultrasound Elasticity Imaging in Predicting Ileal Fibrosis in Crohn’s Disease Patients. Inflamm. Bowel Dis. 21, 2605–12 (2015).
186. Drews, B. H. et al. Comparison of sonographically measured bowel wall vascularity, histology, and disease activity in Crohn’s disease. Eur. Radiol. 19, 1379–86 (2009).
187. Makanyanga, J. et al. MRI texture analysis (MRTA) of T2-weighted images in Crohn’s disease may provide information on histological and MRI disease activity in patients undergoing ileal resection. Eur. Radiol. (2016). doi:10.1007/s00330-016-4324-4
188. Knod, J. L. et al. Angiogenesis and Vascular Endothelial Growth Factor-A Expression Associated with Inflammation in Pediatric Crohn’s Disease. J. Gastrointest. Surg. 20, 624–30 (2016).
189. Danese, S. et al. Angiogenesis as a novel component of inflammatory bowel disease pathogenesis. Gastroenterology 130, 2060–73 (2006).
190. Cromer, W. E., Mathis, J. M., Granger, D. N., Chaitanya, G. V & Alexander, J. S. Role of the endothelium in inflammatory bowel diseases. World J. Gastroenterol. 17, 578–93 (2011).
191. Hatoum, O. a, Binion, D. G. & Gutterman, D. D. Paradox of simultaneous intestinal ischaemia and hyperaemia in inflammatory bowel disease. Eur. J. Clin. Invest. 35, 599–609 (2005).
192. Costa, C., Incio, J. & Soares, R. Angiogenesis and chronic inflammation: Cause or consequence? Angiogenesis 10, 149–166 (2007).
193. Stidham, R. W. et al. Ultrasound elasticity imaging for detecting intestinal fibrosis and inflammation in rats and humans with Crohn’s disease. Gastroenterology 141, 819–826.e1 (2011).
194. Kim, K. et al. Noninvasive ultrasound elasticity imaging (UEI) of Crohn’s disease: animal model. Ultrasound Med. Biol. 34, 902–912 (2008).
195. Rice, T. K., Schork, N. J. & Rao, D. C. Methods for Handling Multiple Testing. Adv. Genet. 60, 293–308 (2008).
196. Schreyer, A. G. et al. Microcirculation and perfusion with contrast enhanced ultrasound (CEUS) in Crohn’s disease: First results with linear contrast harmonic imaging (CHI). Clin. Hemorheol. Microcirc. 40, 143–155 (2008).
197. Martínez, M. J., Ripollés, T., Paredes, J. M., Blanc, E. & Martí-Bonmatí, L. Assessment of the extension and the inflammatory activity in Crohn’s disease: comparison of ultrasound and MRI. Abdom. Imaging 34, 141–148 (2009).
198. Ripollés, T. et al. Effectiveness of contrast-enhanced ultrasound for characterisation of intestinal inflammation in Crohn’s disease: A comparison with surgical histopathology analysis. J. Crohns. Colitis 7,
67
120–8 (2013).
199. Serafin, Z., Białecki, M., Białecka, A., Sconfienza, L. M. & Kłopocka, M. Contrast-enhanced Ultrasound for Detection of Crohn’s Disease Activity: Systematic Review and Meta-analysis. J. Crohn’s Colitis 10, 354–362 (2016).
200. Cheng, W. et al. Preliminary Analysis of Clinical Situations Involved in Quantification of Contrast-enhanced Ultrasound in Crohn’s Disease. Ultrasound Med. Biol. 1–8 (2016). doi:10.1016/j.ultrasmedbio.2016.03.001
201. Quaia, E., Sozzi, M., Angileri, R., Gennari, A. G. & Cova, M. A. Time-Intensity Curves Obtained after Microbubble Injection Can Be Used to Differentiate Responders from Nonresponders among Patients with Clinically Active Crohn Disease after 6 Weeks of Pharmacologic Treatment. Radiology 000, 152461 (2016).
202. Bonnin, P., Coelho, J., Pocard, M., Levy, B. I. & Marteau, P. Anti-TNFα therapy early improves hemodynamics in local intestinal and extraintestinal circulations in active Crohn’s disease. J. Crohns. Colitis 7, 451–9 (2013).
203. RSNA. AIUM/QIBA Ultrasound Volume Blood Flow Biomarker Ctte. at <http://qibawiki.rsna.org/index.php/AIUM/QIBA_Ultrasound_Volume_Blood_Flow_Biomarker_Ctte>
68
Appendices
Appendix I: Harvey-Bradshaw spørgeskema
Forsøgsperson:
Patientvejledning: Udfyld nedenstående. Dine svar skal omfatte de forudgående 24 timer eller dagen i går
Nr. Spørgsmål
Score
1. Antal vandtynde eller meget bløde afføringer pr. dag:
2. Mavesmerter: (0 = ingen, 1 = mild, 2 = moderat, 3 = svær)
3. Generel velvære: (0 = Godt, 1 = lidt under middel, 2= dårlig, 3 = meget dårlig, 4 = forfærdelig)
4. Sum af følgende tilstande pt. har dags dato, 1 point pr. tilstand: 1. Arthritis/artralgier (gigt/ledsmerter) 2. Iritis/uveitis (øjenbetændelse) 3. Erythema nodosum 4. Pyoderma gangrenosum 5. Aphtøs stomatitis (hud / mund læsioner) 6. Anal fissur 7. Fistel 8. absces (perianalsygdom) 9. Andre fistler
5. Abdominal udfyldning (0 = Nej, 1 = Tvivlsom, 2 = Sikker, 3 = Sikker og øm)
SUM:
Udfyldt af: Læge/spl. Dato:
69
Appendix II: CDAI spørgeskema
Forsøgsperson: 7 dage forudgået næste kontrol.
dd/mm
Dato for næste kontrol
Dato: Spørgsmål
/ / / / / / / /
1. Antal af vandtynde eller meget bløde afføringer?
Sum:
2. Mavesmerter? (0 = ingen, 1 = mild, 2 = moderat, 3 = svær)
Sum:
3. Generelt velvære? (0 = Godt, 1 = lidt under middel, 2= dårlig, 3 = meget dårlig, 4 = forfærdelig)
Sum:
Patientvejledning: Udfyld ovenstående hver dag før sengetid. Dine svar skal omfatte de forudgående 24 timer. Skriv score i kolonnen svarende til korrekt dato.
Nr. Spørgsmål Sum Faktor Subtotal 1. Sum af antal vandtynde eller meget bløde afføringer dag 1-7 x 2 = 2. Sum af mavesmerter dag 1-7 x 5 = 3. Sum af generet velvære dag 1-7 x 7 = 4. Sum af 6 følgende tilstande pt. har d.d.
10. Arthritis/artralgier (gigt/ledsmerter) 11. iritis/uveitis (øjenbetændelse) 12. erythema nodosum/pyoderma gangrenosum/aphtøs
stomatitis (hud / mund læsioner) 13. Anal fissur, -fistler eller –absces (perianalsygdom) 14. Andre fistler 15. Temperatur ≥ 37,8 (i løbet af sidste uge)
x
20
=
5. Peristaltikhæmmende midler (Imodium / morfin) (0 = Nej, 1= ja)
x
30
=
6. Abdominal udfyldning (0 = Nej, 2 = Tvivlsom, 5 = Sikker)
x
10
=
7. Mænd (47- Hct) Hæmatokrit (Hct): ____ % Kvinder (42 – Hct) Subtotal:
x
6
=
8. Vægt:______ Standard vægt:______ Procentuel ændring x 1 =
SUM:
Udfyldt af: Læge/spl. Dato:
70
Appendix III
Sequences True FISP
T2w Single-shot Turbo Spin-Echo
T2w Single-shot Turbo Spin-Echo fat sat
T1w Turboflash
fat sat T2w
TRUEFISP T1w Spoiled
3D flash T1w Spoiled
3d flash Turboflash
fat sat T1w VIBE Parameter Image plane(s) Coronal Coronal Coronal Coronal Axial Coronal Coronal Coronal +
Axial Coronal
Field of view (mm)
450 x 366 450 x 338 450 x 338 450 x 366 400 x 300 360 x 240 360 x 240 450 x 366 / 400 x 300
400 x 400
No. of sections 5 20 20 24 30 20 20 24 / 22 96 No. of stacks 2-3 1 1 1 3-4 1 1 1 / 3-4 1 Repetition time (msec)
36.9 2000 2000 212 3.42 3.00 3.00 212 / 203 9:33
Echo time (msec) 1.04 81 81 4.76 1.45 0.82 0.82 4.76 / 4.76 4:44 Acquisition time per stack (min)
0:19 1:20 1:20 0:40 0:12 0:05:4 3:36 0:40 / 0:36 0:20
No. of sequential acquisitions
10 1 1 1 1 1 120 1 1
Matrix 192 x 192 320 x 260 320 x 260 256 x 205 256 x 256 256 x 128 256 x 128 256 x 205 256 x 154 Section thickness (mm)
10 6 6 5 4 5 5 5 / 5 2.5
Section gap (mm) 2 2 2 2.5 0 1 1 2.5 / 0.5 0.5 Turbo factor NA 194 194 NA NA NA NA NA NA Parallel imaging* GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA Flip angle(s) (degrees)
77 150§ 150§ 70 60 5, 10, 15, 20, 25
24 70 / 70 20
Magnetic resonance enterography parameters
*GRAPPA = Generalized autocalibrating partially parallel acquisitions, applied left to right with a factor of two in conjunction with a body matrix coil. § Refocusing flip angles NA = Not applicable
Paper I
Dynamic Contrast Enhanced Magnetic Resonance Enterography
and Dynamic Contrast Enhanced Ultrasonography in Crohn Disease: An observational
comparison study
Rune Wilkens1,2, David A Peters3, Agnete H. Nielsen1, Valeriya P Hovgaard1, Henning
Glerup1*, Klaus Krogh2*
1) Diagnostic Centre, divisions of Medicine and Radiology, University Research Clinic for
Innovative Patient Pathways, Silkeborg Regional Hospital, Silkeborg, Falkevej 1-3,
8600 Silkeborg, Denmark
2) Department of Hepatology and Gastroenterology, Aarhus University Hospital,
Noerrebrogade 44, 8000 Aarhus C, Denmark
3) Department of Clinical Engineering, Aarhus University Hospital, Oluf Palmes Alle 15,
8200 Aarhus N, Denmark
*) These authors contributed equally.
Corresponding author: Rune Wilkens
Diagnostic Centre, University Research Clinic for Innovative Patient Pathways
Silkeborg Regional Hospital
Falkevej 1-3, 8600 Silkeborg, Denmark
Phone +45 7841 5000, E-mail: [email protected]
Manuscript type: Original research
Funding: This study was funded with an unrestricted grant by AbbVie, Denmark.
Peer Review O
nly
Dynamic Contrast Enhanced Magnetic Resonance
Enterography
and Dynamic Contrast Enhanced Ultrasonography in Crohn
Disease: An observational comparison study
Journal: Ultrasound International Open
Manuscript ID Draft
Manuscript Type: Original Article
Keywords: small bowe < AREAS, STRUCTURES & SYSTEMS, MR-diffusion/perfusion < METHODS & TECHNIQUES, ultrasound < METHODS & TECHNIQUES, contrast agents < THEMES, hemodynamics/flow dynamics < THEMES
Abstract:
Purpose Cross sectional imaging methods are important for objective evaluation of small intestinal inflammation in Crohn’s disease (CD). The primary aim was to compare relative parameters of intestinal perfusion between contrast enhanced ultrasonography (CEUS) and dynamic contrast enhanced magnetic resonance enterography (DCE-MRE) in CD. Furthermore, we aimed at testing the repeatability of regions of interest
(ROIs) for CEUS. Methods This prospective study included 25 patients, 12 females (age 37, range 19-66), with moderate to severe CD and a bowel wall thickness >3 mm, evaluated with DCE-MRE and CEUS. CEUS bolus injection was performed twice for repeatability and analysed in VueBox®. Correlations between modalities were described with Spearman’s rho, limits of agreement (LoA) and intraclass correlation coefficient (ICC). ROI repeatability for CEUS was assessed. Results The correlation between modalities was good and very good for bowel wall thickness (ICC= 0.71, P<.001) and length of the inflamed segment (ICC=0.89, P<0.001). Moderate-weak correlations were found for the time-intensity curve parameters: peak intensity (r=0.59, P=0.006),
maximum wash-in-rate (r=0.62, P=0.004), and wash-in perfusion index (r=0.47, P=0.036). Best CEUS repeatability for Peak Enhancement was a mean difference of 0.73 dB (95% CI: 0.17 to 1.28, P=0.01) and 95% LoA from -3.8 to 5.3 dB. Good quality of curve fit improved LoA to -2.3 to 2.8 dB. Conclusions The relative perfusion of small intestinal CD assessed with DCE-MRE and CEUS shows merely a moderate correlation. Applying strict criteria for ROIs are important and allows for good CEUS repeatability.
Note: The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online.
Video 1.wmv
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
Peer Review O
nly
1
Dynamic Contrast Enhanced Magnetic Resonance Enterography
and Dynamic Contrast Enhanced Ultrasonography in Crohn Disease: An observational
comparison study
Manuscript type: Original Articles
Keywords Ultrasonography Magnetic resonance imaging Crohn’s Disease Reproducibility of results Contrast media
Word count; main body: 3388
Abbreviations
AIU = Arbitrary intensity units
CEUS = Contrast enhanced ultrasound
CD = Crohn Disease
CDAI = Crohn Disease Activity Index
CDI = Colour Doppler imaging
DCE-MRE = Dynamic Contrast Enhanced Magnetic Resonance Enterography
HBI = Harvey Bradshaw Index
ICC = Intraclass correlation coefficient
LoA = Limits of agreement
QoF = Quality of Fit
ROI = Region of interest
TIC = Time intensity curve
US = Ultrasonography
Page 2 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
3
Introduction
In Crohn’s disease (CD) grading of disease activity has shifted from subjective clinical scoring
systems towards more objective measurements, in combination with patient reported
outcomes [1]. Endoscopy is a well-recognized gold standard for luminal disease in the colon,
rectum, and the terminal ileum. However, endoscopy is of limited use in stricturing and
proximal disease [2]. This calls for cross sectional imaging methods with objective parameters
of disease severity [1,3]. Currently no single imaging modality exists as a gold standard for
transmural disease of the small intestine [4].
The most consistent characteristic of disease activity on imaging is an increased bowel wall
thickness of more than 3 mm [3,5,6]. Nevertheless, the intestinal wall may be thickened not
only by active disease but also by fibrosis [6,7]. Other features of inflammatory activity
comprise ulcerations, T2-hypersignal, perimural signal, contrast enhancement, comb sign,
enlarged lymph nodes, fistulas, abscesses and strictures described in the development of the
MR intestinal activity score and MR enterography global assessment [6,8,9]. Unfortunately,
experts do not agree about the importance of the individual findings [10].
In recent classifications, increased contrast enhancement is considered a relevant marker of
disease activity [6,8,11,12]. This is in accordance with the characteristics of active
inflammation including dilated leaking vessels [13] and neoangiogenesis [14]. Additionally,
microvascular density has been shown to correlate with intensity on dynamic contrast
enhanced ultrasound (CEUS) [15]. Therefore, dynamic imaging techniques could potentially
be used for evaluating disease activity and efficacy of treatment [6,16].
Page 4 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
4
The two promising modalities to assess relative bowel wall perfusion are CEUS and dynamic
contrast enhanced MR enterography (DCE-MRE). However, there are significant differences
in contrast behaviour between modalities. MR gadolinium based contrast agents are relatively
small and exhibits extravasation over time whereas CEUS gas-filled lipid-shell contrast acts
as a true intravascular agent. The time-intensity curves (TICs) recorded from the former are
therefore a combination of perfusion and permeability, rather than perfusion alone. The
relation between signal intensity and MRI contrast agent concentration is complex and
depends on a number of parameters such as the native tissue relaxation rate, relaxivity of the
contrast agent, local field inhomogeneity’s and the applied flip angle and inversion-recovery
time [17]. US contrast agent on the other hand has a direct correlation with signal intensity
measured in dB [18]. Since perfusion is difficult to measure if the bowel wall is less than 3 mm
thick [19] the parameters should only be used for grading disease activity or to follow
treatment efficacy [5,20].
In the present study, we hypothesized that intensity and time parameters of the initial time-
intensity curves correlate well between modalities as the amount of MR contrast which is
extravacated during the initial pass is low.
The main objective of this study was to compare objective parameters of relative perfusion
obtained with DCE-MRE and CEUS in patients with moderate to severe CD. Our secondary
objectives were to test the repeatability of regions of interest (ROIs) for CEUS and to evaluate
inter-rater reliability of CD characteristics assessed with MRE and US.
Page 5 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
5
Materials and Methods
This GCP monitored prospective double blinded observational study was approved by the
Danish national authorities (2011-005886-19) and the local research ethics committee (1-10-
72-340-12) for the off-label use of US contrast agents. All participants gave written informed
consent before entering the study. Inclusion criteria were known CD with moderate to severe
clinical activity based on either CD activity index [21] (CDAI) >220 or Harvey Bradshaw Index
[22] (HBI) >7. Furthermore, patients had to be ≥18 years of age and have an US detectable
intestinal segment with bowel wall thickness >3 mm. Patients were excluded if they were
pregnant, breast feeding or had any contraindications for DCE-MRE or CEUS.
Twenty-five consecutive patients (mean age, 37 years; range 19-66 years; 12 females) were
recruited for the study during September 2012 to March 2014. Due to screening failure two
patients were excluded and another two were subsequently recruited to reach the desired
inclusion of 25 patients, see Fig. 1. The Montreal classification [23], CDAI, HBI,
gastrointestinal symptoms, smoking status, and medical history were recorded and blood and
stool samples were taken at the first visit. For full patient demographics see Tab. 1.
