Disclosures Diffusion and Perfusion Imaging in the Head ...€¦ · MRI ² diffusion weighted...

12
Ashok Srinivasan, MD Associate Professor Director of Neuroradiology University of Michigan Health System Diffusion and Perfusion Imaging in the Head and Neck Disclosures No relevant financial disclosures Learning objectives Review technical aspects and challenges of diffusion and perfusion imaging Discuss clinical applications in the head and neck 43 y/o M status post parotidectomy ??? CT Perfusion MRI – diffusion weighted imaging MRI – Perfusion imaging Imaging of diffusion of water molecules Strong gradients in 3 orthogonal directions ADC (Apparent diffusion coefficient) Expressed in mm 2 /s Higher the ADC, more is the degree of motion (and vice versa)

Transcript of Disclosures Diffusion and Perfusion Imaging in the Head ...€¦ · MRI ² diffusion weighted...

Ashok Srinivasan, MD

Associate Professor

Director of Neuroradiology

University of Michigan Health System

Diffusion and Perfusion Imaging

in the Head and Neck

Disclosures

No relevant financial disclosures

Learning objectives

Review technical aspects and challenges of diffusion and perfusion imaging

Discuss clinical applications in the head and neck

43 y/o M status post parotidectomy

???

CT Perfusion

MRI – diffusion weighted imaging

MRI – Perfusion imaging

Imaging of diffusion of water molecules

Strong gradients in 3 orthogonal directions

ADC (Apparent diffusion coefficient)

Expressed in mm2/s

Higher the ADC, more is the degree of

motion (and vice versa)

ECHO-PLANAR IMAGING

• Single shot- Limited spatial resolution, greater geometric distortion, shorter acquisition

• Multi shot- Higher resolution, reduced geometric distortion, longer acquisition

ECHO-PLANAR• AXIAL plane• 4 mm slice thickness• 0.5 mm intersection gap• AP phase encoding direction

NON ECHO-PLANAR

HASTE

PROPELLER

BLADE

• Less susceptibility artifact

• Thinner sections, higher imaging matrices

• Longer acquisition

• Lower signal to noise ratio

Challenges with DWI neck

Susceptibility artifacts

• Parallel imaging

• Minimize echo train length (duration of recording)

Geometric distortions (especially at root of neck)

Read out segmented EPI, Reduced FOV DWI

Less cellularity (Benign)

HIGHER ADC

?Surrogate marker for cellularity

More cellularity (Malignant)

LOWER ADC

CLINICAL APPLICATIONS

Benign vs. malignant lesions

Pleomorphic adenoma

T2-weighted Pre-Gad T1-weighted Post Gad T1-weighted

b800

diffusionADC map

ADC = 1.5 x 10-3 mm2/s

Adenocarcinoma

ADC = 0.7 x 10-3 mm2/s

Post-Gad T1-weighted

b1000

diffusion

ADC map

Basal cell adenocarcinoma

Post Gad

T1-weighted

b1000

diffusionADC map

19 y/o female with LE tingling and numbness

Venous vascular malformation

Tagged RBC scan

High grade acinic cell CA parotid with left

level II metastatic node

T2-weighted Post Gad T1W b1000 diffusion

Benign vs. malignant lesions

ADC

0.5 0.75 1.0 1.25 1.5

MAL BEN

ADC value of 1.3 x 10–3

mm2

/s at 3T –

? threshold value for differentiation

65 y/o M with stridor and laryngeal mass

65 y/o M with stridor and laryngeal massb1000

diffusionADC map

High ADC-

Chronic

inflammation

Mucoepidermoid carcinoma

T2-weighted Post Gad

T1-weighted

b1000

diffusionADC map

Post Gad

T1-weightedb1000

diffusionADC map

Chondrosarcoma Paraganglioma

T2W Post Gad

T1-weighted

b1000

diffusionADC map

CLINICAL APPLICATIONS

Benign vs. malignant lesions

Post-therapy changes vs. recurrence

T1-w Post-Gad T1-w

T2-w Post-Gad T1-w

Post- Rx

enhancing

mass

Patient

1

Patient

2

POST-THERAPY

Patient

1

Patient

2

Post-GadT1-w DWI ADC map

POST-THERAPY

Patient

1

Patient

2

ADC = 0.8 x 10-3 mm2/s

Biopsy proven recurrence

ADC = 1.8 x 10-3 mm2/s

Biopsy: Benign granulation

ADC map

POST-THERAPY

ADC can be helpful in differentiating

residual or recurrent tumor from

post therapy changes

Residual or recurrent tumor ADC

Post-therapy changes ADC

ADC of 1.30 x 10–3 mm2/s as a threshold

POST-THERAPY 82 y/o F –

H/o Large cell Lymphoma and neutropenia

High ADC-

Likely

benign

b1000

diffusion ADC map

High ADC-

No malignancy on

path

Two patients- both with new mass in

the surgical bed post parotidectomy

Restricted diffusion

(low ADC)-

Path proven

recurrence

Pt.1

Pt.2

T2-weighted

Post Gad

T1-weighted

b1000

diffusionADC map

43 y/o M status post parotidectomy

High ADC- No malignant cells on path

CLINICAL APPLICATIONS

Benign vs. malignant lesions

Post-therapy changes vs. recurrence

Prediction of therapy

Monitoring of response

PRE-TREATMENT ADC

COMPLETE responders PARTIAL responders

1.04 x 10-3 mm2/s 1.35 x 10-3 mm2/s

• Significant in ADC in complete responders within 1 week of treatment

• Prediction of response

Pre-Rx ADC and change in ADC

PREDICTION OF RESPONSE

ADC = 0.8 x10–3

ADC = 1.4 x10–3

Pt.1

Pt.2

• ADC - Monitoring early therapeutic response

• Increase in ADC during early phase of

therapy relative to baseline value suggesting conversion of solid tumor to necrotic tissue

