1

DEPARTMENT ABSTRACTS – 2005 - 2006

I. Vasculitis pages 2 – 14

II. Microcrystalline arthritis 15-17

III. Scleroderma and Raynaud's vasospasm 18

IV. Epidemiology and Outcome 19-21

V. Psoriatic arthritis 22-23

VI. Metabolic Bone Disease 24-30

2

I. Vasculitis

Wegener granulomatosis: customized treatment using cyclophosphamide and

methotrexate.

A 12 year single-practice experience.

Alexandra Villa-Forte, Tiffany M. Clark, Marcelo Gomes, John Carey, Edward Mascha,

Matthew T. Karafa, Susana Arrigain, Gerald Roberson and Gary S. Hoffman

Purpose: To assess outcomes of therapy for Wegener's granulomatosis using customized

treatment strategies that limits or completely avoids the use of cyclophosphamide.

Methods: Patients were included if they primarily received care within Cleveland Clinic

Center for Vasculitis Care and Research and met American College of Rheumatology criteria

for Wegener's granulomatosis. A primary goal was to assess methotrexate utility in limiting or

avoiding cyclophosphamide use. Patients in whom treatment primarily included cytotoxic

agents other than methotrexate and/or cyclophosphamide were excluded. In the presence of

severe disease, remission-induction was attempted with use of cyclophosphamide. For mild to

moderate disease, methotrexate was the initial treatment for all patients in whom serum

creatinine was less than 2mg/dl. Following initial improvement of severe disease after

cyclophosphamide use, patients were treated with methotrexate if serum creatinine was less

than 2mg/dl. Disease activity was measured based on the Birmingham Vasculitis Activity

Score, as modified for Wegener's granulomatosis.

Results: Eighty-three patients from among 250 referred to Center for Vasculitis Care and

Research with Wegener's granulomatosis met eligibility criteria. Seventy percent of patients

initially had severe disease and received a short course of cyclophosphamide for remission-

induction. In over half of these patients, illness was judged to be severe because of pulmonary

hemorrhage, rapidly progressive glomerulonephritis, including need for dialysis, or neurologic

abnormalities. Sixty-three percent of patients with severe illness presented with serum

creatinine values that were > 2.0 mg/dl. The two disease severity groups did not differ in time

required to achieve remission. All patients improved, remission was achieved in 50% within 6

months and 72% within 12 months. Sustained remission (> 6months) was ultimately achieved

in 77% of all patients and this end point did not differ between severity subsets. Among the 76

patients who achieved remission of any duration, 45% relapsed within 1 year and 66% relapsed

within 2 years. There were no differences in relapse rates or relapse disease activity scores

between groups. Eighty-two percent of relapsed patients achieved subsequent remissions after

additional treatment. About ¾ of relapses were mild and promptly responded to treatment.

Eighty-six percent of relapsed patients suffered disease exacerbations after methotrexate

therapy was withdrawn or in the process of being discontinued. None of the patients developed

CYC-induced cystitis or bladder cancer. Only 3.7% of patients died over a median period of

4.5 years. No deaths were due to active disease.

Conclusions: In patients with Wegener's granulomatosis, a strategy that limits or avoids

cyclophosphamide therapy has produced excellent outcomes as judged by survival and

infrequent long-term cyclophosphamide toxicity. Frequent relapses have been mild in the great

3

majority of patients. This may be related to close clinical and laboratory monitoring of disease

activity and treatment-related adverse events. While this approach has diminished mortality,

relapses still lead to incremental permanent morbidity in most patients.

4

Temporal artery gene expression profiles may predict conditions necessary for

developing giant cell arteritis. Rula Hajj-Ali, Thomas Hamilton, Michael Lee, Jianbo Li, Jose Hernandez-Rodriguez, Maria

C. Cid, Ana García-Martínez, Earl Brown, Richard Prayson, Gary S. Hoffman

Purpose: Pathogenesis of giant cell arteritis (GCA) remains incompletely understood. It is

uncertain whether abnormalities that occur during the aging process of the vessel wall and/or

the immune system play an important role in establishing disease vulnerability. We sought to

identify gene expression changes that occur in the vessel wall before overt inflammation that

may lead to GCA.