Ultrasonography
Participants were investigated after four hours fast. Ultrasonography (US) was performed by
one physician (RW) experienced with the procedure through two years. The investigator was
blinded to the MRE scan and biochemical results; however, the patients´ symptoms were
known. An Acuson S3000 ultrasound machine with a 4-9 MHz linear matrix transducer and a
1-6 MHz curvilinear transducer was used (Siemens Medical Solutions, Malvern, PA). Colour
Doppler imaging (CDI) was set with a transmit frequency of 6.75 MHz, gain 1 dB, pulsed
Page 6 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
6
repetition frequency 1099, low wall filter of 2, and a scale of 6 cm/s. The most severely
inflamed bowel segment was identified based on wall thickness and highest CDI signal score
according to the Limberg classification [24]. Total length of each affected segment, bowel wall
pattern, presence of ulcers, stenosis and prestenotic dilatation were registered.
All CEUS scans were performed on the Acuson machine using the 9L4 probe. Settings were;
fixed mechanical index of 0.06-0.08, dynamic range 80, frame rate of 10 per second,
frequency 4 MHz, and the focal zone beneath the bowel wall. Sulphur hexafluoride
microbubbles (SonoVue®; Bracco Imaging, Milan, Italy) 2.4 ml x 2 was injected by trained
nurses followed by a 5 ml saline flush over 2 seconds. Scans were recorded for 90 seconds.
Scan plane was kept constant and patients were instructed in gentle breathing. More than 5
minutes after first injection, the scan was repeated at the same spot, but in a different scan
plane, to cover the segment in both the transverse and longitudinal axes. The chosen bowel
segments were either terminal ileum, neo-terminal ileum or proximal ileum. The location for
CEUS was determined as the most inflamed area of the segment according to a prior
classification [11].
Analysis of Contrast Enhanced Ultrasonography
Cineloop files were exported in DICOM format, re-linearized, and quantified on VueBox® 5.1
(Bracco Suisse SA, Geneva, Switzerland) as earlier described [16].
If possible, four ROIs were drawn using the following criteria: all ROIs had to be larger than
0.1 cm2 and within the bowel wall at all times. Shapes and placement of ROIs were optimized
to obtain a quality of fit (QoF) of the fitted curve larger than 90 % or as high as possible. Built-
in motion compensation was applied whenever beneficial. The first ROI was drawn as large
Page 7 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
7
as possible and typically covering the full bowel wall thickness of the anterior and posterior
bowel wall avoiding the lumen. VueBox includes the possibility to apply a heat map for the
parameters of interest. Three additional ROIs were placed in areas with highest peak
enhancement according to the heat map and without overlapping, Fig. 2 and video (Online
Resource). Analyses were then compared for repeatability between the largest ROI, the
maximum peak ROI and mean of the three latter ROIs, exhibiting a QOF >85%, entitled mean
ROI. The average (log-converted) values of the best reproducible method were subsequently
chosen for comparison with DCE-MRE results. Data post processing were badge-analysed by
the same investigator more than six months after the US and clinical scoring of the last
patient to ensure effective blinding of data.
Dynamic Contrast Enhanced Magnetic Resonance Enterography
Patients were instructed to fast for four hours and drink one litre of oral contrast one hour
before the scan. Oral contrast comprised a suspension of 125 ml mannitol 15 % (Fresenius
Kabi, Bad Homburg, Germany) in 875 ml of tap water, 30 ml psyllium HUSK® Fibre, and ice
cubes. Peristalsis was suppressed by intravenous injection of 20 mg Hyoscine butylbromid
(Buscopan®; Boehringer Ingelheim, Ingelheim, Germany) prior to non-dynamic sequences
and repeated before contrast injection. Images were acquired using a 1.5 T MR unit (Avanto;
Siemens, Erlangen, Germany) with patients in the prone position. The intravenous contrast
agent used was gadoterate meglumine (Dotarem®; Guerbet, Villepinte, France) with 0.2 mg /
kg bodyweight at 5 ml/s followed by a 24 ml saline flush. Patients were instructed to
hyperventilate prior to a long breath hold followed by gentle breathing. MR scanning protocols
can be seen in Table 2.
Page 8 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
8
Analysis of Magnetic Resonance enterography
Interpretation of MRE based pathoanatomical data was performed individually by two
radiologists with nine (AHN) and four (VPH) years’ experience respectively. Both were blinded
to the findings on US. Maximum wall thickness and total length of disease were described in
continues measurements for the most pathological bowel segment. Average values between
readers were used for comparison with bowel wall thickness and length of involvement
measured on US. Presence of mural oedema, ulcers, wall enhancement pattern, perimural
involvement, and presence of complications like stenosis and penetrating disease were also
registered for the segment, based on the MRE global score [8,9].
The ROI for DCE-MRE analysis was placed in the bowel wall at the site of largest wall
thickness and highest enhancement within the same bowel segment examined by CEUS,
using a custom made program in MATLAB® (MathWorks®, Natick, MA). The ROI was
manually moved in order to stay within the bowel wall during the dynamic series, Fig. 3. TICs
were interpolated using a cubic spline. This interpolated curve was used to derive the
parameters described in Table 3.
Statistical analysis
Statistical analysis was performed using Stata 13.1 for MAC (Stata Corp LP, College Station,
TX). If no disease was observed on MRE, bowel wall thickness was set at 3 mm and length 0
cm. Existing data in the literature were to scarce to allow for a power calculation, however we
estimated 25 patients to be sufficient. None of the linearized CEUS intensity data, expressed
as arbitrary intensity units (AIU), followed a Gaussian distribution. Hence, they were log-
converted as by default in US systems using 10 x log10 (AIU) and expressed in dB for further
Page 9 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
9
analysis [11]. Time parameters for both CEUS and DCE-MRE, C-reactive protein, and faecal-
calprotectin were analysed log-converted. Correlations between DCE-MRE and CEUS TIC
parameters were described with Spearman’s correlation, since DCE-MRE data were slightly
skewed [25] even with log-conversion. Correlation coefficients were interpreted as earlier
suggested [25]. CEUS repeatability was assessed with 95 % limits of agreement (LoA), using
mixed effect model with independent residuals per ROI [26]. Data for length of disease and
MRE global score did not follow a Gaussian distribution regardless of log-conversion. Hence
only Intraclass correlation coefficients (ICC) are reported for these data. P values < 0.05 were
considered statistically significant. Data were not corrected for multiple testing. However, final
conclusions were drawn having multiple testing in mind.
Page 10 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
10
Results
All but two patients had CEUS and DCE-MRE performed within the same day. The remaining
two patients were scanned four days apart. All patients completed both examinations without
adverse events or serious discomfort.
Pathoanatomical data
The thickest bowel wall segments had a mean of 7.9 mm, (range 4 to 12 mm) when assessed
with US and 8.1 mm (range 4 to 14.5 mm) when assessed with MRE. The mean difference
was 0.22 mm (LoA -4.3 to 3.9) and the corresponding ICC was 0.71 (0.44 to 0.86, P < 0.001)
(Fig. 4). The median length of the inflamed segment was 15 cm (range 3 to 57 cm) on US and
12 cm (range 1 to 70 cm) on MRE. The corresponding ICC was 0.89 (0.76-0.95, P < 0.001).
Associations between perfusion data from Contrast Enhanced Ultrasonography and Dynamic
Contrast Enhanced Magnetic Resonance Enterography
Data from three MRE and two CEUS scans were excluded from further analysis, Fig. 1. All
segments compared were either from the terminal ileum (n = 19) or the ileum (n = 1).
Total area under curve, including wash-in and wash-out for CEUS and wash-in and plateau-
phase at 70 seconds for DCE-MRE, had a low and insignificant correlation between the two
methods (r = 0.16, P = 0.494). Wash-in area under curve also showed poor correlation (r =
0.18, P = 0.443). Likewise, rise time and time to peak showed no correlation between
modalities (r = 0.11, P = 0.659 and r = 0.02, P = 0.930 respectively). Slope and maximum
Page 11 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
11
slope for DCE-MRE and wash-in rate for CEUS correlated moderately well (r = 0.60, P =
0.005, and r = 0.62, P = 0.004), Fig. 5. Peak intensity and Wash-in perfusion index
determined by each of the two methods were moderately and moderate to weakly correlated
(r = 0.59, P = 0.006 and r = 0.47, P = 0.036 respectively). No significant correlation was found
between peak enhancement of CEUS and by DCE-MRE (r = 0.41, P = 0.076).
Repeatability of Contrast Enhanced Ultrasonography and reproducibility of Magnetic
Resonance Enterography
For CEUS, the smallest mean difference between two contrast injections was found for the
maximum peak ROI. However, the narrowest limit of agreement was consistently found for
the mean ROIs, see Tab. 4. In a post hoc analysis restricted to ROIs with QoF >90%, or if two
ROIs could not qualify for this, at least one ROI with QoF >85% and the other >90%, LoA
could be further reduced, see Fig. 6 and Tab. 4 for all LoA, Tab. 5 for QoF.
MRE interrater variability for bowel wall thickness showed an ICC = 0.83 (0.66 to 0.92 P <
0.001) and ICC = 0.76 (0.51 to 0.89 P < 0.001) for length of involvement. The mean
difference was 1.2 mm with 95% LoA from -3.8 to 3.6 mm for wall thickness. For
reproducibility on MR enterography global score, see Tab. 6.
Page 12 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
12
Discussion
The present study compares CEUS and MRE for description of the severity of ongoing small
intestinal inflammation in CD. Even though correlations between basic pathoanatomical
findings were good between the two modalities, our main finding was only a moderate to
weak correlation when assessing relative changes in perfusion.
Since clinical activity scores for CD are poorly associated with the presence of active
inflammation and equally poorly predict long-term outcome, their use should be supplemented
by objective markers [1]. Therefore, cross sectional imaging is of paramount importance as an
adjunct to endoscopy [3]. Active inflammation is potentially treatable with effective medication
but needs objective description and repeated follow-up to determine treatment response.
Stenoses caused by fibrosis do not respond to medical treatment and need surgery [27]. In
contrast to fibrosis [28], active inflammation is causing hyperaemia and hyper perfusion [29]
which may be quantified by CEUS and MRE.
A few previous studies have shown significant correlation between dynamic contrast
enhanced cross sectional imaging and clinical disease activity, biochemistry [30], or a
combined score for response [31], the need for surgery [19], and change in medication
[16,20,32]. Other authors aimed at more objective endpoints like micro vessel density [15] or
mucosal healing or inactive disease defined by endoscopy [20]. However, the studies do not
agree about which TIC parameters are important. Romanini et al.[15], Saevik et al.[16] and
Horje et al.[33] found a statistically significant difference for almost all TIC parameters and
disease activity, whereas others only showed significance for time-to-peak [30], area under
curve [31], or peak enhancement [32]. In this present study, we found a significant correlation
Page 13 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
13
between the two modalities when describing peak and slope-related parameters but,
surprisingly, not for area under curve, peak enhancement or rise time.
No consensus on how to perform or quantify intestinal perfusion measurement exists.
Consequently, the heterogeneity between studies makes them difficult to compare or
reproduce. For example, only few authors have described the placement and analysis of
ROIs for CEUS in detail [20] and only one group did log-transformation of data before
statistical analysis [34].
Several MRE studies are using change in contrast enhancement as an indicator for disease
activity [12,28,29]. However, most studies have not applied a dynamic protocol and only use
few image acquisitions or the relative change over a predefined timespan after injection.
Taylor et al. found an inverse correlation with slope of enhancement on MRE and micro
vessel density [35], which is opposite to the finding by Romanini et al. using CEUS [15].
These studies and our findings, showing lack of good correlation, suggests that the two
modalities are measuring somewhat dissimilar components of “perfusion”, with DCE-MRE TIC
measurements being a mixture of perfusion and extravasation. Taylor et al. also found a
direct correlation with slope of enhancement and disease duration and speculated that
increased enhancement could be caused by ischemia and arteriolar stenosis [35].
In the present study, interrater variability for structural MRE findings was comparable to those
reported in previous studies [36]. Surprisingly, only a moderate correlation in wash-in rate and
peak intensity could be established between DCE-MRE and CEUS. Lack of strong correlation
between modalities may likely be due to the dissimilar distribution nature of contrast agents,
relatively poor MR time resolution, and perhaps also the administration technique. In the
Page 14 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
14
optimal setting, absolute perfusion measurements of tissue blood flow, blood volume and
mean transit time should be compared. However, this is complicated, even when using MR
contrast agents which act as a true intra vascular agents, e.g. in cerebral perfusion [37].
The present study demonstrates the consequence of ROI selection in quantification of
perfusion in CD. Our data emphasizes the importance of TIC QoF for reliability and
reproducibility. Poor QoF [33], e.g. by fitting a burst-replenishment curve on a bolus injection
examination [38], will obviously give unreliable results. We therefore recommend that curve
fitting quality should be reported alongside test results in future publications. Also, using low
perfused tissue as reference, will cause high uncertainty of the final results [39].
This study has some limitations. We did not apply Tofts (extended) model or any other model
to reflect pharmacokinetic parameters, like absolute blood flow or permeability measures for
pathological conditions [40], as our T1 measurements employing the variable flip angle
technique gave unreliable results [41]. As an alternative we used the absolute signal
difference technique instead, which has recently been shown to have a linear relation to
contrast agent concentration at low contrast concentrations [42,43].
Furthermore, CEUS was performed without deconvolution [44] thereby only providing
semiquantitative measurements. Deconvolution is complex and relies on several assumptions
[44,45], that are difficult to fulfil and thus rarely used in daily practice nor in scientific work. A
method called bolus-tracking and burst-replenishment is described by Jirik et al.[46] but the
repeatability is not yet established in humans.
Based on existing guidelines, bolus injection techniques were used for CEUS and DCE-MRE.
A fixed dose and manual injection of SonoVue was chosen for CEUS quantification [47]. For
Page 15 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
15
DCE-MRE, gadolinium dose was bodyweight-dependent and administered with an automatic
pump. We chose the bodyweight-dependent dose over the fixed dose based on the general
recommendation for MR contrast administration [12]. We did not measure the exact length
defined from an anatomical landmark, like the ileocecal valve, to ensure identical ROI location
between modalities. However, same bowel segment was analysed for each patient and the
most enhanced and thickest area was chosen.
Since no guidelines exist on the optimal scan plane, two different scan planes were employed
for assessment of the repeatability of CEUS ROIs. Full repeatability of findings from the same
segment in an identical scan plane along with reproducibility between investigators are still
warranted. However, based on the present findings within patient repeatability seems
acceptable for the clinical use of CEUS in CD, especially when applying strict criteria for size
and QoF. Lack of strict criteria or using a low perfused tissue as reference tissue will lead to
poor reproducibility[34,39].
We chose to restrict the inclusion of patients to those with moderate to severe disease activity
based on clinical symptoms. Investigating perfusion in a normal bowel wall is difficult because
of peristalsis and small ROI size results in poor QoF. However, clinical symptoms are often
poorly correlated to objective signs of active disease. Based on wall thickness and
biochemical findings, we covered the full disease spectrum of active small bowel disease.
In summary, there is only a moderate to weak correlation between CEUS and DCE-
MRE slope-related and peak intensity parameters in CD. This is likely to be caused by the
inherent different nature of the contrast agents and scanning modalities. Additionally, we have
elucidated the importance of quality of fit for ROI selection in CEUS. The value of perfusion
Page 16 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
16
measurements as activity assessment in CD still remains to be clarified and validated against
more objective endpoints. Further studies should aim at a global standardization of methods
for assessment of intestinal perfusion.
Page 17 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
17
References
[1] Peyrin-Biroulet L, Sandborn W, Sands BE, Reinisch W, Bemelman W, Bryant R V, et al.
Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE): Determining
Therapeutic Goals for Treat-to-Target. Am J Gastroenterol 2015;110:1324–38.
doi:10.1038/ajg.2015.233.
[2] Wilkens R, Novak KL, Lebeuf-Taylor E, Wilson SR. Impact of intestinal ultrasound on
classification and management of Crohn disease patients with inconclusive
colonoscopy. Can J Gastroenterol Hepatol 2015;29:1–7.
[3] Panes J, Bouhnik Y, Reinisch W, Stoker J, Taylor SA, Baumgart DC, et al. Imaging
techniques for assessment of inflammatory bowel disease: Joint ECCO and ESGAR
evidence-based consensus guidelines. J Crohns Colitis 2013;7:556–85.
doi:10.1016/j.crohns.2013.02.020.
[4] Taylor S, Mallett S, Bhatnagar G, Bloom S, Gupta A, Halligan S, et al. METRIC
(MREnterography or ulTRasound in Crohn’s disease): a study protocol for a
multicentre, non-randomised, single-arm, prospective comparison study of magnetic
resonance enterography and small bowel ultrasound compared to a reference standard
in those . BMC Gastroenterol 2014;14:142. doi:10.1186/1471-230X-14-142.
[5] Horsthuis K, Bipat S, Bennink RJ, Stoker J. Inflammatory Bowel Disease Diagnosed
with US, MR, Scintigraphy, and CT: Meta-analysis of Prospective Studies. Radiology
2008;247:64–79. doi:10.1148/radiol.2471070611.
[6] Murphy DJ, Smyth AE, McEvoy SH, Gibson DJ, Doherty GA, Malone DE.
Subclassification of small bowel Crohn’s disease using magnetic resonance
Page 18 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
18
enterography: A review using evidence-based medicine methodology. Clin Radiol
2015;70:1336–43. doi:10.1016/j.crad.2015.07.008.
[7] Zappa M, Stefanescu C, Cazals-Hatem D, Bretagnol F, Deschamps L, Attar A, et al.
Which magnetic resonance imaging findings accurately evaluate inflammation in small
bowel Crohn’s disease? A retrospective comparison with surgical pathologic analysis.
Inflamm Bowel Dis 2011;17:984–93. doi:10.1002/ibd.21414.
[8] Steward MJ, Punwani S, Proctor I, Adjei-Gyamfi Y, Chatterjee F, Bloom S, et al. Non-
perforating small bowel Crohn’s disease assessed by MRI enterography: derivation and
histopathological validation of an MR-based activity index. Eur J Radiol 2012;81:2080–
8. doi:10.1016/j.ejrad.2011.07.013.
[9] Makanyanga JC, Pendsé D, Dikaios N, Bloom S, McCartney S, Helbren E, et al.
Evaluation of Crohn’s disease activity: Initial validation of a magnetic resonance
enterography global score (MEGS) against faecal calprotectin. Eur Radiol
2014;24:277–87. doi:10.1007/s00330-013-3010-z.