INTRA-THERAPY

Intra therapy tumor response

assessment

Rx

b1000

diffusion

ADC map

CLINICAL APPLICATIONS

Benign vs. malignant lesions

Post-therapy changes vs. recurrence

Prediction of therapy

Monitoring of response

Recurrent cholesteatoma

Recurrent Cholesteatoma + Scar

b1000

diffusionADC map

Post mastoidectomy for cholesteatoma

Recurrent Cholesteatoma + Scar

b1000

diffusionADC map

Post-Op for petrous apex cholesteatoma

CLINICAL APPLICATIONS

Benign vs. malignant lesions

Post-therapy changes vs. recurrence

Prediction of therapy

Monitoring of response

Miscellaneous

Recurrent cholesteatoma

Intraocular

abscess

Melanoma,

Evaluate brain

metastases

Extracranial and

brain metastases

present

MRI – diffusion weighted imaging

MRI – Perfusion imaging

• With or without contrast administration

• T2*- Dynamic susceptibility contrast or

T1 - Dynamic contrast enhanced

• Blood flow and blood volume, capillary permeability and transfer coefficients

Axial plane

• 2D-multislice, T1-weighted fast-field echo

• 3D spoiled Gradient T1-weighted

• Slice thickness of 5-6 mm (3 mm overlap)

• Craniocaudal length of coverage – 5 to 6 cm

• 0.1 mmol/kg of Gadolinium at 5 mL/sec

Then, 20 mL saline flush at 5 mL/sec

MR Perfusion – DCE Technique

• Patient motion

• Skull base susceptibility artifacts

• Good cardiovascular function

• Good renal function

• Values are relative, so semi-quantitative

MR Perfusion - Challenges

Benign vs. malignant

MRI – Perfusion imaging

Razek AA et al. Eur J Radiol. 2011;77:73-79. Razek AA et al. JCAT 2011;35:21-25.

DSC % threshold value

Could be helpful for differentiating

malignant from benign nodes and

metastatic from lymphomatous nodes

Mean DSC% of malignant tumor >>

Mean DSC% of benign lesions

Ve WIR

Benign vs. malignant

Prediction of outcome

MRI – Perfusion imaging

MRI scans before therapy and 2 weeks into chemo-RT

Blood volume in the primary tumor after 2 weeks of chemo-RT was increased significantly in the local control patients compared with the local failure patients (p < 0.03)

Reduction in tumor volume after 2 weeks of chemo-RT did not predict local control

Prediction of outcome using DCE MRI

Cao Y et al. Int J Radiat Oncol Biol Phys. 2008;72:1287-90.

Pre-Tx During Tx Post-Tx

(Courtesy: Srinivasan et al. Biologic imaging of head and neck cancer. AJNR 2012;33:586-94)

Blood volume Blood volume Blood volume

Kim S et al. AJNR 2010;31:262-8.

Pre-Rx

HIGH

K(TRANS)

Pre-Rx

LOW

K (TRANS)

COMPLETE RESPONSE

PARTIAL RESPONSE

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Benign vs. malignant

Recurrent tumor

MRI – Perfusion imaging

Prediction of outcome

Recurrent tumor versus

Post-therapy benign changes

DCE MR perfusion

Post-therapy benign changes showed

TTP & RWO (relative washout ratio)

than recurrent tumor

Furukawa et al. Head Neck 2013;35:923-929.

Recurrent squamous cell carcinoma

WIR TTP

CT Perfusion

MRI – diffusion weighted imaging

MRI – Perfusion imaging

CT Perfusion

• Requires contrast administration and dynamic imaging

• Post-processing using deconvolutionalgorithm

• Blood volume, blood flow, mean transit time, time to peak, capillary permeability

Cine mode

40 mm detector coverage

5 mm slice thickness

50 cc Non-ionic contrast at 4 cc/s

20 cc saline flush at 4 cc/s

5 sec delay

50 sec scan duration

CT Perfusion - Technique

• Patient motion

• Z axis coverage limited

• Streak from dental amalgam

• Good renal function

CT Perfusion - Challenges

Pre-Rx

• Higher BF and BV correlate with better response

• BV correlates with microvascular density, an important prognostic indicator

• CP correlates with EGFR overexpression

Post Rx

• Responders show lower BV and BF

Ash L et al. Radiology 2009;251:422-428.

Zima A et al. AJNR 2007;28:328-334.

Surlan-Popovic K et al. AJNR 2010;31:570-575.

Jin G et al. J Comput Assist Tomogr 2011;35:26-30.

CT Perfusion

CT PERFUSION

Post treatment

Increased blood volume and blood flow – recurrent tumor

Blood volume Blood flow

Thrombosed vein CT PERFUSION

Post treatment

Increased blood volume and permeability– residual tumor

Blood volume Permeability

Summary

Pre-Rx During Rx Post Rx

Low ADC- Malignant

Low ADC

Elevated blood volume/K-Trans

Good prognosis

Increasing ADC

Transient increase in BV

Good response

Low ADC- Malignant

Elevated BV or Ve

Recurrent tumor

Always interpret diffusion/perfusion imaging along with anatomic information

ALLERGIC FUNGAL SINUSITIS