Methods: Gene expression profiling using RNA obtained from thirteen temporal artery

biopsies (5 from patients with GCA and 8 from controls) was carried out using Affymetrix

HGU133 Plus 2 GeneChips. Patients with GCA met modified 1990 ACR criteria for GCA.

Samples from patients with unequivocal GCA in which temporal artery biopsies showed no

evidence of histologic disease were compared with biopsies from patients who did not have

GCA. The data sets were analyzed using the random forests algorithm, a Classification and

Regression Tree based procedure. Gene probes exhibiting significantly increased or diminished

expression in samples from patients with GCA as compared to controls were ranked according

to the magnitude of difference and the top 50 probes were further examined for their biological

relevance.

Results: Genes showing markedly elevated expression in temporal artery biopsies from GCA

patients included those involved in transcriptional regulation (ZNF709, ZNF568, CRSP9,

ZNF131, ZNF585A) and apoptosis/anti-apoptosis (BNIP1, PAFAH2). In contrast, genes

exhibiting reduced expression included metallopeptidases (ADAMTSL3), DNA damage

response (GTSE1), anti-inflammatory activity (IL18BP) and fibroblast growth factor receptor

(FGFR1).

Conclusion: These findings suggest that the gene expression profile in temporal arteries from

GCA patients exhibits changes that precede overt inflammation and vessel injury. These

changes may set the stage in vessel substrate to “invite” the injury that is later recognized as

GCA.

5

Takayasu’s arteritis: guarded outcomes in an American cohort Kathleen Maksimowicz-McKinnon, Tiffany M. Clark, Gary S. Hoffman

Purpose: Describe clinical, laboratory, and radiographic manifestations of Takayasu’s

arteritis (TAK) in an American cohort. 2) Evaluate response to interventions, remission and

relapse rates, and disease progression. 3) Compare these observations to cohorts from the

United States, Japan, India, Italy and Mexico.

Methods: Seventy-five patients were retrospectively studied using a uniform database that

included clinical, laboratory and imaging data. Vascular imaging studies were performed at

least yearly to monitor disease progression.

Results: Ninety-two percent of patients were Caucasian; 89% female. Median age at onset

was 26 years; median duration of follow up was 3.0 years. Common manifestations

included loss or asymmetry of pulse (57%), limb blood pressure discrepancy (53%), and

bruits (53%). Eleven percent of patients were asymptomatic prior to disease diagnosis.

Angiographic studies demonstrated aortic abnormalities in 79% of patients. Subclavian

(65%) and carotid (43%) arteries were also frequently affected.

Ninety-three percent of longitudinally followed patients attained disease remission of any

duration, but only 28% attained a sustained remission of at least 6 months duration while

receiving less than 10mg of prednisone/day. Both angioplasty and vascular surgery

procedures were initially successful, but recurrent stenosis occurred in 78% of angioplasty

and 36% of bypass/reconstruction procedures.

Disease manifestations in our cohort are similar to the NIH, Italian, Japanese and Mexican

cohorts in female predominance and disease manifestations, but differed from the Indian

cohort in that the latter group had a higher frequency of males, abdominal aortic

involvement, and hypertension.

Conclusions: While symptomatic improvement usually follows glucocorticosteroid

therapy, relapses usually occur in Takayasu's arteritis with dose reduction. Attempts to

restore vascular patency are often initially successful, but restenosis occurs frequently.

Chronic morbidity and disability occurs in most patients with Takayasu arteritis in the

United States.

6

Anti-Endothelial Cell Antibodies in Patients with Wegener’s Granulomatosis:

Prevalence and Correlation with Disease Manifestations and Activity JK Sebastian, A Mahr, JH Stone, Z Romay-Penabad, JC Davis, GS Hoffman, WJ McCune,

EW St. Clair, U Specks, R Spiera, S Pierangeli, and PA Merkel, for the Wegener’s

Granulomatosis Etanercept Trial Research Group.