[10] Ziech MLW, Bossuyt PMM, Laghi A, Lauenstein TC, Taylor SA, Stoker J. Grading
luminal Crohn’s disease: which MRI features are considered as important? Eur J Radiol
2012;81:e467–72. doi:10.1016/j.ejrad.2011.05.033.
[11] Medellin-Kowalewski A, Wilkens R, Wilson A, Ruan J, Wilson SR. Quantitative
Contrast-Enhanced Ultrasound Parameters in Crohn Disease: Their Role in Disease
Activity Determination With Ultrasound. AJR Am J Roentgenol 2016;206:64–73.
doi:10.2214/AJR.15.14506.
Page 19 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
19
[12] Rimola J, Rodriguez S, García-Bosch O, Ordás I, Ayala E, Aceituno M, et al. Magnetic
resonance for assessment of disease activity and severity in ileocolonic Crohn’s
disease. Gut 2009;58:1113–20. doi:10.1136/gut.2008.167957.
[13] Darwin F. On the Primary Vascular Dilatation in Acute Inflammation. J Anat Physiol
1875;10:1–16.
[14] Danese S, Sans M, de la Motte C, Graziani C, West G, Phillips MH, et al. Angiogenesis
as a novel component of inflammatory bowel disease pathogenesis. Gastroenterology
2006;130:2060–73. doi:10.1053/j.gastro.2006.03.054.
[15] Romanini L, Passamonti M, Navarria M, Lanzarotto F, Villanacci V, Grazioli L, et al.
Quantitative analysis of contrast-enhanced ultrasonography of the bowel wall can
predict disease activity in inflammatory bowel disease. Eur J Radiol 2014;83:1317–23.
doi:10.1016/j.ejrad.2014.05.012.
[16] Saevik F, Nylund K, Hausken T, Odegaard S, Gilja OH. Bowel Perfusion Measured with
Dynamic Contrast-enhanced Ultrasound Predicts Treatment Outcome in Patients with
Crohn’s Disease. Inflamm Bowel Dis 2014;20:2029–37.
doi:10.1097/MIB.0000000000000159.
[17] Buckley DL. Uncertainty in the analysis of tracer kinetics using dynamic contrast-
enhancedT1-weighted MRI. Magn Reson Med 2002;47:601–6. doi:10.1002/mrm.10080.
[18] Payen T, Coron A, Lamuraglia M, Le Guillou-Buffello D, Gaud E, Arditi M, et al. Echo-
power estimation from log-compressed video data in dynamic contrast-enhanced
ultrasound imaging. Ultrasound Med Biol 2013;39:1826–37.
Page 20 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
20
doi:10.1016/j.ultrasmedbio.2013.03.022.
[19] Nylund K, Jirik R, Mezl M, Leh S, Hausken T, Pfeffer F, et al. Quantitative Contrast-
Enhanced Ultrasound Comparison Between Inflammatory and Fibrotic Lesions in
Patients With Crohn’s Disease. Ultrasound Med Biol 2013;39:1197–206.
doi:10.1016/j.ultrasmedbio.2013.01.020.
[20] Quaia E, Sozzi M, Angileri R, Gennari AG, Cova MA. Time-Intensity Curves Obtained
after Microbubble Injection Can Be Used to Differentiate Responders from
Nonresponders among Patients with Clinically Active Crohn Disease after 6 Weeks of
Pharmacologic Treatment. Radiology 2016;000:152461.
doi:10.1148/radiol.2016152461.
[21] Best WR, Becktel JM, Singleton JW, Kern Jr F. Development of a Crohn’s disease
activity index. National Cooperative Crohn's Disease Study. Gastroenterology
1976;70:439.
[22] Harvey RF, Bradshaw JM. A simple index of Crohn’s-disease activity. Lancet (London,
England) 1980;1:514.
[23] Silverberg MS, Satsangi J, Ahmad T, Arnott IDR, Bernstein CN, Brant SR, et al. Toward
an integrated clinical, molecular and serological classification of inflammatory bowel
disease: report of a Working Party of the 2005 Montreal World Congress of
Gastroenterology. Can J Gastroenterol 2005;19 Suppl A:5A – 36A.
[24] Limberg B. [Diagnosis of chronic inflammatory bowel disease by ultrasonography].
Zeitschrift Für Gastroenterol 1999;37:495–508.
Page 21 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
21
[25] Zou KH, Tuncali K, Silverman SG. Correlation and simple linear regression. Radiology
2003;227:617–22. doi:10.1148/radiol.2273011499.
[26] Raunig DL, McShane LM, Pennello G, Gatsonis C, Carson PL, Voyvodic JT, et al.
Quantitative imaging biomarkers: a review of statistical methods for technical
performance assessment. Stat Methods Med Res 2015;24:27–67.
doi:10.1177/0962280214537344.
[27] Spinelli A, Correale C, Szabo H, Montorsi M. Intestinal fibrosis in Crohn’s disease:
medical treatment or surgery? Curr Drug Targets 2010;11:242–8.
[28] Rimola J, Planell N, Rodríguez S, Delgado S, Ordás I, Ramírez-Morros A, et al.
Characterization of Inflammation and Fibrosis in Crohn’s Disease Lesions by Magnetic
Resonance Imaging. Am J Gastroenterol 2015;110:432–41. doi:10.1038/ajg.2014.424.
[29] Röttgen R, Grandke T, Grieser C, Lehmkuhl L, Hamm B, Lüdemann L. Measurement of
MRI enhancement kinetics for evaluation of inflammatory activity in Crohn’s disease.
Clin Imaging 2010;34:29–35. doi:10.1016/j.clinimag.2009.03.008.
[30] Girlich C, Schacherer D, Jung EM, Schreyer A, Büttner R. Comparison between a
clinical activity index (Harvey-Bradshaw-Index), laboratory inflammation markers and
quantitative assessment of bowel wall vascularization by contrast-enhanced ultrasound
in Crohn’s disease. Eur J Radiol 2012;81:1105–9. doi:10.1016/j.ejrad.2011.02.054.
[31] Quaia E, Cabibbo B, De Paoli L, Toscano W, Poillucci G, Cova MA. The value of time-
intensity curves obtained after microbubble contrast agent injection to discriminate
responders from non-responders to anti-inflammatory medication among patients with
Page 22 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
22
Crohn’s disease. Eur Radiol 2013;23:1650–9. doi:10.1007/s00330-012-2754-1.
[32] Moreno N, Ripollés T, Paredes JM, Ortiz I, Martínez MJ, López A, et al. Usefulness of
abdominal ultrasonography in the analysis of endoscopic activity in patients with
Crohn’s disease: changes following treatment with immunomodulators and/or anti-TNF
antibodies. J Crohns Colitis 2014;8:1079–87. doi:10.1016/j.crohns.2014.02.008.
[33] Horjus Talabur Horje CS, Bruijnen R, Roovers L, Groenen MJM, Joosten FBM, Wahab
PJ. Contrast Enhanced Abdominal Ultrasound in the Assessment of Ileal Inflammation
in Crohn’s Disease: A Comparison with MR Enterography. PLoS One
2015;10:e0136105. doi:10.1371/journal.pone.0136105.
[34] Socaciu M, Ciobanu L, Diaconu B, Hagiu C, Seicean A, Badea R. Non-Invasive
Assessment of Inflammation and Treatment Response in Patients with Crohn’s Disease
and Ulcerative Colitis using Contrast-Enhanced Ultrasonography Quantification. J
Gastrointestin Liver Dis 2015;24:457–65.
[35] Taylor SA, Punwani S, Rodriguez-Justo M, Bainbridge A, Greenhalgh R, De Vita E, et
al. Mural Crohn Disease: Correlation of Dynamic Contrast-enhanced MR Imaging
Findings with Angiogenesis and Inflammation at Histologic Examination—Pilot Study.
Radiology 2009;251:369–79. doi:10.1148/radiol.2512081292.
[36] Makanyanga JC, Pendsé D, Dikaios N, Bloom S, McCartney S, Helbren E, et al.
Evaluation of Crohn’s disease activity: initial validation of a magnetic resonance
enterography global score (MEGS) against faecal calprotectin. Eur Radiol
2014;24:277–87. doi:10.1007/s00330-013-3010-z.
Page 23 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
23
[37] Calamante F. Arterial input function in perfusion MRI: a comprehensive review. Prog
Nucl Magn Reson Spectrosc 2013;74:1–32. doi:10.1016/j.pnmrs.2013.04.002.
[38] Wong DD, Forbes GM, Zelesco M, Mason R, Pawlik J, Mendelson RM. Crohn’s disease
activity: quantitative contrast-enhanced ultrasound assessment. Abdom Imaging
2012;37:369–76. doi:10.1007/s00261-011-9792-z.
[39] Zink F, Kratzer W, Schmidt S, Oeztuerk S, Mason RA, Porzner M, et al. Comparison of
Two High-End Ultrasound Systems for Contrast-Enhanced Ultrasound Quantification of
Mural Microvascularity in Crohn’s Disease. Ultraschall Med 2016;37:74–81.
doi:10.1055/s-0034-1398746.
[40] Tofts PS, Brix G, Buckley DL, Evelhoch JL, Henderson E, Knopp M V, et al. Estimating
kinetic parameters from dynamic contrast-enhanced t1-weighted MRI of a diffusable
tracer: Standardized quantities and symbols. J Magn Reson Imaging 1999;10:223–32.
doi:10.1002/(SICI)1522-2586(199909)10:3<223::AID-JMRI2>3.0.CO;2-S.
[41] Sourbron SP, Buckley DL. On the scope and interpretation of the Tofts models for DCE-
MRI. Magn Reson Med 2011;66:735–45. doi:10.1002/mrm.22861.
[42] Sharman A, Zealley IA, Greenhalgh R, Bassett P, Taylor SA. MRI of small bowel
Crohn’s disease: determining the reproducibility of bowel wall gadolinium enhancement
measurements. Eur Radiol 2009;19:1960–7. doi:10.1007/s00330-009-1371-0.
[43] Wang P, Xue Y, Zhao X, Yu J, Rosen M, Song HK. Effects of flip angle uncertainty and
noise on the accuracy of DCE-MRI metrics: comparison between standard
concentration-based and signal difference methods. Magn Reson Imaging
Page 24 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
24
2015;33:166–73. doi:10.1016/j.mri.2014.10.005.
[44] Mezl M, Jirik R, Harabis V, Kolar R, Standara M, Nylund K, et al. Absolute ultrasound
perfusion parameter quantification of a tissue-mimicking phantom using bolus tracking
[Correspondence]. IEEE Trans Ultrason Ferroelectr Freq Control 2015;62:983–7.
doi:10.1109/TUFFC.2014.006896.
[45] Gauthier M, Tabarout F, Leguerney I, Polrot M, Pitre S, Peronneau P, et al.
Assessment of quantitative perfusion parameters by dynamic contrast-enhanced
sonography using a deconvolution method: an in vitro and in vivo study. J Ultrasound
Med 2012;31:595–608.
[46] Jirik R, Nylund K, Gilja OH, Mezl M, Harabis V, Kolar R, et al. Ultrasound perfusion
analysis combining bolus-tracking and burst-replenishment. IEEE Trans Ultrason
Ferroelectr Freq Control 2013;60:310–9. doi:10.1109/TUFFC.2013.2567.
[47] Piscaglia F, Nolsøe C, Dietrich CF, Cosgrove DO, Gilja OH, Bachmann Nielsen M, et
al. The EFSUMB Guidelines and Recommendations on the Clinical Practice of Contrast
Enhanced Ultrasound (CEUS): Update 2011 on non-hepatic applications. Ultraschall
Med 2011:33–59. doi:10.1055/s-0031-1281676.
Page 25 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Figure 1. Flow chart of inclusion and analysis. purple coloured boxes show reason for no inclusion, exclusion or no analysis. The large number of patients with insufficient contrast analysis of 2nd contrast injection, is due to in-and-out-of-plane motion artefacts in the non-optimal scan plane. CEUS = Contrast enhanced
ultrasound, DCE-MRE = Dynamic contrast enhanced magnetic resonance enterography, QoF = Quality of fit, inj. = injection(s)
279x200mm (150 x 150 DPI)
Page 26 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Figure 2. Contrast enhanced ultrasonography with SonoVue in a 66-year old woman. Quantification using VueBox. Upper left: Axial view of first bolus injection. On the left the contrast image is seen, on the right the corresponding b-mode image is shown. Outer turquoise oval shaped ROI is the area of investigation and
motion compensation. Green region of interest (ROI) is ROI1 and the largest possible. Yellow ROI is ROI2, purple ROI is ROI3, and the fourth ROI is white. Lower left: Corresponding time intensity curves. Upper
right: bowel in longitudinal scan after second bolus injection with 4 new ROIs. Lower right: TICs for injection 2. NB. Y-axis is slightly different from injection 1. Quality of Fit is shown in the box on the right, indicating the largest ROI (ROI1) has the best curve fit. ROI2 and ROI3 are almost identical. ROI = Region of interest.
316x162mm (150 x 150 DPI)
Page 27 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Figure 3. RoiTool. Dynamic contrast enhanced MR Enterography quantification, using RoiTool. Coronal T1 weighted spoiled 3d flash sequence of a 35-year old woman. A region of interest is drawn within the
thickened bowel at the terminal ileum. Corresponding graphs are produced in MatLab. Red line indicates the baseline. Bold blue line indicates the initial slope. Bold green line indicates the maximum slope. Yellow area shows the wash-in area under curve. The two thin lines can calculate the plateau over time (not utilized in
our study). ROI = region of interest. 240x189mm (150 x 150 DPI)
Page 28 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Figure 4: Limits of agreement for bowel wall thickness measured by ultrasound (US) and magnetic resonance enterography (MRE). Purple line shows the observed average agreement. Red lines indicate 95 %
limits of agreements and green line is the perfect average agreement. MRE = Magnetic resonance
enterography, US = Ultrasonography 201x146mm (150 x 150 DPI)
Page 29 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Figure 5 Scatter plot showing correlation between dynamic contrast enhanced magnetic resonance enterography and contrast enhanced ultrasound for maximum wash-in rate. Spearman’s rho = 0.618, P = .004. MRE = Magnetic resonance enterography. CEUS = Contrast enhanced ultrasound. AIU = Arbitrary
intensity units. 208x208mm (144 x 144 DPI)
Page 30 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Figure 6: Limits of agreement (LoA) for peak enhancement mean regions of interest. Purple line shows the observed average agreement. Red lines indicate 95 % limits of agreements and green line is the perfect
average agreement. ROI = region of interest. Inj. = injection.
325x237mm (145 x 145 DPI)
Page 31 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Table 1 Patient Demographics
Parameter No. of Patients
Included patients 25
Female 13 (52)
Age, years 37 [19-66]
Body mass index (kg/m2) 24.5 ± 4.4
Disease duration
< 2 years 10 (40)
2-10 years 6 (24)
>10 years 7 (28)
Unknown 2 (8)
Location of disease
Terminal Ileum 16 (64)
Colon 1 (4)
Ileocolon 6 (24)
Upper disease 0 (0)
Unknown 2 (8)
Medical therapy, n (%)
None 11 (44)
Corticosteroids 5 (20)
Immunomodulators 6 (24)
Biological therapy 2 (8)
Combo treatment 1 (4)
Crohn Disease Activity Index 298 ± 85
Harvey Bradshaw Index 9.9 ± 3.5
Fecal Calprotectin (μg/g)* 356 [63-3600]
C-Reactive Protein (mg/l)* 5.9 [0.7-34.4]
Hemoglobin (mmol/l) 8.6 ± 0.8
Albumin (g/l) 36.7 ± 4.5
Vitamin D (nmol/l) 65 ± 20.5
Hematocrit 0.40 ± 0.035
Time between examinations, days* 0 [0-4]
Symptoms within last flair, n (%), days*
Pain 23 (92), 157 days [11-2906]
Nausea 17 (68), 70 days [3-2495]
Vomit 11 (44), 35 days [3-265]
Diarrhea 19 (76), 303 days [3-5751]
Bloody stools 6 (24), 29.5 days [5-105]
Bloating 17 (68), 166 days [26-4093]
Weight loss 16 (64), 108 days [3-2468]
Fatigue 5 (20), 189 days [22-1764]
Note – Numbers in parenthesis are percentages. Numbers in brackets are ranges.
Unless indicated otherwise, data are means ± standard deviations.
* Median values and ranges.
Page 32 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Table 2 Magnetic resonance enterography parameters
Sequences
True FISP
T2w
Single-
shot
Turbo
Spin-Echo
T2w Single-
shot Turbo
Spin-Echo
fat sat
T1w
Turboflash
fat sat
T2w
TRUEFISP
T1w
Spoiled
3D flash
T1w
Spoiled
3d flash
Turboflash
fat sat T1w VIBE Parameter
Image plane(s) Coronal Coronal Coronal Coronal Axial Coronal Coronal Coronal +
Axial
Coronal
Field of view
(mm)
450 x 366 450 x 338 450 x 338 450 x 366 400 x 300 360 x 240 360 x 240 450 x 366 /
400 x 300
400 x 400
No. of sections 5 20 20 24 30 20 20 24 / 22 96
No. of stacks 2-3 1 1 1 3-4 1 1 1 / 3-4 1
Repetition time
(msec)
36.9 2000 2000 212 3.42 3.00 3.00 212 / 203 9:33
Echo time
(msec)
1.04 81 81 4.76 1.45 0.82 0.82 4.76 / 4.76 4:44
Acquisition
time per stack
(min)
0:19 1:20 1:20 0:40 0:12 0:05:4 3:36 0:40 / 0:36 0:20
No. of
sequential
acquisitions
10 1 1 1 1 1 120 1 1
Matrix 192 x 192 320 x 260 320 x 260 256 x 205 256 x 256 256 x 128 256 x 128 256 x 205 256 x 154
Section
thickness (mm)
10 6 6 5 4 5 5 5 / 5 2.5
Section gap
(mm)
2 2 2 2.5 0 1 1 2.5 / 0.5 0.5
Turbo factor NA 194 194 NA NA NA NA NA NA
Parallel
imaging*
GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA GRAPPA
Flip angle(s)
(degrees)
77 150§ 150
§ 70 60 5, 10, 15,
20, 25
24 70 / 70 20
*GRAPPA = Generalized autocalibrating partially parallel acquisitions, applied left to right with a factor of two in conjunction with a body matrix coil. §
Refocusing flip angles
NA = Not applicable
Page 33 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Table 3 Time intensity curve parameters, dynamic contrast enhanced magnetic resonance enterography
Value Description
Baseline Mean of initial frames before rapid rise in enhancement. First frame was discarded.
Peak Highest enhancement within first 7 frames (15 seconds). In the upslope, all preceding values
should present in an increasing manner. Only one single dip was allowed.
Rise time Time between end of Baseline and Peak
Peak enhancement Absolute value between Peak and Baseline
Slope Peak enhancement divided by Rise time
Robust Slope Best line fitted between values from 25 % to 75 % of peak enhancement
Max Slope Steepest slope over an average of 1 sec
Wash-in AUC Area under curve from Baseline to Peak – subtracted by baseline
AUC70 s Area under the curve within the first 70 seconds
Time to Peak Calculated time to Peak Enhancement value based on extrapolation of the Robust Slope.
AUC = Area under curve
Page 34 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Table 4 Repeatability of time intensity curve parameters, dynamic contrast enhanced ultrasonography (CEUS)
CEUS parameter
Region of Interest (ROI)
Mean difference
between inj 1 and inj 2 P value
Limits of
agreement
Difference from
Large ROI P value
Peak Enhancement
Large ROI 1.36 dB (0.77 to 1.96) P < .001 [-4.0 to 6.8] dB reference NA
Good QoF -0.14 dB (-0.66 to 0.38) P = .588 [-4.2 to 3.9] dB reference NA
Maximum Peak ROI 0.63 dB (0.05 to 1.20) P = .032 [-4.4 to 5.7] dB 1.34 dB (0.93 to 1.75) P < .0001
Good QoF -0.49 dB (-0.89 to -0.08) P = .018 [-3.7 to 2.7] dB 1.78 dB (1.45 to 2.10) P < .0001
Mean ROI 0.73 dB (0.17 to 1.28) P = .010 [-3.8 to 5.3] dB 0.90 dB (0.50 to 1.30) P < .0001
Good QoF 0.24 dB (-0.13 to 0.61) P = .198 [-2.3 to 2.8] dB 1.18 dB (0.87 to 1.50) P < .0001
Area under curve
Large ROI 1.46 dB (0.78 to 2.13) P < .0001 [-4.6 to 7.5] dB reference NA
Good QoF 0.46 dB (-0.03 to 0.95) P = .068 [-3.4 to 4.3] dB Reference NA
Maximum Peak ROI 0.18 dB (-0.34 to 0.71) P = .489 [-4.3 to 4.7] dB 0.88 dB (0.46 to 1.30) P < .0001
Good QoF 0.16 dB (-0.32 to 0.63) P = .515 [-3.5 to 3.8] dB 1.31 dB (0.98 to 1.64) P < .0001
Mean ROI 0.64 dB (0.15 to 1.13) P = .010 [-3.3 to 4.6] dB 0.32 dB (-0.09 to 0.73) P = .122
Good QoF 0.75 dB (0.32 to 1.17) P < .001 [-2.2 to 3.7] dB 0.79 dB (0.48 to 1.11) P < .0001
Wash-in Rate
Large ROI 1.41 dB/s (0.74 to 2.09) P < .0001 [-4.7 to 7.6] dB/s Reference NA
Good QoF -0.59 dB/s (-1.09 to -0.08) P = .023 [-4.6 to 3.4] dB/s Reference NA
Maximum Peak ROI 0.90 dB/s (0.20 to 1.59) P = .011 [-5.2 to 7.0] dB/s 1.54 dB/s (1.06 to 2.02) P < .0001
Good QoF -0.68 dB/s (-1.13 to -0.24) P = .003 [-4.2 to 2.8] dB/s 1.94 dB/s (1.61 to 2.27) P < .0001
Mean ROI 0.61 dB/s (-0.00 to 1.23) P = .051 [-4.4 to 5.6] dB/s 1.17 dB/s (0.72 to 1.62) P < .0001
Good QoF -0.16 dB/s (-0.53 to 0.21) P = .393 [-2.8 to 2.4] dB/s 1.35 dB/S (1.04 to 1.66) P < .0001
Wash-in Perfusion index
Large ROI 1.34 dB/s (0.75 to 1.93) P < .0001 [-4.0 to 6.7] dB/s Reference NA
Good QoF -0.13 dB/s (-0.64 to 0.38) P = .616 [-4.2 to 3.9] dB/s Reference NA
Maximum Peak ROI 0.57 dB/s (0.01 to 1.14) P = .045 [-4.4 to 5.5] dB/s 1.31 dB/s (0.91 to 1.72) P < .0001
Good QoF -0.50 dB/s (-0.90 to -0.10) P = .016 [-3.7 to 2.7] dB/s 1.74 dB/s (1.42 to 2.06) P < .0001
Mean ROI 0.71 dB/s (0.17 to 1.26) P = .011 [-3.8 to 5.2] dB/s 0.87 dB/s (0.48 to 1.27) P < .0001
Good QoF 0.25 dB/s (-0.12 to 0.61) P = .191 [-2.3 to 2.8] dB/s 1.16 dB/s (0.85 to 1.47) P < .0001
Note – Numbers in parenthesis are 95 % confidence intervals. Numbers in brackets are 95 % Limits of agreement.
ROI = region of interest, QoF = quality of fit, CEUS = Contrast enhanced ultrasonography, inj. = injection.
Page 35 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Table 5 CEUS region of interest quality of fit
Quality of fit Injection 1 Injection 2
Large ROI 93.4 (67-99) 97.1 (80-99)
Good QoF 97.4 (82-99) 97.4 (90-99)
Maximum peak ROI 91.9 (69-96) 94.3 (69-98)
Good QoF 92.6 (86-96) 94.2 (86-96)
Mean ROI 93.7 (82-98) 93.1 (75-97)
Good QoF 95.0 (90-98) 94.4 (86-97)
Note – Numbers are per cent, parentheses are ranges in per cent.
Page 36 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
Table 6 MR enterography reproducibility
MR enterography global
score (MEGS)
Kappa value P value
Total score ICC = 0.79 (0.59-0.90) P < .0001
Bowel wall thickness κ = 0.41 ± 0.14 P = .0016
Length of involvement κ = 0.42 ± 0.12 P = .0004
Lymph nodes κ = 0.51 ± 0.19 P = .0046
Enhancement pattern κ = 0.16 ± 0.22 P = .2313
Mural T2 signal κ = 0.51 ± 0.14 P = .1816
Perimural T2 signal κ = 0.30 ± 0.12 P = .0056
Comb sign κ = 0.39 ± 0.18 P = .0148
Fistulas κ = 0.65 ± 0.19 P = .0003
Note – Numbers in parenthesis are 95 % confidence intervals.
Unless indicated otherwise, data are means ± standard error.
ICC = intraclass correlation coefficient, κ = kappa
Page 37 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Peer Review O
nly
ABSTRACT
Purpose Cross sectional imaging methods are important for objective evaluation of small intestinal
inflammation in Crohn’s disease (CD). The primary aim was to compare relative parameters of
intestinal perfusion between contrast enhanced ultrasonography (CEUS) and dynamic contrast
enhanced magnetic resonance enterography (DCE-MRE) in CD. Furthermore, we aimed at testing the
repeatability of regions of interest (ROIs) for CEUS.
Methods This prospective study included 25 patients, 12 females (age 37, range 19-66), with
moderate to severe CD and a bowel wall thickness >3 mm, evaluated with DCE-MRE and CEUS.
CEUS bolus injection was performed twice for repeatability and analysed in VueBox®. Correlations
between modalities were described with Spearman’s rho, limits of agreement (LoA) and intraclass
correlation coefficient (ICC). ROI repeatability for CEUS was assessed.
Results The correlation between modalities was good and very good for bowel wall thickness (ICC=
0.71, P<.001) and length of the inflamed segment (ICC=0.89, P<0.001). Moderate-weak correlations
were found for the time-intensity curve parameters: peak intensity (r=0.59, P=0.006), maximum wash-
in-rate (r=0.62, P=0.004), and wash-in perfusion index (r=0.47, P=0.036). Best CEUS repeatability for
Peak Enhancement was a mean difference of 0.73 dB (95% CI: 0.17 to 1.28, P=0.01) and 95% LoA
from -3.8 to 5.3 dB. Good quality of curve fit improved LoA to -2.3 to 2.8 dB.
Conclusions The relative perfusion of small intestinal CD assessed with DCE-MRE and CEUS shows
merely a moderate correlation. Applying strict criteria for ROIs are important and allows for good
CEUS repeatability.
Page 38 of 37
Georg Thieme Verlag KG. P. O. Box 30 11 20, D-70451 Stuttgart, Germany. http://www.thieme.de/fz/ultraschall/index.html
Submitted to Editorial Office, Ultraschall in der Medizin/European Journal of Ultrasound
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
Paper II
Validity of Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced MR Enterography in the assessment of Crohn´s Disease Rune Wilkens1,2, Rikke H Hagemann-Madsen3,4, David A Peters5, Agnete H Nielsen1, Charlotte B Nørager6, Anders Tøttrup6, Henning Glerup1*, Klaus Krogh2*
1) Diagnostic Centre, Division of Medicine and Radiology, University Research Clinic for Innovative Patient Pathways, Silkeborg Regional Hospital, Silkeborg, Denmark
2) Department of Hepatology and Gastroenterology, Aarhus University Hospital, Aarhus, Denmark 3) Department of Clinical Pathology, Lillebaelt Hospital, Vejle, Denmark 4) Department of Pathology, Aarhus University Hospital, Aarhus, Denmark 5) Department of Clinical Engineering, Aarhus University Hospital, Aarhus, Denmark 6) Department of Colorectal Surgery P, Aarhus University Hospital, Aarhus, Denmark
*These authors contributed equally Short title: INTESTINAL PERFUSION IN CROHN´S DISEASE Abbreviations: CD = Crohn’s Disease MRE = Magnetic Resonance Enterography US = Ultrasonography CEUS = Contrast enhanced ultrasonography MEGS = Magnetic Resonance Enterography Global Score MaRIA = Magnetic Resonance Index of Activity AIU = Arbitrary Intensity Units SES-CD = the simplified endoscopic activity score for CD CDEIS = endoscopic index of severity
Corresponding author: Rune Wilkens
Diagnostic Centre, University Research Clinic for Innovative Patient Pathways
Silkeborg Regional Hospital
Falkevej 1-3, 8600 Silkeborg, Denmark
Phone +45 7841 5000, E-mail: [email protected]
Manuscript type: Original research
Funding:
This work was supported by AbbVie Denmark; Health Research Fund of Central Denmark Region and Danish Inflammatory Bowel Disease Association.
Abstract Background and Aims Increased small intestinal wall thickness correlates with both inflammatory activity and fibrosis in Crohn’s disease (CD). Assessment of perfusion holds promise as an objective marker distinguishing between the two conditions. Our primary aim was to determine if relative bowel wall perfusion measurements correlate with histopathological scores for inflammation or fibrosis in CD. Methods Twenty-five patients were investigated prior to elective surgery for small intestinal CD. Unenhanced Ultrasonography (US) and MR Enterography (MRE) were applied to describe bowel wall thickness. Perfusion was assessed with Contrast Enhanced US (CEUS) and Dynamic Contrast Enhanced MRE (DCE-MRE). Histopathology was used as gold standard. Results Compared to histopathology, the mean wall thickness was 0.4 mm greater on US (range -0.3 to 1.0, p=0.24) and 1.4 mm greater on MR (0.4 to 2.3, p=0.006). No correlation was found between the severity of inflammation or fibrosis on histopathology and neither DCE-MRE (r=-0.13, p=0.54 for inflammation and r=0.41, p=0.05 for fibrosis) nor CEUS (r=0.16, p=0.45 for inflammation and r=-0.28, p=0.19 for fibrosis). Wall thickness assessed with US was correlated with both histological inflammation (r=0.611, p=0.0012) and fibrosis (r=0.399, p=0.048). The same was not true for MR (r=0.41, p=0.047 for inflammation and r=0.29, p=0.16 for fibrosis)
Conclusions Bowel wall thickness assessed with US is a valid marker of small intestinal CD. However, relative contrast enhancement of US or MRE cannot distinguish between inflammatory activity and fibrosis.
Background As suggested by Peyrin-Biroulet et al., the severity of Crohn´s disease (CD) includes: impact on the patient, inflammatory burden, and complicated disease course1. Therefore, the assessment of treatment efficacy has shifted from either purely clinical endpoints or patient reported impact on quality of life towards the inclusion of objective evaluation of inflammatory burden2. It has been suggested that assessment of treatment response should be based on patient reported outcomes in combination with absence of ulcerations at endoscopy3. Unfortunately, endoscopy only allows assessment of luminal disease and it is mainly useful for evaluation of the colon and terminal ileum. CD isolated to the upper small intestine is relatively rare, but often the terminal ileum is not reached during endoscopy4.
Cross sectional imaging methods, including MRI, ultrasonography (US) or CT, are of paramount importance in the evaluation of small intestinal CD. However, no cross sectional imaging modality or single parameter is considered gold standard for the purpose5. Thus, surgical specimens are regarded as gold standard when evaluating the validity of imaging modalities for transmural assessment and complications6. Bowel wall thickness is the most consistently used parameter in cross sectional imaging, since it is well correlated with disease activity6–8. However, bowel wall thickness does not only correlate with active disease but also with chronic disease or fibrosis. Consequently, bowel wall thickness may reflect burden of disease, rather than active inflammation. The distinction is important because inflammation may respond to medical treatment while fibrosis traditionally is believed not to respond9. Active inflammation is causing vessel dilatation and neoangiogenesis resulting in increased blood flow10,11. Contrast enhancement, reflecting blood flow, measured with magnetic resonance enterography (MRE) or US has shown encouraging results as a biomarker for disease activity12–15.
For the present study we hypothesized, that the degree of perfusion in segments of the small intestine affected by CD is directly correlated with the histological grading of ongoing inflammatory activity, but inversely correlated with disease chronicity in terms of fibrosis. Thus, our primary objective was to determine the correlation between relative perfusion parameters and pathology indices of inflammatory activity and fibrosis in small intestinal specimens from patients with CD. Our secondary objectives were to determine correlations between perfusion parameters and clinical indices or biochemical markers of disease activity.
Material and methods
This GCP-monitored double blinded prospective observational study was approved by our local research ethics committee (1-10-72-339-12) and national Danish authorities (2011-005846-36). All patients gave written informed consent prior to participation. Inclusion criteria were patients with known CD scheduled for elective ileo-cecal or small intestinal resection. Additional criteria were age ≥ 18 years and detectable inflammation or small intestinal stricture on US, defined as wall thickness > 3 mm. Exclusion criteria were pregnancy, breast feeding or other contraindications for either MRE or contrast enhanced US (CEUS).
Clinical data
At the time of first US scan, participants reported their age, time since diagnosis of CD, prior abdominal surgery, and type and duration of their present symptoms. The CD activity index (CDAI) and Harvey Bradshaw index (HBI) were determined, blood samples for inflammatory parameters were drawn and stool samples were sent for assessment of f-calprotectin.
Dynamic Contrast Enhanced Ultrasonography
B-mode and CEUS were performed as previously described in details16. In brief, patients were fasting for four hours prior to scan. All scans were performed by the same experienced investigator (RW) using a Acuson S3000 (Siemens Medical Solutions, Malvern, PA) and the 9L4 probe. Large ulcers were registered if there were transverse linear hypoechoic regions across the circumferential layered bowel wall pattern or deep impression/concavity in the mucosa, Figure 1. The bowel segment with the greatest bowel wall thickness and most abundant color Doppler signal was identified and chosen for the site of CEUS, Figure 2a. US contrast with 2.4 ml Sulphur hexafluoride microbubbles (SonoVue®; Bracco Imaging, Milan, Italy), was injected as a bolus followed by 5 ml Saline flush and recorded for 90 s with low mechanical index and fixed acquisition settings16, Figure 2b. To ensure blinding to clinical data, post processing were badge analyzed more than 6 months after last US scan using VueBox® 5.1 (Bracco Suisse SA, Geneva, Switzerland) as earlier described16. Mean values of three regions of interests, all larger than 0.1 cm2 with a quality of fit 90 % or greater, were used for analysis. Motion compensation was applied whenever appropriate.
Dynamic Contrast Enhanced MR Enterography
Dynamic contrast enhanced MRE (DCE-MRE) was conducted after 4 hours fast followed by ingestion of 1 l oral contrast solution. The scanning protocol has been described in detail previously16. After initial dynamic scans, peristalsis was suppressed by i.v. injection of 20 mg Hyoscine butylbromid (Buscopan®; Boehringer Ingelheim, Ingelheim, Germany), repeated just before i.v. contrast injection of 0.2 mg / kg bodyweight gadoterate meglumine (Dotarem®; Guerbet, Villepinte, France). Patients were lying in the prone position and scanned with a 1.5 T MR scanner (Avanto; Siemens, Erlangen, Germany). The dynamic contrast enhanced sequence was 416 s long. MRE interpretation was performed by a consultant radiologist (AHN) experienced in MRE for > 9 years. Bowel wall thickness and length of involvement were registered in continues numbers while semi quantitative scoring of the most severely affected segment was done according to the MRE global score (MEGS)17. The individual components of the MRI index of activity (MaRIA) score18 were registered and the total score was calculated.
Using a home written program in MATLAB® (MathWorks®, Natick, MA), the dynamic sequences were analyzed for the most severely affected segment defined by the radiologist ((neo-)terminal ileum proximal small bowel). A region of interest (ROI) was drawn inside the bowel wall at the most enhanced part, Figure 2. The ROI was manually moved in order to stay within the bowel wall during the dynamic scan16.