Purpose: In small vessel vasculitis, such as Wegener’s granulomatosis (WG), histological

examination reveals endothelial cell damage. Previous studies in small cohorts of patients with

anti-neutrophil cytoplasmic antibody (ANCA) associated vasculitis have yielded conflicting

data regarding the prevalence of AECA and its use as a measure of disease activity ranging

from 20% to 100%. This study examined a large, well characterized cohort of patients with

WG and active disease for the presence of AECA.

Methods: Serum from subjects with WG who participated in a clinical therapeutic trial for

WG was collected at baseline, when all subjects had active disease. Disease activity was

measured using the BVAS/WG and clinical manifestations documented. Serum AECA (IgG)

was measured in all specimens by cyto-ELISA using unfixed human umbilical vein endothelial

cells (HUVEC). Analyses were computed by Student's t-tests and linear regression.

Results: Serum samples were evaluated for 173 patients enrolled in the trial. At baseline, 34

of 173 (19.6%) had elevated titers of AECA. Disease activity did not differ between patients

with AECA (mean BVAS/WG = 7.32 ± 3.18) or without AECA (mean BVAS/WG = 6.96 ±

3.32, p = 0.579 for difference) (Figure). Among patients positive for AECA, AECA antibody

titer did not correlate with BVAS/WG score; R = 0.09 (p = 0.57). There were no differences in

the frequency of major clinical manifestations between patients with AECA, including upper

airway, pulmonary, renal, and neurologic systems, or thrombosis.

Conclusion: AECA, as measured using HUVEC, are not highly prevalent among patients

with active WG (19.6%), are not associated with specific clinical manifestations, and do not

correlate with level of disease activity. These data question the usefulness of AECA as a

marker of disease activity in WG. Future studies should consider utilizing organ-specific adult

endothelial cells as the target for the AECA.

7

Reevaluation of the Weighting System of the Birmingham Vasculitis Activity

Score for Wegener’s Granulomatosis AD Mahr

1, T Neogi

1, H Xie

1, JC Davis

2, GS Hoffman

3, WJ McCune

4, U Specks

5, R Spiera

6, EW

St. Clair7, JH Stone

8, PA Merkel

1 for the WGET Research Group.

1Boston University, Boston, MA,

2University of California, San Francisco, CA,

3Cleveland Clinic, Cleveland, OH,

4University of

Michigan, Ann Arbor, MI, 5Mayo Clinic, Rochester, MN,

6Hospital for Special Surgery, New York,

NY, 7Duke University, Durham, NC,

8Johns Hopkins University, Baltimore, MD.

Purpose: The Birmingham Vasculitis Activity Score for Wegener’s granulomatosis (BVAS/WG) is a

validated instrument that quantifies disease activity as the weighted sum of defined manifestations plus

‘other’ signs that can be added. The item weighting was originally developed based on expert opinion

and not a data-driven approach. The aim of this study was to generate and evaluate a data-driven

weighting system for the BVAS/WG.

Methods: We used data from the Wegener’s granulomatosis etanercept trial (WGET). At each trial

visit, both BVAS/WG and a 10-cm visual analog scale for physician global assessment (PGA) were

completed. Taking the PGA as the explanatory variable, we calculated weights for: i) all 34 BVAS/WG

items (rated in ≥ 5 visits); ii) ‘other’ manifestations; and iii) an additional ‘new/worse’ variable

(invoked when ≥ 1 item was rated ‘new/worse’). Construct validity of the ‘modified’ BVAS/WG (using

the data-driven weights) were assessed for the baseline visits by correlations with PGA and compared

to the ‘original’ BVAS/WG (weights of 3 for severe and 1 for non-severe manifestations) and an

‘unweighted’ BVAS/WG (all items weighted 1). Analyses used linear regression with generalized

equation estimates and Pearson’s correlations.

Results: For the 180 WGET subjects, BVAS/WG and PGA data were available for 2020 visits.