Standardized tissue preparations and histopathology analyses
Immediately after surgical resection the bowel segment was rinsed in tap water. To preserve the luminal diameter and still have the opportunity to make cross sectional slices comparable to images, we did not open the bowel segment longitudinally. Fixation was made in neutral buffered formaldehyde for 48 hours using intraluminal wicks to aid fixative permeation. Well fixed specimens were transversely sectioned in 3-6 mm slices and digitally photographed before several blocks were selected representing most stenotic part, most inflamed part and other diagnostic findings. Large slices were embedded in total in macro-blocks to ensure full cross sectional coverage of the bowel. For examination of the smallest lumen and the largest wall thickness for each patient, sections of 3 μm where stained with Haematoxylin and Eosin-stain and Masson-trichrome-stain respectively. If this occurred in two different blocks, two were examined. All slides were digitalized with a NanoZoom 2.0 HT Digital Slide scanner and, and for slides from macro blocks, using a NanoZoomer 2.0-RS Digital slide scanner (both Hamamatsu, Hamamatsu City, Japan). Glasses where examined by senior GI-pathologist (RHH) and active inflammation was scored as previously described by Borley et al.19. In brief, the score semi quantitatively describes acute inflammation based on grading mucosal ulceration (0-3), edema (0-3) and depth of neutrophilic infiltration (0-4). Chronic inflammation was scored according to Chiorean et al20. Depth of collagen deposition, was scored as described by Baumgart et al.21 with increasing numbers related to the layers with increased deposition (0-5). Bowel wall thickness was digitally measured as the greatest measurement of three in the NDP.view2 Viewing software U12388-01, by RW more than 6 months from initial US scan to be effectively blinded (Hamamatsu, Hamamatsu City, Japan), Figure 2.
Statistical methods
Non-computer analyzed data were registered in web-based databases by REDCap22. Data from VueBox, REDCap and MATLAB were exported in csv-format. Statistical analysis was performed with Stata for Mac 13.1 (Stata Corp LP, College Station, TX). If no detectable disease could be found on MRE, bowel wall thickness was registered at 3 mm with a length of 0 cm. None of the linearized CEUS intensity data, expressed as arbitrary intensity units (AIU), followed a Gaussian distribution. Hence, they were log-converted as by default in US systems using 10 x log10 (AIU) and expressed in dB for further analysis16,23. Agreement between bowel wall thickness, US and MRE was described with 95 % limits of agreement (LoA) and paired t-test. Spearman’s correlation coefficient was computed between dynamic contrast enhancement and histological (primary endpoints), clinical and biochemical findings (secondary endpoints), and MRE scores of activity (post-hoc). A p-value < 0.05 was regarded as significant. The strength of our conclusions take into account, that we did not correct for multiple testing.
Results
Twenty-five patients with known CD were recruited between October 2012 and March 2014 from the colorectal surgical department of Aarhus University Hospital, a tertiary referral Centre. Twenty-three patients underwent CEUS and DCE-MRE on the same day while two had their DCE-MRE performed three and four days later. No patients experienced any severe adverse events. Elective surgery was performed with a median of seven days from the US scan (range 1-26 days), Figure 3 for eligible and included patients and Table 1 for full demographics.
Patho-anatomical data
Bowel wall thickness measured on histopathology was 8.7 ± 1.6 mm, on US 9.1 ± 2.1 mm and on MRE 10.0 ± 2.6 mm. US measurements were 0.4 (-0.3 to 1.0) mm thicker, p = 0.238 (LoA -2.7 to 3.5 mm), whereas MRE measured a mean of 1.4 (0.4 to 2.3) mm thicker bowel wall compared to histopathology, p = 0.006 (LoA -3.0 to 5.7 mm). Ulcers defined as linear, rake, confluent, or large were present in 19 patients on histopathology. Sensitivity and specificity for ulcer detection by US and MRE are shown in Table 2. Compared to patients without ulcers patients with histological proven ulcers had a thicker bowel wall; 9.2 mm (8.0 to 9.4) on histology, 9.7 mm (8.9 to 10.6) on US and 10.5 mm (9.6 to 11.4) on MRE opposed to bowel wall thickness of 7.0 mm (5.9 to 8.2, p = 0.002) on histology, 7.0 mm (5.0 to 9.0, p = 0.003) on US and 8.3 mm (4.0 to 12.7, p = 0.080) on MRE.
Relative perfusion
A total of 187 ROIs were drawn on CEUS with a median QOF of 95 % (range 65.2 to 99.6 %). Predefined good QOF > 90 was obtained in 90.9 % of ROIs. There was no statistical significant association between any CEUS or DCE-MRE based perfusion parameters and the acute inflammation index by Borley19 or the fibrosis index by Chiorean20, p > 0.38 for all p-values.
For secondary endpoints, there were weak to moderate association between CEUS intensity parameters and f-calprotectin (r = 0.49, p = 0.017), Figure 4. However, no statistical significant association was present between CEUS and CDAI, HBI or C-reactive protein.
DCE-MRE showed a moderate to weak inverse association to CDAI that was statistical significant for peak enhancement and area under curve at 25 and 70 seconds. Also, an inverse association was found for area under curve at 70 seconds and HBI. However, there was no association between DCE-MRE and biochemical markers of activity. For all results on perfusion parameters, Table 3.
Post-hoc analyses of MR enterography global score, MR index of activity and bowel wall thickness
In a post-hoc analysis, the histology index of activity were moderately associated with MEGS, (r = 0.53, p = 0.008) and with bowel wall thickness measured by US, (r = 0.61, p = 0.001) but not with MaRIA score (r= 0.17, p= 0.434). However, MaRIA and MEGS were both weakly associated with the semiquantitative histological score of collagen depth (r = 0.47 p = 0.022 and r = 0.45, p = 0.027, respectively). Bowel wall thickness also showed a moderate association with collagen depth (r = 0.53, p = 0.006).
Discussion
In the present study, unenhanced and dynamic contrast enhanced US and MRE were compared to histopathological descriptions of resections, which are generally considered the gold standard for assessment of transmural inflammation and fibrosis in CD. Our main findings do not suggest any direct correlation between relative perfusion parameters and the applied histological classification. However, there were indications towards direct correlations between f-calprotectin and CEUS and between CDAI and DCE-MRE. Additionally, we find good agreement between histology and the appreciation of bowel wall thickness and ulcers on cross-sectional imaging, especially US. Thus, increased bowel wall thickness still seems to be the single most important factor when assessing both activity and chronicity of small intestinal CD.
Although not true for all CD patients, the disease course is considered progressive from inflammation to stricture and/or penetrating disease behavior24. There is common belief that successful medical treatment is possible against active inflammation but not the chronic fibrotic disease stage25,26. Since inflammation precedes and drives fibrosis, early aggressive treatment may be beneficial27,28. This calls for non-invasive objective modalities able to map the small intestine and distinguish between the two conditions. Furthermore, objective evaluation of efficacy is important during the course of treament29. In recent years, intestinal perfusion has been suggested as a marker to reflect disease activity30–32. However, most of the prior studies used clinical markers like CDAI33 and HBI or pseudo markers of activity, like C-reactive protein, f-calprotectin34, CD34 count of mucosal biopsies35 or even compared CEUS to MRE without acknowledging that neither is considered gold-standard5. Some authors have used endoscopic indices, but only to distinguish between active and no inflammation rather than grading activity8,14,34.
We applied histology from the most severely affected part of the specimen as the gold standard for inflammation and fibrosis. Activity was measured by the score from Borley et al.19 and fibrosis by Chiorean et al.20 These scores were previously applied by Ripollés et al.36 and showed a moderate correlation with CEUS. However, theirs was a combined score including several variables including wall thickness and an arbitrary contrast intensity increase of > 46 %. The same score had an almost identical correlation with fibrosis. The likely component of this correlation could therefore be wall thickness, which we also found having a moderate correlation with both activity and fibrosis. Romanini et al.35 used a rather vague histological description and divided patients into two categories; active and inactive. She found a good correlation to several perfusion parameters. However, the analyses were performed on biopsies, rather than transmural specimens. Nylund et al.37 compared patients undergoing surgery with patients in medical treatment and defined the surgical group to be mainly fibrostenotic by excluding 10 % of patients without significant fibrosis. He found significantly lower blood volume in the surgery group.
We only recruited patients undergoing surgery and the patient group was highly heterogeneous in terms of histopathology, f-calprotectin, CDAI and perfusion parameters. Still, we are not able to show any significant correlations. Our choice of gold standard is debatable, since the acute inflammatory score was not designed for this purpose19 and has not been validated. This also accounts for the fibrosis score. It is unfortunate, that histology is considered the gold standard for transmural disease without having a solid validated score38. This paradox further impedes other pseudo-markers of activity like CEUS and DCE-MRE. The best existing activity index may thus be the simplified endoscopic activity score for CD39 (SES-CD) or CD endoscopic index of severity40 (CDEIS), with the latter being less reproducible41. These indices are, however, restricted to the
mucosa and unless double balloon endoscopy is performed, they are constrained to the large intestine and terminal ileum.
In our secondary outcomes, we found a moderate correlation between CEUS and f-calprotectin, a pseudo-marker for active inflammation. This could suggest that the chosen histological scores were inadequate. In contrast to previous studies, we found that DCE-MRE had an inverse correlation with CDAI which contradicts what is earlier shown42,43. However, CDAI is now considered less useful and sometimes unreliable to assess disease activity44. Although Giusti et al.43 found complete match between active disease based on CDAI and mucosal biopsies, six patients in our study, having CDAI < 150, underwent surgery with severely increased bowel wall thickness and thus a large element of pathological changes.
In our post-hoc analysis, existing MRE scores of CD activity showed a moderate correlation with the acute activity index for the MEGS score. Depth of collagen deposition correlated weakly with both MEGS and the MaRIA score. These scores may therefore be better at predicting inflammation or chronicity than perfusion parameters. However, solid conclusions should not be drawn from post-hoc analyses. Also, it is important to notice, that both scores include bowel wall thickness which independently correlates even better than the total scores.
This study has some limitations. First of all, the histopathological indices, considered gold standards, are semiquantitative by nature and have never been validated. Although used and slightly modified in different trials, they may not truly reflect disease activity / chronicity or susceptibility for medical treatment. As by the nature of this complex disease, we do not have a true gold standard for grading activity or fibrosis38. Therefore, our choice of gold standard was based on a pragmatic choice.
There is some selection bias in our study, since we did not include patients without inflammatory bowel changes. This is obviously owing to the fact that such patients would not undergo elective surgery. Additionally, there are several issues with the performance and reliability for dynamic perfusion in a bowel wall < 3 mm37. We aimed at grading activity rather than investigating presence of disease. For that purpose, we believe bowel wall thickness is sufficient45,46. The importance of increased bowel wall perfusion within a bowel wall of less than 3 mm is still debatable.
Due to lack of correction for the arterial input function, our methodology for measuring perfusion only allows relative or semiquantitative quantification. The impact of this is unknown, but utilizing full deconvolution for bolus injection is extremely complex47. Jirik et al.48 proposed a method to diminish this factor, which unfortunately has not yet been validated by others. The infusion-burst-replenishment technique could potentially overcome some of the limitations. Nevertheless, the method is expensive, cumbersome and has not yet been used in IBD. These methods were inconsistent with our aim, which was to validate the two existing and relatively easy-to-perform methods: CEUS and DCE-MRE. The best future application of bolus administered CEUS in CD may be in follow-up examinations on the same patient, since this will exclude between-patient variation49.
We did not apply Tofts pharmacokinetic model. However, as earlier discussed, fulfilment of the underlying assumptions for this model is also problematic16,50.
Finally, our study population is rather small. This may underestimate correlations between perfusion parameters and histological outcomes. However, our sample size was not largely different from others having
investigated the same modalities. The fact that testing existing MRE scores were post-hoc analysis will induce a risk of type I error and these results should be investigated in separate prospective studies.
In conclusion, we found no correlation between relative perfusion measurements and histological grading of inflammatory activity or fibrosis. This could be due to the chosen gold standard and the complicated nature of perfusion measurements. Bowel wall thickness remains the single parameter that correlates best with both activity and fibrosis and can be reliably estimated by MRE and especially US.
References
1. Peyrin-Biroulet L., Panés J., Sandborn WJ., Vermeire S., Danese S., Feagan BG., et al. Defining Disease Severity in Inflammatory Bowel Diseases: Current and Future Directions. Clin Gastroenterol Hepatol 2016;14(3):348–54.e17. Doi: 10.1016/j.cgh.2015.06.001.
2. Sandborn WJ., Hanauer S., Van Assche G., Panés J., Wilson S., Petersson J., et al. Treating beyond symptoms with a view to improving patient outcomes in inflammatory bowel diseases. J Crohns Colitis 2014;8(9):927–35. Doi: 10.1016/j.crohns.2014.02.021.
3. Peyrin-Biroulet L., Sandborn W., Sands BE., Reinisch W., Bemelman W., Bryant R V., et al. Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE): Determining Therapeutic Goals for Treat-to-Target. Am J Gastroenterol 2015;110(9):1324–38. Doi: 10.1038/ajg.2015.233.
4. Wilkens R., Novak KL., Lebeuf-Taylor E., Wilson SR. Impact of intestinal ultrasound on classification and management of Crohn disease patients with inconclusive colonoscopy. Can J Gastroenterol Hepatol 2015;29(X):1–7.
5. Taylor S., Mallett S., Bhatnagar G., Bloom S., Gupta A., Halligan S., et al. METRIC (MREnterography or ulTRasound in Crohn’s disease): a study protocol for a multicentre, non-randomised, single-arm, prospective comparison study of magnetic resonance enterography and small bowel ultrasound compared to a reference standard in those . BMC Gastroenterol 2014;14(1):142. Doi: 10.1186/1471-230X-14-142.
6. Zappa M., Stefanescu C., Cazals-Hatem D., Bretagnol F., Deschamps L., Attar A., et al. Which magnetic resonance imaging findings accurately evaluate inflammation in small bowel Crohn’s disease? A retrospective comparison with surgical pathologic analysis. Inflamm Bowel Dis 2011;17(4):984–93. Doi: 10.1002/ibd.21414.
7. Castiglione F., Testa A., Rea M., De Palma GD., Diaferia M., Musto D., et al. Transmural healing evaluated by bowel sonography in patients with Crohn’s disease on maintenance treatment with biologics. Inflamm Bowel Dis 2013;19(9):1928–34. Doi: 10.1097/MIB.0b013e31829053ce.
8. Moreno N., Ripollés T., Paredes JM., Ortiz I., Martínez MJ., López A., et al. Usefulness of abdominal ultrasonography in the analysis of endoscopic activity in patients with Crohn’s disease: changes following treatment with immunomodulators and/or anti-TNF antibodies. J Crohns Colitis 2014;8(9):1079–87. Doi: 10.1016/j.crohns.2014.02.008.
9. Eshuis EJ., Stokkers PC., Bemelman W a. Decision-making in ileocecal Crohn’s disease management: surgery versus pharmacotherapy. Expert Rev Gastroenterol Hepatol 2010;4(2):181–9. Doi: 10.1586/egh.10.3.
10. Darwin F. On the Primary Vascular Dilatation in Acute Inflammation. J Anat Physiol 1875;10(Pt 1):1–16.
11. Danese S., Sans M., de la Motte C., Graziani C., West G., Phillips MH., et al. Angiogenesis as a novel component of inflammatory bowel disease pathogenesis. Gastroenterology 2006;130(7):2060–73. Doi: 10.1053/j.gastro.2006.03.054.
12. Saevik F., Nylund K., Hausken T., Odegaard S., Gilja OH. Bowel Perfusion Measured with Dynamic Contrast-enhanced Ultrasound Predicts Treatment Outcome in Patients with Crohn’s Disease. Inflamm Bowel Dis 2014;20(11):2029–37. Doi: 10.1097/MIB.0000000000000159.
13. Panés J., Ricart E., Rimola J. New MRI modalities for assessment of inflammatory bowel disease. Gut 2010;59(10):1308–9. Doi: 10.1136/gut.2010.214197.
14. Ripollés T., Martínez MMJ., Paredes JMJJM., Blanc E., Flors L., Delgado F. Crohn Disease: Correlation of Findings at Contrast-enhanced US with Severity at Endoscopy. Radiology 2009;253(1):241–8. Doi: 10.1148/radiol.2531082269/-/DC1.
15. Tielbeek J a W., Ziech MLW., Li Z., Lavini C., Bipat S., Bemelman W a., et al. Evaluation of conventional, dynamic contrast enhanced and diffusion weighted MRI for quantitative Crohn’s disease assessment with histopathology of surgical specimens. Eur Radiol 2013:619–29. Doi: 10.1007/s00330-013-3015-7.
16. Wilkens R., Peters DA., Nielsen AH., Hovgaard VP., Glerup H., Krogh K. Dynamic Contrast Enhanced Magnetic Resonance Enterography and Dynamic Contrast Enhanced Ultrasonography in Crohn Disease: An observational comparison study. Ultrasound Int Open (under Rev 2016.
17. Makanyanga JC., Pendsé D., Dikaios N., Bloom S., McCartney S., Helbren E., et al. Evaluation of Crohn’s disease activity: initial validation of a magnetic resonance enterography global score (MEGS) against faecal calprotectin. Eur Radiol 2014;24(2):277–87. Doi: 10.1007/s00330-013-3010-z.
18. Rimola J., Ordás I., Rodriguez S., García-Bosch O., Aceituno M., Llach J., et al. Magnetic resonance imaging for evaluation of Crohn’s disease: validation of parameters of severity and quantitative index of activity. Inflamm Bowel Dis 2011;17(8):1759–68. Doi: 10.1002/ibd.21551.
19. Borley NR., Mortensen NJ., Jewell DP., Warren BF. The relationship between inflammatory and serosal connective tissue changes in ileal Crohn’s disease: evidence for a possible causative link. J Pathol 2000;190(2):196–202. Doi: 10.1002/(SICI)1096-9896(200002)190:2<196::AID-PATH513>3.0.CO;2-5.
20. Chiorean M V., Sandrasegaran K., Saxena R., Maglinte DD., Nakeeb A., Johnson CS. Correlation of CT enteroclysis with surgical pathology in Crohn’s disease. Am J Gastroenterol 2007;102(11):2541–50. Doi: 10.1111/j.1572-0241.2007.01537.x.
21. Baumgart DC., Müller HP., Grittner U., Metzke D., Fischer A., Guckelberger O., et al. US-based Real-time Elastography for the Detection of Fibrotic Gut Tissue in Patients with Stricturing Crohn Disease. Radiology 2015;275(3):889–99. Doi: 10.1148/radiol.14141929.
22. Harris PA., Taylor R., Thielke R., Payne J., Gonzalez N., Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42(2):377–81. Doi: 10.1016/j.jbi.2008.08.010.