Twenty-five BVAS/WG items were reported in ≥ 5 visits; the derived weights for these 25 items plus

‘other severe’, ‘other non-severe’, and ‘new/worse’ are shown in the Table.

Correlation coefficients with PGA of the ‘modified’ vs. ‘original’ vs. ‘unweighted’ BVAS/WG

were as follows: all baseline visits (N = 177), 0.70 vs. 0.62 vs. 0.52; baseline visits with ‘severe’ disease

(N = 106), 0.72 vs, 0.67 vs. 0.56; baseline visits with ‘limited’ disease (N = 71)*, 0.51 vs. 0.28 vs. 0.28

(*P = 0.02 for comparison of ‘modified’ and ‘original’ correlation coefficients).

Data-Driven Weights for 29 Variables of BVAS/WG (original weights in parentheses)

Sign/symptom Weight Sign/symptom Weight

Alveolar hemorrhage 20 (3) Nasal discharge/crusting 6 (1)

Red blood cell casts 18 (3) Other non-severe 6 (1)

≥ 30% creatinine rise 17 (3) Motor neuropathy 5 (3)

Sensory neuropathy 16 (3) Arthralgia/arthritis 5 (1)

Pulmonary nodules or cavities 12 (1) Purpura 5 (1)

Other infiltrates 12 (1) Skin ulcers 5 (1)

Hematuria 12 (1) Retroorbital mass/proptosis 5 (1)

Other severe 11 (3) Fever 4 (1)

Cranial nerve palsy 9 (3) Conductive deafness 2 (1)

Sensorineural deafness 9 (1) Endobronchial involvement 1 (1)

Sinus involvement 8 (1) Mouth ulcers 1 (1)

Subglottic inflammation 8 (1) Conjunctivitis/episcleritis 1 (1)

Pleurisy 8 (1) Swollen salivary gland 1 (1)

Scleritis 6 (3) New/worse 11 (–)

Conclusions: This study suggests that a data-driven weighting system enhances the construct validity

of BVAS/WG, especially for patients with low-medium disease activity. The potentially greater

precision and wider range of scores of such a modified BVAS/WG may better capture the wide

spectrum of disease activity in WG, and detect more subtle therapeutic effects. These findings warrant

confirmation in an independent sample.

8

Limited versus Severe Wegener’s Granulomatosis: Outcomes and response to

therapy in the Wegener’s Granulomatosis Etanercept Trial Seo P, Min Y-I, Holbrook JT, Hoffman GS, Merkel PA, Spiera R, Davis JC, Ytterberg S, St.

Clair EW, Specks U, and Stone JH for the WGET Research Group

Purpose: In the Wegener’s Granulomatosis Etanercept Trial (WGET), all patients received

standard immunosuppressive therapy. The addition of etanercept to standard therapies had no

effect on disease outcomes. Because the experimental therapy had no discernable effect, the

WGET cohort provides a unique opportunity to compare outcomes among patients with limited

versus severe disease. This study compares the accumulation of damage and response to

therapy of patients with limited or severe Wegener’s granulomatosis enrolled in the WGET.

Methods: Patients in the WGET cohort were stratified by disease classification (i.e., limited

versus severe disease), and the frequency with which patients with limited or severe disease

experienced disease flares, remission, sustained remission, or sustained low disease activity

was calculated. Disease activity was scored at each trial visit by means of the Birmingham

Vasculitis Activity Score for Wegener’s Granulomatosis (BVAS/WG). Patients’ quality of life

was self-reported at each trial visit on the Short-Form 36 version 2 health survey (SF-36). Data

relating to the Vasculitis Damage Index (VDI) were collected at baseline and at the end of the

trial.