23. Medellin-Kowalewski A., Wilkens R., Wilson A., Ruan J., Wilson SR. Quantitative Contrast-Enhanced Ultrasound Parameters in Crohn Disease: Their Role in Disease Activity Determination With Ultrasound. AJR Am J Roentgenol 2016;206(1):64–73. Doi: 10.2214/AJR.15.14506.
24. Pariente B., Cosnes J., Danese S., Sandborn WJ., Lewin M., Fletcher JG., et al. Development of the Crohn’s disease digestive damage score, the Lémann score. Inflamm Bowel Dis 2011;17(6):1415–22. Doi: 10.1002/ibd.21506.
25. Spinelli A., Correale C., Szabo H., Montorsi M. Intestinal fibrosis in Crohn’s disease: medical treatment or surgery? Curr Drug Targets 2010;11(2):242–8.
26. Latella G., Sferra R., Speca S., Vetuschi A., Gaudio E. Can we prevent, reduce or reverse intestinal fibrosis
in IBD? Eur Rev Med Pharmacol Sci 2013;17(10):1283–304.
27. D’Haens G., Baert F., van Assche G., Caenepeel P., Vergauwe P., Tuynman H., et al. Early combined immunosuppression or conventional management in patients with newly diagnosed Crohn’s disease: an open randomised trial. Lancet 2008;371(9613):660–7. Doi: 10.1016/S0140-6736(08)60304-9.
28. Khanna R., Bressler B., Levesque BG., Zou G., Stitt LW., Greenberg GR., et al. Early combined immunosuppression for the management of Crohn’s disease (REACT): a cluster randomised controlled trial. Lancet (London, England) 2015;386(10006):1825–34. Doi: 10.1016/S0140-6736(15)00068-9.
29. Bouguen G., Levesque BG., Feagan BG., Kavanaugh A., Peyrin-Biroulet L., Colombel J-F., et al. Treat to target: a proposed new paradigm for the management of Crohn’s disease. Clin Gastroenterol Hepatol 2015;13(6):1042–50.e2. Doi: 10.1016/j.cgh.2013.09.006.
30. Horjus Talabur Horje CS., Bruijnen R., Roovers L., Groenen MJM., Joosten FBM., Wahab PJ. Contrast Enhanced Abdominal Ultrasound in the Assessment of Ileal Inflammation in Crohn’s Disease: A Comparison with MR Enterography. PLoS One 2015;10(8):e0136105. Doi: 10.1371/journal.pone.0136105.
31. Ma X., Li Y., Jia H., Zhang J., Wang G., Liu X., et al. Contrast-Enhanced Ultrasound in the Diagnosis of Patients Suspected of Having Active Crohn’s Disease: Meta-analysis. Ultrasound Med Biol 2015;41(3):659–68. Doi: 10.1016/j.ultrasmedbio.2014.09.013.
32. Murphy DJ., Smyth AE., McEvoy SH., Gibson DJ., Doherty GA., Malone DE. Subclassification of small bowel Crohn’s disease using magnetic resonance enterography: A review using evidence-based medicine methodology. Clin Radiol 2015;70(12):1336–43. Doi: 10.1016/j.crad.2015.07.008.
33. Quaia E., De Paoli L., Stocca T., Cabibbo B., Casagrande F., Cova MA. The value of small bowel wall contrast enhancement after sulfur hexafluoride-filled microbubble injection to differentiate inflammatory from fibrotic strictures in patients with Crohn’s disease. Ultrasound Med Biol 2012;38(8):1324–32. Doi: 10.1016/j.ultrasmedbio.2012.04.008.
34. De Franco A., Veronica A Di., Armuzzi A., Roberto I., Marzo M., De Pascalis B., et al. Ileal Crohn Disease: Mural Microvascularity Quantified with Contrast-enhanced US Correlates with Disesase Activity. Radiology 2012;262(2):680–8.
35. Romanini L., Passamonti M., Navarria M., Lanzarotto F., Villanacci V., Grazioli L., et al. Quantitative analysis of contrast-enhanced ultrasonography of the bowel wall can predict disease activity in inflammatory bowel disease. Eur J Radiol 2014;83(8):1317–23. Doi: 10.1016/j.ejrad.2014.05.012.
36. Ripollés T., Rausell N., Paredes JM., Grau E., Martínez MJ., Vizuete J. Effectiveness of contrast-enhanced ultrasound for characterisation of intestinal inflammation in Crohn’s disease: A comparison with surgical histopathology analysis. J Crohns Colitis 2013;7(2):120–8. Doi: 10.1016/j.crohns.2012.03.002.
37. Nylund K., Jirik R., Mezl M., Leh S., Hausken T., Pfeffer F., et al. Quantitative Contrast-Enhanced Ultrasound Comparison Between Inflammatory and Fibrotic Lesions in Patients With Crohn’s Disease. Ultrasound Med Biol 2013;39(7):1197–206. Doi: 10.1016/j.ultrasmedbio.2013.01.020.
38. Rieder F., de Bruyn JR., Pham BT., Katsanos K., Annese V., Higgins PDR., et al. Results of the 4th Scientific Workshop of the ECCO (Group II): Markers of intestinal fibrosis in inflammatory bowel disease. J Crohn’s Colitis 2014;8(10):1166–78. Doi: 10.1016/j.crohns.2014.03.009.
39. Daperno M., D’Haens G., Van Assche G., Baert F., Bulois P., Maunoury V., et al. Development and validation of a new, simplified endoscopic activity score for Crohn’s disease: the SES-CD. Gastrointest Endosc 2004;60(4):505–12. Doi: 10.1016/S0016-5107(04)01878-4.
40. Mary JY., Modigliani R. Development and validation of an endoscopic index of the severity for Crohn’s disease: a prospective multicentre study. Groupe d'Etudes Therapeutiques des Affections Inflammatoires du Tube Digestif (GETAID). Gut 1989;30(7):983–9. Doi: 10.1136/gut.30.7.983.
41. Khanna R., Zou G., D’Haens G., Rutgeerts P., McDonald JWD., Daperno M., et al. Reliability among central readers in the evaluation of endoscopic findings from patients with Crohn’s disease. Gut 2015:1–7. Doi: 10.1136/gutjnl-2014-308973.
42. Florie J., Wasser MNJM., Arts-Cieslik K., Akkerman EM., Siersema PD., Stoker J. Dynamic contrast-enhanced MRI of the bowel wall for assessment of disease activity in Crohn’s disease. AJR Am J Roentgenol 2006;186(5):1384–92. Doi: 10.2214/AJR.04.1454.
43. Giusti S., Faggioni L., Neri E., Fruzzetti E., Nardini L., Marchi S., et al. Dynamic MRI of the small bowel: usefulness of quantitative contrast-enhancement parameters and time-signal intensity curves for differentiating between active and inactive Crohn’s disease. Abdom Imaging 2010;35(6):646–53. Doi: 10.1007/s00261-010-9624-6.
44. Peyrin-Biroulet L., Reinisch W., Colombel J-F., Mantzaris GJ., Kornbluth A., Diamond R., et al. Clinical disease activity, C-reactive protein normalisation and mucosal healing in Crohn’s disease in the SONIC trial. Gut 2014;63(1):88–95. Doi: 10.1136/gutjnl-2013-304984.
45. Calabrese E., Maaser C., Zorzi F., Kannengiesser K., Hanauer SB., Bruining DH., et al. Bowel Ultrasonography in the Management of Crohnʼs Disease. A Review with Recommendations of an International Panel of Experts. Inflamm Bowel Dis 2016;0(0):1. Doi: 10.1097/MIB.0000000000000706.
46. Horsthuis K., Bipat S., Bennink RJ., Stoker J. Inflammatory Bowel Disease Diagnosed with US, MR, Scintigraphy, and CT: Meta-analysis of Prospective Studies. Radiology 2008;247(1):64–79. Doi: 10.1148/radiol.2471070611.
47. Gauthier M., Tabarout F., Leguerney I., Polrot M., Pitre S., Peronneau P., et al. Assessment of quantitative perfusion parameters by dynamic contrast-enhanced sonography using a deconvolution method: an in vitro and in vivo study. J Ultrasound Med 2012;31(4):595–608.
48. Jirik R., Nylund K., Gilja OH., Mezl M., Harabis V., Kolar R., et al. Ultrasound perfusion analysis combining bolus-tracking and burst-replenishment. IEEE Trans Ultrason Ferroelectr Freq Control 2013;60(2):310–9. Doi: 10.1109/TUFFC.2013.2567.
49. Quaia E., Sozzi M., Angileri R., Gennari AG., Cova MA. Time-Intensity Curves Obtained after Microbubble Injection Can Be Used to Differentiate Responders from Nonresponders among Patients with Clinically Active Crohn Disease after 6 Weeks of Pharmacologic Treatment. Radiology 2016;000(0):152461. Doi: 10.1148/radiol.2016152461.
50. Sourbron SP., Buckley DL. On the scope and interpretation of the Tofts models for DCE-MRI. Magn Reson Med 2011;66(3):735–45. Doi: 10.1002/mrm.22861.
Table 1 Demographics
Parameter No. of Patients Included patients 25 Female 17 (68) Age, years 37 [19-66] Body mass index (kg/m2) 24.8 [20.2-32] Disease duration 9.4 [0-25.3]
< 2 years 6 (24) 2-10 years 8 (32) >10 years 11 (44)
Location of disease L1, Terminal Ileum 17 (68) L2, Colon 0 (0) L3, Ileocolon 6 (24) L4, Upper disease 2 (8)
Crohn Disease Activity Index 202 ± 94 Harvey Bradshaw Index 7.5 ± 4.4 Fecal Calprotectin (μg/g)* 340 [30-3600] C-Reactive Protein (mg/l)* 4.3 [0.5-29.4] Hemoglobin (mmol/l) 8.6 ± 0.9 Albumin (g/l) 34.9 ± 4.8 Vitamin D (nmol/l) 63 ± 21.7 Hematocrit 0.42 ± 0.039 Time between examinations, days* 0 [0-4] Time between US and Surgery, days* 7 [1-26] Symptoms within last flair, n (%), days*
Pain 21 (84), 180 days [9-2495] Nausea 18 (72), 130 days [8-2495] Vomit 10 (40), 32.5 days [8-193] Diarrhea 15 (60), 358 days [9-5751] Bloody stools 7 (28), 105 days [11-3173] Bloating 17 (68), 159 days [9-4093] Weight loss 11 (44), 193 days [47-4452] Fatigue 4 (16), 242.5 days [32-1764]
IBD Treatment, n (%) None 8 (32) Steroids 6 (24) Immunosuppressors 5 (20) Biological therapy 5 (20) Combo treatment 1 (4)
Note – Numbers in parenthesis are percentages. Numbers in brackets are ranges. Unless indicated otherwise, data are means ± standard deviations. * Median values and ranges. US = Ultrasonography
Table 2. Accuracy for ulcer detection on cross sectional imaging. Sensitivity Specificity Accuracy PPV NPV P-value Ulcers on US 18/19 (94.7 %)
[74.0 - 99.9 %] 4/6 (66.7 %) [22.3 – 95.7 %]
80.7 % [59.4 – 100 %]
90 % [68.3 – 98.8 %]
80 % [28.4 – 99.5 %]
P = 0.005
Ulcers on MRE 15/18 (83.3 %) [58.6 – 96.4 %]
2/6 (33.3 %) [4.3- 77.7 %]
58.3 % [35.9 – 80.8 %]
78.9 % [54.4 – 93.9 %]
40 % [5.3 – 85.3 %]
P = 0.366
Difference P = 0.625 P = 0.5 P = 0.2188 US = Ultrasonography; MRE = Magnetic Resonance Enterography; PPV = positive predictive value; NPV = negative predictive value Nominator is True positive or true negative and denominator is total ulcers on histology Brackets are 95 % confidence interval.
Table 3 Correlation between relative perfusion and histological, biochemical and clinical markers of activity
DCEUS parameter Acute inflammation Fibrosis f-calprotectin CRP CDAI HBI P-value
Peak enhancement NS NS 0.48 NS NS NS 0.021§* Total area under curve NS NS 0.49 NS NS NS 0.017§* Wash-in area under curve NS NS 0.46 NS NS NS 0.029§* Wash-out area under curve NS NS 0.51 NS NS NS 0.013§* Wash-in perfusion index NS NS 0.48 NS NS NS 0.021§* Wash-in Rate NS NS 0.44 NS NS NS 0.034§* Rise time NS NS -0.02 NS NS NS 0.932§ Fall time NS NS 0.13 NS NS NS 0.547§ Mean transit time local NS NS 0.39 NS NS NS 0.068§ DCE-MRE parameters Rise time NS NS NS NS 0.11 NS 0.629# Peak enhancement NS NS NS NS -0.46 NS 0.026#* Area under curve washin NS NS NS NS -0.18 NS 0.406# Area under curve 25 s NS NS NS NS -0.49 NS 0.019# Area under curve 70 s NS NS NS NS -0.56 -0.43 0.006#** / 0.040† Initial slope NS NS NS NS -0.41 NS 0.052# Maximum slope NS NS NS NS -0.30 NS 0.157# Wash-in perfusion index NS NS NS NS -0.36 NS 0.089#
§P value for Calprotectin; #P value for CDAI; †P value for HBI; * P < 0.05, ** P < 0.01; DCEUS = Dynamic Contrast Enhanced Ultrasonography; DCE-MRE = Dynamic Contrast Enhance Magnetic Resonance Enterography; CRP = C-reactive protein; HBI = Harvey Bradshaw Index; CDAI = Crohn’s Disease Activity Index; NS = Not significant.
Figure 1 – Ulcer on ultrasonography
Left: B-mode image with color Doppler Imaging, transverse scan plane on severely thickened small intestinal loop. Arrows indicate an ulcer.
Right: B-mode image in the longitudinal scan plane. Arrows shows a vertical low echogenic disturbance in wall layering of the posterior bowel wall indicating an ulcer. In the anterior bowel wall, a broad impression is seen, indicated by arrow heads.
Figure 2
Top left: Ultrasonography b-mode image of neo-terminal ileum with clear bowel wall thickening, measured to 9.3 mm. Slight mucosal impressions are seen and partially disturbed wall layering.
Top mid: Corresponding dual image from VueBox analysis with contrast enhanced ultrasonography on the left side and b-mode on the right side. Turquois box delineate the analysis and motion compensation area, Green region of interest (ROI) is the largest ROI possible. White, yellow and purple ROIs are all > 0.1 cm2 with a quality of fit > 90 %. The heat map shows peak enhancement intensity with dark red as the highest value and turquois / blue the lowest.
Right: MR enterography, coronal T 2 HASTE fat sat sequence showing neo-terminal ileum with edema and a thickened bowel segment, measured to 9.4 mm in the right lower quadrant.
Bottom left: Scanned histopathological hole section of the worst segment stained Masson-Trichrome and digitalized for morphometric measurements. Clearly thickened bowel walls with three measurements of bowel wall thickness. The greatest measurement is 9.15 mm in NDP.view2. Deep ulcers, disturbed submucosa with lymphocyte infiltration and fat deposition. Smooth muscle cells (red) and fibrosis (blue) spiculates from the muscular layer into the adventitia and prominent hyperplasia of the sub-serosal fatty tissue (often described as “creeping fat”) is seen.
Bottom mid: Time intensity curve from DCE-MRE analysis. Blue and turquois lines represent slope of late enhancement.
Figure 3 Flowchart
Flow chart of patients finally included in the study.
Figure 4 Scatterplot and best fitted line for correlation between contrast enhanced ultrasonography area under curve and fecal calprotectin.
Paper III
Which cross-sectional imaging parameters predict stiffness of strictures in Crohn’s Disease?
Rune Wilkens1,2, Dong-Hua Liao3, Hans Gregersen4, Henning Glerup1, Agnete H Nielsen1, David A Peters5,
Charlotte Buchard6, Anders Tøttrup6, Klaus Krogh2.
1) Diagnostic Centre, Divisions of Medicine and Radiology, University Research Clinic for Innovative
Patient Pathways, Silkeborg Regional Hospital, Denmark
2) Department of Hepatology and Gastroenterology, Aarhus University Hospital, Denmark
3) GIOME Academy, Department of Clinical Medicine, Aarhus University, Denmark
4) GIOME and Key Laboratory for Biorheological Science and Technology of Ministry of Education;
Bioengineering College of Chongqing University, Chongqing, 400044, China
5) Department of Clinical Engineering, Aarhus University Hospital, Denmark
6) Department of Surgery, Aarhus University Hospital, Denmark
Corresponding author:
Rune Wilkens
Diagnostic Centre, University Research Clinic for Innovative Patient Pathways
Silkeborg Regional Hospital
Falkevej 1-3, 8600 Silkeborg, Denmark
Phone +45 7841 5000, E-mail: [email protected]
Manuscript type: Original research
ABSTRACT
Objectives The primary aim was to explore which parameters obtained by preoperatively cross-sectional
imaging of small bowel Crohn’s Disease (CD) strictures are associated with the biomechanical properties of the
strictures.
Methods Patients with proven CD that underwent elective surgery for small bowel disease were preoperatively
examined with dynamic contrast enhanced MR enterography (DCE-MRE) and contrast enhanced
ultrasonography (CEUS) in the fasting state. Small intestinal specimens were examined in a bath containing
Krebs solution and distended with the EndoFLIP probe. Luminal and outer wall cross sectional areas were
measured with ultrasonography (US) for calculation of stiffness by young’s modulus (E). Statistical analysis was
performed using Spearman’s correlation and Students t-test.
Results Stiffness of strictures was associated with prestenotic dilatation on US (r = 0.60, P = 0.009) but not on
MRE (r = 0.10, P = 0.704). Obvious strictures identified by US were stiffer E: 19.0 kPa (14.0 to 25.8) than those
with no or uncertain strictures E: 12.6 kPa (10.5 to 15.1), P = 0.026. The same was not true on MRE E: 15.4 kPa
(12.8 to 18.6) for certain strictures and 16.6 kPa (8.9 to 31.0) for uncertain/no stricture, P = 0.726. MR
enterography global score (MEGS) was associated with E (r = 0.66, P = 0.003). Of the dynamic contrast
enhanced parameters, stiffness was only associated with the initial slope of enhancement on DCE-MRE (r =
0.718, P = 0.001) and with none of the CEUS parameters.