Results: The 180 patients in the WGET cohort experienced a total of 261 disease flares,

including 210 limited disease flares and 51 severe disease flares. Patients with limited

Wegener’s granulomatosis were at substantially higher risk for disease flare (hazard ratio (HR)

1.80; 95% confidence interval (CI) 1.29, 2.52; P<0.001), largely due to an increased risk of

limited disease flares (HR 1.89, 95% CI 1.31-2.73; P<0.001). Patients with severe disease

were significantly more likely to achieve disease remission (HR 1.49, 95% CI 1.06-2.12;

P=0.023), sustained disease remission (HR 1.49, 95% CI 1.01-2.22; P=0.047) and sustained

low disease activity (HR 1.67, 95% CI 1.16-2.38, P=0.005) than patients with limited disease.

At trial entry, patients with limited disease were more likely to report otolaryngologic damage

(53.8% versus 25.0%, P<0.01), while patients with severe disease were more likely to report

renal damage (1.9% versus 22.6%, P<0.01). The distribution of other forms of damage did not

differ significantly among patients with limited versus severe disease phenotypes. The median

VDI score at the end of the trial for patients in both groups was 1 (P=0.53). The mental and

physical component summary scales of the SF-36 did not differ between patients with limited

versus severe disease, either at trial entry or at the end of the trial.

Conclusions: Despite different phenotypes and medication exposures, patients with limited

and severe Wegener’s granulomatosis experience overlapping forms of damage and report

similar effects on their quality of life. Patients with the limited form of Wegener’s

granulomatosis, however, are at substantially higher risk of disease flare than patients with

severe Wegener’s granulomatosis.

9

Antineutrophil Cytoplasmic Antibodies Against Proteinase 3 Do Not Predict

Disease Relapses in Wegener’s Granulomatosis Javier D. Finkielman, Peter A. Merkel, Darrell Schroeder, Gary S. Hoffman, Robert Spiera, E.

William St. Clair, John C. Davis, W. Joseph McCune, Andrea K. Tibbs, Steven R. Ytterberg,

Amber M. Hummel, Margaret A. Viss, Tobias Peikert, John H. Stone, and Ulrich Specks for

the WGET Research Group

Background: The clinical utility of ANCA levels to guide the management of patients with

Wegener’s granulomatosis (WG) remains controversial. A recent study suggested that ANCA

against pro-PR3 might correlate better with disease activity than ANCA against mature-PR3.

The main objectives of this study were to determine if the levels of pro- versus mature-PR3-

ANCA correlate better with disease activity; and if increases in their levels predict relapses.

Methods: Clinical data and serum samples of 156 patients with WG collected prospectively

during a recent randomized trial were included. PR3-ANCA levels were measured by capture

ELISA, and disease activity by the Birmingham Vasculitis Activity Score for WG.

Results: A weak association with disease activity was found for both mature- (r2=0.030,

P=0.031) and pro-PR3-ANCA levels (r2=0.038, P=0.016) at baseline. Similar findings in

individual patients were obtained from repeated assessments over time using multiple linear

regression analysis (partial r2=0.069, P<0.001 for mature-PR3-ANCA; partial r

2=0.049,

P<0.001 for pro-PR3-ANCA).

The median time to sustained remission was 12 months among patients whose mature-PR3-

ANCA levels decreased over the initial four visits (approximately 6 months) versus 19 months

among patients whose mature-PR3-ANCA levels did not decrease (unadjusted P= 0.017,

adjusted P= 0.043). This analysis did not achieve statistical significance for pro-PR3-ANCA

levels (unadjusted P= 0.123; adjusted P= 0.284).

Cox proportional hazards regression analyses showed that neither increases in mature- (hazard

ratio: 1.0, 95% C.I.: 0.5 to 1.9, P=0.973) nor pro-PR3-ANCA levels (hazard ratio: 1.0, 95%

C.I.: 0.5 to 1.9, P=0.915) predicted relapses. The proportion of patients that relapsed within 1-

year following an increase in ANCA levels was 40% for mature-PR3 (95% C.I.: 18 to 56), and

43% for pro-PR3 (95% C.I.: 22 to 58).

Conclusion: No advantage of measuring pro- versus mature-PR3-ANCA was found in this

study. Increases in PR3-ANCA levels do not predict relapses and should not be used to guide

immunosuppressive therapy.