Conclusions Prestenotic dilatation on US is an objective marker of the stiffness of small intestinal strictures in
CD. Increasing severity defined by MEGS and initial slope of enhancement on DCE-MRE are also associated with
stiffness.
Background
The natural course of Crohn’s Disease (CD), usually evolving from active inflammation to stricturing and
penetrating disease, is well known and has been described earlier1,2. Active inflammation is mainly treated with
medical therapy3 aiming at mucosal healing and normalization of bowel habits without pain4. There is no
evidence for successful medical treatment of fibrotic strictures5,6. Consequently, fibrotic strictures and
penetrating disease are considered the main indications for small intestinal surgery in CD.
Pre-operative classification and detection of strictures in CD are relatively good for all cross sectional
modalities7. From imaging and endoscopy, strictures are defined as localized, persistent narrowing of the
lumen8. Narrowing of the lumen combined with upstream dilatation characterize a functional stricture9.
Unfortunately, no consensus seems to exist on the classification of strictures in clinical trials10 and objective
methods for evaluating properties of strictures in CD are warranted. In lack of a strict definition, strictures are
also classified as either symptomatic or asymptomatic. A classification based purely on symptomatology is
highly subjective and depends, among other factors, on food composition and intake11,12.
Persistent inefficient medical treatment can lead to corticosteroid dependency and prolongation of severe
abdominal pain13. On the other hand, early and aggressive resection of an inflammatory segment may lead to
early relapse and increased risk of short bowel syndrome14. Based on conventional imaging, sub-classification
of strictures into those with active inflammation, mixed components of inflammation, or fibrotic is still
difficult15 and new techniques should be developed. In a contrast enhanced ultrasonography (CEUS) study by
Ripollés et al.16 the authors showed a significantly higher contrast enhancement of inflammatory areas
compared with fibrotic areas. Rimola et al. showed that the percentage of enhancement gain after seven
minutes associated with the degree of fibrosis17. Both Magnetic Resonance Enterography (MRE) and
Ultrasonography (US) have a high sensitivity and specificity for detecting strictures based on structural
findings18. The studies mentioned above suggest that addition of dynamic contrast enhancement may provide a
valid tool for the differentiation between inflammatory and fibrotic strictures.
Mechanical properties of the gastrointestinal tract are important for normal function. Unfortunately, our
knowledge about the mechanical properties of small intestinal strictures and the ability to predict these
without surgery is very limited. It has been shown, that longitudinal tensiometry strain (stretch) of strictures is
associated with US Elastography and fibrosis19. Johnson et al. showed that extracellular matrix stiffness is an
independent driver of fibroblast acitivation20 and Tsamis et al. computed the theoretical stress and strain in
strictureplasty surgery21.
Impedance planimetry measures pressure-cross-sectional area (P-CSA) during distension of non-sphincteric
regions of the gastrointestinal (GI) tract22. The data can be used for computation of simple mechanical
parameters such as tension-strain relations23. The Functional Luminal Imaging Probe (FLIP) provides a
geometric profile of the distended site from multiple CSA recordings, often in relation to studies of the
esophagogastric junction or the anal sphincter24–26. The FLIP technology is commercially available (EndoFLIP®,
Crospon Ltd., Galway, Ireland) but has to the best of our knowledge not been used to study the geometric
profile and stiffness of pathological strictures. Furthermore, it does not measure the wall thickness which is an
important parameter in stiffness computations22.
The primary aim of the present study was to explore which parameters obtained by preoperatively CEUS and
Dynamic Contrast Enhanced MRE (DCE-MRE) of small bowel CD were associated with the biomechanical
properties of the strictures. Our main hypothesis was that strictures with low perfusion on CEUS or DCE-MRE
would be less compliant than those with high perfusion, as the former would be mainly fibrotic while the latter
would be inflammatory16. Furthermore, we hypothesized that the degree of prestenotic dilatation is associated
with the stiffness of the stenosis.
Materials and Methods
The present study was a GCP-monitored double-blinded prospective observational study approved the Danish
research ethics committee (1-10-72-339-12) and Danish national authorities (2011-005846-36). All patients
gave written informed consent prior to participation. Patients were 18 years or older and had small bowel
inflammation or stricture on US27. Inclusion flowchart and full demographics are shown in Figure 1 and Table 1.
Data on associations between imaging methods, perfusion and histology from the same group of patients will
be presented elsewhere27. At time of inclusion, patients scored their symptoms subjectively into four
categories; no symptoms, mild symptoms, moderate symptoms, and severe symptoms.
Cross-sectional imaging
All preoperative imaging procedures were previously described in detail27,28. In brief, bowel ultrasonography
was done using an Acuson S3000™ with a 9L4 linear matrix probe (Siemens Medical Solutions, Malvern, PA)
after the patient fasted for four hours. CEUS was continuously recorded for 90 seconds after a dual injection of
2.4 ml sulphur hexafluoride (SonoVue®; Bracco Imaging, Milan, Italy). Time intensity curves were analysed in
VueBox® 5.1 (Bracco Suisse SA, Geneva, Switzerland). Mean values of three ROIs from each of the two
injections were utilized for comparison with biomechanical parameters.
MRE is performed routinely in our department using 4 hours fast and 1 L of oral Mannitol solution. Patients lie
in the prone position throughout the scan. Prior to scanning non-dynamic sequences 20 mg of Hyoscine
butylbromid (Buscopan®; Boehringer Ingelheim, Ingelheim, Germany) were injected i.v. and repeated
immediately before dynamic contrast sequences. The full MRE protocol has been described in detail28. In
addition to DCE-MRE analysis, components of the MR enterography global assessment (MEGS)29 and MR index
of activity (MaRIA)30 were assessed and the full scores calculated.
For both imaging modalities prestenotic dilatation is graded as: mild (more than stricture but no more than
adjacent loops), moderate (more than adjacent loops), severe (> 25 mm). Strictures were regarded as
“certain”, for moderate and severe prestenotic dilatation.
Functional luminal imaging and in vitro ultrasonography
Immediately after surgical excision, the bowel specimen was opened at the stapled ends, rinsed in lukewarm
water and placed in a water tank containing 37 °C buffering Krebs Solution infused with 100 % O2 1 L/min. The
Krebs solution was comprised of 118 mmol/L NaCl, 4.7 mmol/L KCl, 25 mmol/L NaHCO3, 1.0 mmol/L NaH2PO4,
1.2 mmol/L MgSO4, 2.5 mmol/L CaCl2-H2O, 11 mmol/L Glucose, and 0.11 mmol/L ascorbic acid31. The water
tank was preheated and constantly kept at 37 °C using an electric warming tray (Bartscher, Salzkotten,
Germany). An EndoFLIP probe was inserted from the anal side of the bowel segment and the bag was slightly
distended to determine the most stenotic part of the specimen. After deflation and relocation to ensure
optimal positioning, the bowel and probe were fixated in both ends (Figure 2). At this stenotic site, a tiny
acupuncture needle was inserted in the mesentery or serosa in the axial direction of the bowel. The needle,
easily identified on ultrasound, ensured the same location of ultrasound still images. Two times two additional
needles were inserted one and two centimetres on each side of the first needle to allow US still images exactly
at the five locations. The EndoFLIP probe was then connected to a water column system filled with the
EndoFLIP solution delivered by Crospon Ltd. The level container was aligned with the EndoFLIP probe and
enough time to reach a steady-state calibration was allowed before pressure was reset to the zero pressure
point. A HI VISION Preirus US machine with a EUP-L73S probe (Hitachi Medical Corporation, Tokyo, Japan) was
used to obtain b-mode still images at each site of needle placement and annotated with 0, +1, +2 or -1, -2 and
pressure level. After five measurements were obtained, the water column was elevated 10 cm and the next
images were obtained after equilibrium of pressure was obtained. The procedure was repeated with 10 cm
increments until 100 cm H2O.
Data Analysis
US images were analysed using OsiriX 5.7.1. 64 bit (Pixmeo SARL, Bernex, Switzerland). A circular/ovoid region
of interest were drawn for the luminal and the serosal border of the bowel wall respectively. ROIs were named
according to the picture annotations and ROI data were exported in .csv files and then imported into MATLAB®
(MathWorks®, Natick, MA) for further analysis.
The bowel segment was assumed to be composed of five circular cylinders. The ultrasound images were
analysed for: circumference (C) and luminal areas (LA) of inner and outer surfaces at the five points along the
stenosis. At each location, the circumferential Green strain (εθ) and circumferential Kirchhoff stress (Sθ ) for
each distension pressure were calculated as32:
Circumferential Green strain:
)1(21 2
−= θθ λε (1)
Circumferential Kirchhoff stress:
hrPS inner
.2θ
θλ⋅∆
= (2)
where)()(
00outerinner
outerinner
CCCC
++
=θλ is circumferential stretch ratio, and Cinner, Couter, are circumference for inner and
outer surface at pressurize state, 00 , outerinner CC are circumference at pressure 0 cmH2O, P∆ is the transmural
pressure difference, π/innerinner LAr = is the inner radius calculated from measured luminal area of the inner
surface (LAinner), π/outerouter LAr = is the outer radius calculated from the outer surface (LAouter) and
innerouter rrh −= is wall thickness. The calculated strain- stress curves was fitted to the exponential function as:
)1( −= θαεθ β eS (3)
where α and β are material constants and α relates to the wall stiffness. The circumferential Young's modulus (E) was thus calculated as:
θ
θ
εddSE = (4)
In this paper, E for 2.5 < Sθ < 5 kPa was calculated, under the assumption of linearity.
Statistical analysis
Descriptive statistics were computed for all demographic variables. These included means, standard deviations
and range for continues variables and frequencies and percentages for categorical factors. Several parameters
did not follow a Gaussian distribution. Young’s Modulus was log-converted for further analysis and CEUS
intensity parameters (expressed as arbitrary intensity units (AIU)) were converted to dB using the equation 10 x
log10 (AIU) = dB. Spearman’s correlation was computed between E and CEUS or DCE-MRE parameters.
Spearman’s correlation was also computed for associations between E and MEGS, MaRIA and categorical
variables including the degree of prestenotic dilatation, MRE mural edema and enhancement pattern.
Categorical variables were also examined with one-way analysis of variance before investigation of pairwise
difference. Finally, difference in E was tested with Students t-test for dichotomous variables like presence of
strictures and ulcers.
Results
Of the 25 patients included in the original study undergoing elective small bowel surgery, EndoFLIP data were
available for 18 (Figure 1).
Stiffness of the stricture and Ultrasonography or MR enterography
Young’s modulus for the stricture were moderately associated with the grading of prestenotic dilatation on US,
r = 0.60, P = 0.009, but not on MRE, r = -0.10, P = 0.704, (Figure 3). Strictures classified as “certain” on US were
stiffer, E: 19.0 kPa (14.0 to 25.8) than those classified as “no” or “uncertain” stricture, E: 12.6 kPa (10.5 to
15.1), P = 0.026. No association was found between certainty of strictures on MRE and E: 15.4 kPa (12.8 to
18.6) for certain strictures and 16.6 (8.9 to 31.0) for uncertain/no stricture, P = 0.726, Figure 4. Patients with
large ulcers on US had a trend towards also having a stiffer bowel wall, E: 17.5 kPa (13.6 to 22.6) than those
with no ulcers, E: 12.1 kPa (9.6 to 15.3), P = 0.082. Normal (layered) versus disturbed bowel wall pattern on US
was not associated with E: 17.2 kPa (12.0 to 24.7) and E: 14.2 kPa (12.0 to 18.8) respectively, P = 0.291. In
contrast, increasing levels of bowel edema on MRE were associated with E of the stricture, r = 0.527, P = 0.025.
Compound MRE scores were without agreement. No correlation was found between E and MaRIA, r = 0.222, P
= 0.376, on the other hand a moderate association existed between MEGS and E, r = 0.66, P = 0.003 (Figure 5).
Stiffness of the stricture and Dynamic Contrast Enhanced MR Enterography or Ultrasonography
The enhancement pattern of DCE-MRE showed a weak but statistically significant association to E of the
stricture (r = 0.484, P = 0.042). A moderate correlation was found between E and DCE-MRE Initial slope of
enhancement, r = 0.718, P = 0.001, (Figure 6). No significant correlation was found between E and any of the
CEUS parameters. All correlation parameters are listed in Table 2
Stiffness of the stenosis and abdominal pain
Most patients 14/18 (78 %) reported severe pain. There was no correlation between stiffness of the stenosis
and the degree of self-reported pain, r = -0.08, P = 0.75, (Figure 7).
Discussion
Our main aim was to explore which imaging parameters would be associated with increasing wall stiffness of
small intestinal strictures in Crohn´s disease. Such associations would support the clinical use of specific
imaging methods in the evaluation of CD and, potentially, support the planning of the best treatment strategy.
The bowel wall stiffness was well associated with prestenotic dilatation on US but not on MRE. In contrast,
both DCE-MRE initial slope of enhancement and MEGS were moderately associated with stiffness of the
stricture while CEUS did not. This is opposite of our hypothesis, since we were expecting an inverse correlation
between relative perfusion and stricture stiffness. Our hypothesis may have been too simple, though. In the
present study, we found that the stiffness of small intestinal strictures in CD varied largely between patients
even though their clinical presentation made them all candidates for surgery. The classification of intestinal
strictures should probably be more complex than just inflammatory, compound and fibrotic16 as the co-
existence of inflammation and fibrosis may occur in most patients33. In cases where severe fibrosis is present
with concomitant high degree of inflammation and corresponding high perfusion, a direct correlation between
perfusion and stiffness may occur. Thus, the term ”fibrostenosis” may often be imprecise thereby restricting
correct classification of patients and possibly resulting in poor treatment outcomes34.
It is well known, that clinical symptoms correlate poorly to objective signs of mucosal inflammation in CD35.
Likewise, our data showed no correlation between abdominal pain and stiffness of the stricture. This may not
be surprising as pain is a subjective parameter and, furthermore, the number of patients in our study was
relatively small. However, having a diagnostic tool that can predict a stiff stricture may allow for better
selection of patients undergoing endoscopic balloon dilatation, strictureplasty or even resection. This may be
especially important since cross sectional imaging methods are not always capable of reliably differentiation
between inflammation and fibrosis36.
Stidham et al.37 utilized strain Elastography to predict bowel wall elasticity in rat colitis model and seven CD
patients undergoing surgery for intestinal strictures. Elasticity reference standard was tissue stiffness expressed
as Young’s modulus kPa measured by a micro-elastometer. They were able to find a clear difference between
normal and fibrotic tissue. The micro-elastometer measures the tissue displacement vs. force under
compression38,39. Fraquelli et al.40 also investigated CD strictures with Elastography. However, they used a
quantitative tissue analysis of fibrotic content as gold standard rather than biomechanical properties.
Nevertheless, they were able to find a direct association with ultrasound Elastography and fibrotic content in
the strictures. Another study by Havre et al. also evaluated strain Elastography on human intestinal specimens,
and found no association with increasing histological inflammation. Furthermore, the authors conclude that
reproducibility was only fair and strain Elastography could not differentiate between CD and
adenocarcinoma41. A study by Baumgart et al. compared strain Elastography with tissue tensiometry by
applying a 250 g weight and the percentage of stretch. However, the authors did not specify when the tissue
tensiometry was performed. In a study by Massalou et al.42 longitudinal stress of the normal colonic wall of
cadavers was tested until complete rupture showing the outer bowel layer including the serosa and external
muscular membranes to initially breach after 30 to 40 % elongation.
All of the above mentioned studies either investigated tissue Elastography or longitudinal stress-strain
relationship for a section of bowel. None have so far investigated the more important circumferential stress of
the bowel wall, and particularly not in CD strictures. Circumferential stiffness is more relevant for any
pathophysiological changes than longitudinal stiffness. Classification of a stenosis on US had a significantly
higher E than inflammation without a stricture or uncertain stricture on US, whereas this was not the case for
MRE. Although oral contrast consumption prior to US scan has shown to be superior to non-oral intake43, we
believe that any prestenotic dilatation in the fasting patient is a clear sign of a stricture, whereas the amount of
prestenotic dilatation in enterography can be more difficult to determine since the bowel is already distended.
Both bowel wall oedema and pattern of contrast enhancement on DCE-MRE were associated with stiffness.
Since both variables have been proven to correlate with active inflammation, it is likely, that both inflammation
and chronic changes are causing increasing stiffness, and not only increased collagen deposition36,44.
Our study has some limitations. Although we tested the specimen in Krebs solution at 37 °C, this is not an in
vivo study. Without the normal hormonal and neurogenic feedback mechanisms, we investigated the
mechanical properties only. Active contraction of smooth muscle cells within a stricture will increase wall
stiffness. Our study could theoretically have been performed intra-operatively. However, this would increase
duration of surgery and have the significant drawback of elevated risk of contamination, since the bowel
segment would have to be opened during the procedure. Although the EndoFLIP probe is regarded reliable for
CSA measurements, we experienced some issues with pressure measurements and a couple of the included
specimens were only investigated until 70 cmH2O, whereas three others had to be fully excluded due to
technical failures, which may likely be due to the fact that we bypassed the original system.
In spite of meticulous efforts, we were unable to ensure the exact same location investigated by
impedance planimetry and the pre operative dynamic contrast enhanced modalities, since the exact distance
from the ileocecal valve may vary from pre operative till intraoperative and post operative. However, we
investigated the same segment, and generally the site with the thickest bowel wall also caused the most
abundant stricture. Our aim was to investigate cross sectional imaging’s ability to predict the biomechanical
properties of the bowel without any further prerequisite information about the exact stricture location.
In the analysis of both DCE-MRE and CEUS, we did not take the arterial input function into account45. This
may limit the inter-patient correlation significantly27. Although Jirik et al.46 proposes a way to overcome this,
reproducibility in humans has not yet been proved nor consensus for applying this method is obtained.
Although prestenotic dilatation and bowel wall pattern are objective findings, we applied several
semiquantitative assessments, which was not tested for reproducibility in this study.
Finally, we did not include normal tissue or histology grading in our study as reference standard. A normal
bowel can easily distend to more than 3 cm in diameter, which is the upper limit for the EndoFLIP probe.
Existing histopathology grading has not been validated and has a semiquantitative design. Since elasticity of the
normal bowel may be affected by multiple layers in multiple steps42, a quantitative analysis based on individual
histological layers, like the outer and inner muscular layers and the abundance of collagen may be more useful
in future studies.
In conclusion, we describe the diversity of a simple ex-vivo passive biomechanical property of resected CD
strictures. Circumferential stiffness correlates with increasing prestenotic dilatation acknowledged on pre-
surgical US without oral contrast agents. There is also a direct correlation with objective measurements like the
initial slope of enhancement and the MEGS of severity. No correlation was found with sensation of abdominal
pain.
References
1. Pariente B, Cosnes J, Danese S, et al. Development of the Crohn’s disease digestive damage score, the Lémann score. Inflamm Bowel Dis. 2011;17(6):1415-1422. doi:10.1002/ibd.21506.
2. Cosnes J, Cattan S, Blain A, et al. Long-term evolution of disease behavior of Crohn’s disease. Inflamm Bowel Dis. 2002;8(4):244-250. doi:10.1097/00054725-200207000-00002.
3. Dignass a, Van Assche G, Lindsay JO, et al. The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Current management. J Crohns Colitis. 2010;4(1):28-62. doi:10.1016/j.crohns.2009.12.002.
4. Peyrin-Biroulet L, Sandborn W, Sands BE, et al. Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE): Determining Therapeutic Goals for Treat-to-Target. Am J Gastroenterol. 2015;110(9):1324-1338. doi:10.1038/ajg.2015.233.
5. Bharadwaj S, Fleshner P, Shen B. Therapeutic Armamentarium for Stricturing Crohnʼs Disease. Inflamm Bowel Dis. 2015;21(9):2194-2213. doi:10.1097/MIB.0000000000000403.
6. Latella G, Sferra R, Speca S, Vetuschi A, Gaudio E. Can we prevent, reduce or reverse intestinal fibrosis in IBD? Eur Rev Med Pharmacol Sci. 2013;17(10):1283-1304.
7. Panés J, Bouzas R, Chaparro M, et al. Systematic review: the use of ultrasonography, computed tomography and magnetic resonance imaging for the diagnosis, assessment of activity and abdominal complications of Crohn’s disease. Aliment Pharmacol Ther. 2011;34(2):125-145. doi:10.1111/j.1365-2036.2011.04710.x.
8. Van Assche G, Dignass A, Reinisch W, et al. The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Special situations. J Crohns Colitis. 2010;4(1):63-101. doi:10.1016/j.crohns.2009.09.009.
9. Van Assche G, Dignass A, Panes J, et al. The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Definitions and diagnosis. J Crohns Colitis. 2010;4(1):7-27. doi:10.1016/j.crohns.2009.12.003.
10. Rieder F, Zimmermann EM, Remzi FH, Sandborn WJ. Crohn’s disease complicated by strictures: a systematic review. Gut. 2013;62(7):1072-1084. doi:10.1136/gutjnl-2012-304353.
11. Tanaka M, Kawakami A, Iwao Y, Fukushima T, Yamamoto-Mitani N. Coping Strategies for Possible Flare-Ups and Their Perceived Effectiveness in Patients With Inflammatory Bowel Disease. Gastroenterol Nurs. 39(1):42-47. doi:10.1097/SGA.0000000000000201.
12. Charlebois A, Rosenfeld G, Bressler B. The Impact of Dietary Interventions on the Symptoms of Inflammatory Bowel Disease: A Systematic Review. Crit Rev Food Sci Nutr. 2015;8398(April). doi:10.1080/10408398.2012.760515.
13. Alós R. Timing of surgery in Crohn’s disease: A key issue in the management. World J Gastroenterol. 2008;14(36):5532. doi:10.3748/wjg.14.5532.
14. Poggioli G, Pierangeli F, Laureti S, Ugolini F. Review article: indication and type of surgery in Crohn’s disease. Aliment Pharmacol Ther. 2002;16 Suppl 4:59-64. doi:10.1046/j.1365-2036.16.s4.9.x.
15. Lenze F, Wessling J, Bremer J, et al. Detection and differentiation of inflammatory versus fibromatous Crohn’s disease strictures: prospective comparison of 18F-FDG-PET/CT, MR-enteroclysis, and transabdominal ultrasound versus endoscopic/histologic evaluation. Inflamm Bowel Dis. 2012;18(12):2252-2260. doi:10.1002/ibd.22930.
16. Ripollés T, Rausell N, Paredes JM, Grau E, Martínez MJ, Vizuete J. Effectiveness of contrast-enhanced ultrasound for characterisation of intestinal inflammation in Crohn’s disease: A comparison with surgical histopathology analysis. J Crohns Colitis. 2013;7(2):120-128. doi:10.1016/j.crohns.2012.03.002.
17. Rimola J, Planell N, Rodríguez S, et al. Characterization of Inflammation and Fibrosis in Crohn’s Disease Lesions by Magnetic Resonance Imaging. Am J Gastroenterol. 2015;110(October 2014):432-441. doi:10.1038/ajg.2014.424.
18. Panés J, Bouzas R, Chaparro M, et al. Systematic review: the use of ultrasonography, computed tomography and magnetic resonance imaging for the diagnosis, assessment of activity and abdominal complications of Crohn’s disease. Aliment Pharmacol Ther. 2011;34(2):125-145. doi:10.1111/j.1365-2036.2011.04710.x.
19. Baumgart DC, Müller HP, Grittner U, et al. US-based Real-time Elastography for the Detection of Fibrotic Gut Tissue in Patients with Stricturing Crohn Disease. Radiology. 2015;275(3):889-899. doi:10.1148/radiol.14141929.
20. Johnson L a, Rodansky ES, Sauder KL, et al. Matrix stiffness corresponding to strictured bowel induces a fibrogenic response in human colonic fibroblasts. Inflamm Bowel Dis. 2013;19(5):891-903. doi:10.1097/MIB.0b013e3182813297.
21. Tsamis A, Pocivavsek L, Vorp DA. Elasticity and geometry: a computational model of the Heineke-Mikulicz strictureplasty. Biomech Model Mechanobiol. 2014;13(6):1185-1198. doi:10.1007/s10237-014-0565-9.
22. Zhao J, Jørgensen CS, Liao D, Gregersen H. Dimensions and circumferential stress-strain relation in the porcine esophagus in vitro determined by combined impedance planimetry and high-frequency ultrasound. Dig Dis Sci. 2007;52(5):1338-1344. doi:10.1007/s10620-006-9238-6.
23. Liao D, Frøkjaer JB, Yang J, et al. Three-dimensional surface model analysis in the gastrointestinal tract. World J Gastroenterol. 2006;12(18):2870-2875.
24. Lottrup C, Gregersen H, Liao D, et al. Functional lumen imaging of the gastrointestinal tract. J Gastroenterol. 2015;50(10):1005-1016. doi:10.1007/s00535-015-1087-7.
25. Fynne L, Luft F, Gregersen H, et al. Distensibility of the anal canal in patients with systemic sclerosis: a study with the functional lumen imaging probe. Colorectal Dis. 2013;15(1):e40-e47. doi:10.1111/codi.12063.
26. McMahon BP, Rao SSC, Gregersen H, et al. Distensibility testing of the esophagus. Ann N Y Acad Sci. 2011;1232:331-340. doi:10.1111/j.1749-6632.2011.06069.x.
27. Wilkens R, Hagemann-Madsen RH, Peters DA, et al. Validity of Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced MR Enterography in the assessment of Crohn’s Disease. Under Constr.
28. Wilkens R, Peters DA, Nielsen AH, Hovgaard VP, Glerup H, Krogh K. Dynamic Contrast Enhanced
Magnetic Resonance Enterography and Dynamic Contrast Enhanced Ultrasonography in Crohn Disease: An observational comparison study. Ultrasound Int Open (under Rev. 2016.
29. Makanyanga JC, Pendsé D, Dikaios N, et al. Evaluation of Crohn’s disease activity: initial validation of a magnetic resonance enterography global score (MEGS) against faecal calprotectin. Eur Radiol. 2014;24(2):277-287. doi:10.1007/s00330-013-3010-z.
30. Rimola J, Ordás I, Rodriguez S, et al. Magnetic resonance imaging for evaluation of Crohn’s disease: validation of parameters of severity and quantitative index of activity. Inflamm Bowel Dis. 2011;17(8):1759-1768. doi:10.1002/ibd.21551.
31. Bailey LE, Ong SD. Krebs-Henseleit solution as a physiological buffer in perfused and superfused preparations. J Pharmacol Methods. 1978;1(2):171-175. doi:10.1016/0160-5402(78)90022-0.
32. Dou Y, Gregersen S, Zhao J, Zhuang F, Gregersen H. Effect of re-feeding after starvation on biomechanical properties in rat small intestine. Med Eng Phys. 2001;23(8):557-566. http://www.ncbi.nlm.nih.gov/pubmed/11719078.
33. Zappa M, Stefanescu C, Cazals-Hatem D, et al. Which magnetic resonance imaging findings accurately evaluate inflammation in small bowel Crohn’s disease? A retrospective comparison with surgical pathologic analysis. Inflamm Bowel Dis. 2011;17(4):984-993. doi:10.1002/ibd.21414.
34. Rieder F, Lawrance IC, Leite A, Sans M. Predictors of fibrostenotic Crohn’s disease. Inflamm Bowel Dis. 2011;17(9):2000-2007. doi:10.1002/ibd.21627.
35. Peyrin-Biroulet L, Reinisch W, Colombel J-F, et al. Clinical disease activity, C-reactive protein normalisation and mucosal healing in Crohn’s disease in the SONIC trial. Gut. 2014;63(1):88-95. doi:10.1136/gutjnl-2013-304984.
36. Makanyanga J, Punwani S, Taylor S a. Assessment of wall inflammation and fibrosis in Crohn’s disease: value of T1-weighted gadolinium-enhanced MR imaging. Abdom Imaging. 2012;37(6):933-943. doi:10.1007/s00261-011-9821-y.
37. Stidham RW, Xu J, Johnson L a, et al. Ultrasound elasticity imaging for detecting intestinal fibrosis and inflammation in rats and humans with Crohn’s disease. Gastroenterology. 2011;141(3):819-826.e1. doi:10.1053/j.gastro.2011.07.027.
38. Labout JWA. The Micro-Elastometer A micro-device for measuring the modulus of “stiffness” of small samples of thermoplastic materials. Rheol Acta. 1958;1(2-3):186-189. doi:10.1007/BF01968864.
39. Kim K, Johnson LAL a, Jia C, et al. Noninvasive ultrasound elasticity imaging (UEI) of Crohn’s disease: animal model. Ultrasound Med Biol. 2008;34(6):902-912. doi:10.1016/j.ultrasmedbio.2007.11.020.
40. Fraquelli M, Branchi F, Cribiù FM, et al. The Role of Ultrasound Elasticity Imaging in Predicting Ileal Fibrosis in Crohn’s Disease Patients. Inflamm Bowel Dis. 2015;21(11):2605-2612. doi:10.1097/MIB.0000000000000536.
41. Havre RF, Leh S, Gilja OH, et al. Strain assessment in surgically resected inflammatory and neoplastic bowel lesions. Ultraschall der Medizin. 2014;35(2):149-158. doi:10.1055/s-0032-1325535.
42. Massalou D, Bège T, Masson C, et al. Influence of loading speed on the mechanical properties of the
colon. Comput Methods Biomech Biomed Engin. 2013;16 Suppl 1(April 2016):189-190. doi:10.1080/10255842.2013.815876.
43. Pallotta N, Vincoli G, Montesani C, et al. Small intestine contrast ultrasonography (SICUS) for the detection of small bowel complications in crohn’s disease: a prospective comparative study versus intraoperative findings. Inflamm Bowel Dis. 2012;18(1):74-84. doi:10.1002/ibd.21678.
44. Rimola J, Rodriguez S, García-Bosch O, et al. Magnetic resonance for assessment of disease activity and severity in ileocolonic Crohn’s disease. Gut. 2009;58(8):1113-1120. doi:10.1136/gut.2008.167957.
45. Gauthier M, Tabarout F, Leguerney I, et al. Assessment of quantitative perfusion parameters by dynamic contrast-enhanced sonography using a deconvolution method: an in vitro and in vivo study. J Ultrasound Med. 2012;31(4):595-608.
46. Jirik R, Nylund K, Gilja OH, et al. Ultrasound perfusion analysis combining bolus-tracking and burst-replenishment. IEEE Trans Ultrason Ferroelectr Freq Control. 2013;60(2):310-319. doi:10.1109/TUFFC.2013.2567.
Table 1 Demographics
Parameter No. of Patients Included patients 25 Female 17 (68) Age, years 37 [19-66] Body mass index (kg/m2) 24.8 [20.2-32] Disease duration 9.4 [0-25.3]
< 2 years 6 (24) 2-10 years 8 (32) >10 years 11 (44)
Location of disease L1, Terminal Ileum 17 (68) L2, Colon 0 (0) L3, Ileocolon 6 (24) L4, Upper disease 2 (8)
Crohn Disease Activity Index 202 ± 94 Harvey Bradshaw Index 7.5 ± 4.4 Fecal Calprotectin (μg/g)* 340 [30-3600] C-Reactive Protein (mg/l)* 4.3 [0.5-29.4] Hemoglobin (mmol/l) 8.6 ± 0.9 Albumin (g/l) 34.9 ± 4.8 Vitamin D (nmol/l) 63 ± 21.7 Hematocrit 0.42 ± 0.039 Time between examinations, days* 0 [0-4] Time between US and Surgery, days* 7 [1-26] Symptoms within last flair, n (%), days*
Pain 21 (84), 180 days [9-2495] Nausea 18 (72), 130 days [8-2495] Vomit 10 (40), 32.5 days [8-193] Diarrhea 15 (60), 358 days [9-5751] Bloody stools 7 (28), 105 days [11-3173] Bloating 17 (68), 159 days [9-4093] Weight loss 11 (44), 193 days [47-4452] Fatigue 4 (16), 242.5 days [32-1764]
IBD Treatment, n (%) None 8 (32) Steroids 6 (24) Immunosuppressors 5 (20) Biological therapy 5 (20) Combo treatment 1 (4)
Note – Numbers in parenthesis are percentages. Numbers in brackets are ranges. Unless indicated otherwise, data are means ± standard deviations. * Median values and ranges. US = Ultrasonography
Table 2 Correlation between Young’s modulus and imaging
Parameter Spearman’s rho P value CEUS Peak Enhancement 0.06 P = .815 Area Under Curve -0.05 P = .837 Wash-in Rate 0.18 P = .492 Wash-in Perfusion index 0.06 P = .815 Rise Time -0.37 P = .147 Fall Time -0.25 P = .343 DCE-MRE Peak Enhancement 0.11 P = .667 Area Under Curve 70 s -0.21 P = .422 Initial Slope 0.72 P = .001 Initial Slope Max 0.41 P = .103 Wash-in Perfusion Index -0.12 P = .660 Rise Time -0.49 P = .048 Wash-out Slope 60 s 0.07 P = .794 Wash-out Slope 120 s -0.14 P = .586 Symptoms Pain -0.08 P = .750 / P = .845§ Enhancement pattern (DCE-MRE) 0.48 P = .042 / P = .370§
Note – § P value for One-way analysis of variance. CEUS = Contrast enhanced ultrasonography, DCE-MRE = Dynamic Contrast Enhanced MR Enterography.
Figure 1: Flow chart
Figure 2: Experimental set-up.
Water container with oxygenated Krebs solution and the specimen stretched to original length using magnetic clamps (top left). The water container placed on a table heater (37 °C) with ruler and bag for application of pressure (top right). Acupuncture needles inserted as location marker for ultrasound (bottom left). Ultrasound probe during scanning the bowel segment (bottom right).
Figure 3. Stiffness and degree of prestenotic dilatation. Associations between stiffness of the stricture and the degree of prestenotic dilatation on ultrasonography (left) and MRE (right). Oneway ANOVA Ultrasonography, P = 0.021 and MR Enterography, P = 0.276. Boxes are inter-quartile ranges and medians. * P < 0.05
Figure 4. Associations between stiffness and certainty of stricture on ultrasonography (left) and MR enterography (right). Boxes are inter-quartile ranges and medians. * P < 0.05
Figure 5. Association between stiffness of the stricture and MR enterography global score (MEGS).
Figure 6. Stiffness of the stricture is associated with the initial slope of Dynamic Contrast Enhanced MR Enterography (DCE-MRE) (right) but not with wash-in rate of Contrast Enhanced Ultrasonography (CEUS) (left).
Figure 7. Stiffness according to pre-surgery maximum pain level. Stiffness is not correlated with pain, Oneway ANOVA, P = 0.845.
AARHUS
UNIVERSITY
Declaration of co-authorship
Full name of the PhD student: Rune Thordahl Wilkens
This declaration concerns the foliowing article/manuscript:
Title: Validity of Contrast Enhanced Ultrasonography and Dynamic Contrast Enhanced
Authors: Rune Wilkens, Rikke H Hagemann-Madsen, David A Peters, Agnete H Nielsen, <!, Beooing..Glew,p,...ll.J..al.ll:i.AJ:�1-------------l
The article/manuscript is: Published D Accepted D Submitted D In preparation �
If published, state ful) reference:
If accepted or submitted, statejournal:
Has the article/manuscript previously been used in other PhD or doctoral dissertations?
No 181 Yes D If yes, give details:
The PhD student has contributed to the elements of this article/manuscript as follows:
A. No or little contributionB. Has contributed (10-30 %)C. Has contributed considerably (40-60 %)D. Has done most of the work (70-90 %)E. Has essentially done all the work
Signatures of the co-authors
Date Name
3o - \ l, David A Peters
Agnete H Nielsen
�/S · /6 Rikke H Hagemann-Madsen
1 of 1
':Ju/s · l l?
30-05-16
30.05.16
黝鲆 黢髁鼍§w
Date∶ 30· m旬 2° ls
∷∷ ∷∷
∫
∷0∷ ∫ 。∷矿
`∷
In-,ase;f fui;ther co-authors please attach appendix
Sienature ofthe PhD student
J扯 V祗 Klaus Klogh 掰 溺 撼 鼬Char'lotte Buchard
s卜 Qf-″ Hans Gregersen 辊 饧
''
2of1
31-05-16