10

THE TREATMENT OF GIANT CELL ARTERITIS WITH BIOLOGIC

AGENTS (2005 International workshop on GCA)

Gary S. Hoffman M.D., M.S.

Giant Cell (Temporal) Arteritis (GCA)

Although the signs and symptoms of GCA are exquisitely sensitive to treatment with high

doses of glucocorticosteroids (GCS), morbidity from treatment and GCA itself is substantial.

Morbidity includes visual loss in 5-25% of patients, which although better than in the era

preceding GCS (30-60% blindness), is still a terrible handicap for the already impaired elderly

patient. In one population-based study, 17% of GCA patients developed aortic aneurysms that

were sometimes associated with dissection or vessel rupture. Aortic branch vessel stenoses

may cause extremity (upper > lower) claudication (15%). Patients may also experience

polymyalgia rheumatica (PMR) (~50%), constitutional symptoms (~50%) and stroke (0-5%).

Whereas some early reports of GCA treatment suggested that GCS may only be necessary for

6-12 months, in 1973, Beevers et al. recognized the chronic nature of this illness and noted that

in many cases GCS therapy may be required for several years. Indeed, this is now a widely

accepted observation. Relapse rates in the course of GCS tapering have been reportedly ~30 –

> 80% over 1-4 years of follow up. After 2-3 years of therapy about 50% of patients remain

GCS-dependent. The risk of fractures and cataracts are 5 and 3 times greater, respectively, in

patients with GCA compared to age-matched controls not treated with GCS. Nesher et al.

found that among 43 patients followed for a mean period of 3 years, 35% had fractures and

21% had severe infections, which in 2/3 of cases lead to death. An important role for GCS

could be implicated in 37% of all deaths. The need for prolonged GCS therapy to control

GCA, and the goal of reducing disease- and treatment-related morbidity and mortality, has lead

investigators to explore the use of adjunctive agents to improve outcomes.

Adjunctive Therapy in GCA Numerous studies have explored the utility of either methotrexate or azathioprine as a means of

achieving improved disease control and less dependence on GCS therapy. Two randomized,

double-blind, placebo-controlled studies of weekly methotrexate (MTX) have been completed.

In both, the rate of GCS taper was rapid, so that in the absence of relapse, GCS withdrawal

could be accomplished in four months or 6 months. In both studies relapses were frequent and

the first relapse occurred with equal frequency in the GCS-only and GCS + MTX groups.

However, the frequency with which more than one relapse occurred differed between groups in

one study and not the other. Jover et. al. found that MTX diminished second relapses and

cumulative GCS use, whereas Hoffman et al. did not find MTX to be beneficial. The reason

for these different conclusions is uncertain. Consequently, what role MTX or other adjunctive

therapies may play in GCA remains unsettled.

Rationale for Biologic Agents in the Treatment of GCA

Although the pathogenesis of GCA has not been completely elucidated, our understanding of

the disease has grown substantially. Biopsy specimens obtained at different stages in the

evolution of vascular lesions have revealed that inflammatory cells are initially concentrated in

the adventitia and are absent or sparse in the intima, and there being an intermediate presence

in the media. Mononuclear cells migrate into the vessel wall from the adventitia. Slow flow of

blood in the adventitial microcirculation (vasa vasorum) compared to the larger vessel lumen

favors developing initial vessel wall inflammation in this location. CD4+ T cells, and CD68+

11

macrophages are more commonly found in the adventitia and the intima than in the media.

Giant cells appear to be formed from macrophages that are recruited from the adventitia and

traverse the media, later also appearing in the intima of the large vessel lumen. CD4+ T cells

are prevalent in this infiltrate, and may play a key role in driving the inflammatory attack.

Production of IL-2, TNFα�� and IFNγ� by CD4+ T cells indicates a predominant Th1 response.

Products of activated macrophages include IL-1 and TNF� which are pro-inflammatory

cytokines that further stimulate the Th1 response. Granuloma formation depends on Th1

cytokines, and in animal models, anti-TNFα therapy has been shown to block granuloma

formation. Blockade of these cytokines could theoretically play an important role in selective

interference with disease progression.

TNFα inhibitors such as infliximab, eternacept, adalimumab and IL-1 receptor antagonist (IL-

1ra) have been shown to abrogate inflammatory responses and limit tissue damage in patients

with rheumatoid arthritis and are being studied in other illnesses in which macrophage and Th1-

mediated responses may be important. Our current understanding of the pathogenesis of GCA

suggests that interfering with vascular injury due to the products of activated macrophages and

Th1 lymphocytes would be worthy of investigation. Because of their apparent important roles in

GCA pathogenesis, blockade of either IL-1, IFNγ� and TNF are particularly attractive

considerations.

The use of infliximab (IFX) as adjunctive therapy to GCS in new onset GCA has recently been

investigated in a multicenter, randomized, double-blind, controlled trial (21). Forty-four

patients were randomized in a 1:2 ratio (placebo, n=16, IFX, n=28) at 22 centers in 5 countries.

Mean age in both groups was 71.0 years (range: 50-93). After a 22 week interim analysis, the

study was terminated. There were no differences between groups in regard to: 1) proportion of

patients remaining relapse-free (50% placebo vs. 43% IFX, p=0.651), 2) cumulative doses of

GCS therapy (placebo mean ±SD = 3117 ±971 mg vs. IFX = 3051 ±770 mg, p=NS) and 3)

among patients who had 1st relapse, days to 1st relapse (placebo median = 84.5, IFX median =

96.5, p=NS). There was one case of heart failure in an IFX-treated patient. Infections were

infrequent and limited to upper respiratory tract (12.5% placebo vs. 14.3% IFX). In this elderly

population of patients with newly diagnosed GCA, IFX 5mg/kg every 8 weeks, was well-

tolerated, but during 22 weeks of treatment did not reduce number of 1st relapses or

cumulative GCS dosage.

While these results were unexpected and disappointing, they should not discourage further

exploration of blockade of other seemingly critical cytokines in the future. The morbidity of

GCA and current therapies are far too great to accept the status quo.

12

Evolution of the syndrome of Benign Angiopathy of the Central Nervous System

(BACNS): Retrospective review of 48 patients Swati S. Bharadwaj, Rula A. Hajj-Ali, Jeffrey P. Hammel, Leonard H. Calabrese. Cleveland

Clinic Foundation, Cleveland, OH

Purpose: The entity of BACNS was first proposed in 1993 as a subset of Primary Angiitis of

Central Nervous System (PACNS), defined by acute headache and/stroke, benign

cerebrospinal fluid, high probability cerebral angiogram and a favorable clinical outcome

without intense immunosuppression. In 2002, we described a series of 16 patients with

BACNS with rapid reversal of angiographic changes suggesting that the underlying mechanism

was reversible vasoconstriction rather than true vasculitis. Recognizing this entity is clinically

important as such patients can be spared aggressive studies such a brain biopsy and toxicities

of unnecessary immunosuppression. We now extend the clinical description of BACNS with a

larger cohort of 48 patients in an attempt to refine our diagnostic and treatment strategies.

Methods: A retrospective chart review was conducted on BACNS patients. Diagnosis was

based on prior descriptions (Hajj-Ali et al. Arth Rheum 47(6): 662-669, 2002). Information

was collected on demographics, triggers, strokes, headaches, visual symptoms, seizures,

laboratory markers, spinal taps, brain biopsies, initial and follow up neuroimaging, treatments,

outcomes and dates. The information was entered into an Oracle Database and a descriptive

analysis was conducted.

Results: 48 patients were included. Mean age was 43.8 years and 90% were females.

Associated factors were marijuana in 14.6 % and sympathomimetic amines in 8.3%. Past

history of migraine was seen in 22.9%. Headache was most common and present in 89.6%,

stroke in 14.6%, isolated visual symptoms in 20.8% and seizure in 4.2%. Of the 83.3% of

patients who had spinal taps, 82.5% had normal WBCs and 85% had normal CSF proteins

(7.5% were not reliable due to trauma/bleed, 7.5% had mild elevations). Brain biopsy was

performed in 22.9% and was normal in all except one patient with a concomitant glioma. Initial

MRI was normal in 17.1%. Subsequent MRIs revealed abnormalities in 81.4% (44.2%

ischemia, 18.6% cerebral hemorrhage, 14% subarachnoid hemorrhage). Stroke was bilateral or

multiple in 53%. All patients had high probability neurovascular imaging ¡®consistent¡¯ with

vasculitis. Follow up neurovascular imaging performed in 79.2% demonstrated reversibility of

the vascular abnormality in all, except 1 recent patient with a premature MRA. The median

time to reversibility was 92 days. Treatment was non-standardized with brief course

glucocorticoids used most frequently and more recently encountered patients receiving calcium

channel blockers. Total clinical recovery occurred in 47.9%, recovery with residual deficit in

33.3% and relapse in 3%.

Conclusions: BACNS is a common mimic of PACNS and appears to represent a syndrome of

reversible cerebral vasoconstriction and not true cerebral vasculitis. Initial neuroimaging can

be negative in many patients. Outcome appears to be limited by stroke morbidity. Studies to

define validated diagnostic criteria and optimal therapy are needed.

13

Tumor-like Mass Lesion (ML) - an Under-recognized Presentation of Primary

Angiitis of the Central Nervous System (PACNS)

Eamonn S. Molloy, Leonard H. Calabrese.

Objectives: ML is an under-recognized presentation of PACNS. The purpose of this work was

to document the occurrence of ML in PACNS and to assess the utility of diagnostic testing and

treatment.

Methods: We examined the case records of the Cleveland Clinic Foundation (CCF) and the

English language medical literature for biopsy-proven PACNS cases presenting as ML.

Relevant clinical variables were extracted and analyzed with JMP software.

Results: We identified 34 ML among 626 PACNS cases (5.4%). A higher percentage (11/42;

26%) was seen in the amyloid-related angiitis subset. Mean age was 46 years (range 6-74yrs).

18 patients were male, 16 female. Headache was the most common presenting symptom

(74%), with seizures occurring in 47%. 18/27 (67%) patients presented within 1 month of onset

of symptoms, only 3/27 (11%) had symptom duration of 6 months or more. ML were identified

by MR in 20/20 patients and by CT in 27/28 cases in which these investigations were

performed. Edema, enhancement and hemorrhage were associated with the ML in 14, 12 and 5

patients respectively. Cerebrospinal fluid was abnormal in 14 of the 19 patients in which a

lumbar puncture was performed (74%), generally elevation of protein ± cell count. Cerebral

angiography was abnormal in 8/14 patients; 7/8 showed a mass effect but no suggestion of

vasculitis. The diagnosis of PACNS was confirmed by histology in all cases, with

granulomatous vasculitis in 56% and lymphocytic vasculitis in 44%. Positive staining for

amyloid was found in 11/34 (32%). 29/33 (88%) of patients entered remission, 18 (62%) of

whom had residual deficits, most commonly hemianopia, seizures or cognitive deficits.

Patients were followed for a median of 18 months. 6/27 (22%) patients developed disease

relapse. Five patients (15%) died during the period of follow up; of these, four had amyloid-

related angiitis and one had lymphoma. Ten patients were treated with high-dose steroids for

more than three months in combination with cyclophosphamide, all of whom entered

remission. Three of these relapsed, none of whom were considered adequately treated with

cyclophosphamide. Of the 15 patients not receiving high-dose steroids and cyclophosphamide,

three died, three relapsed and two underwent surgical resection. Of the 11 patients who

underwent resection of the ML, 2 died shortly after surgery; while none of the others had a

disease relapse, 7 had residual deficits.

Conclusions: Although rare, PACNS should be considered in the differential diagnosis of ML.

There are no distinctive neuroimaging findings and angiography does not show changes

consistent with vasculitis. Biopsy evidence of angiitis is required for diagnosis and specimens

should routinely be stained for amyloid. Greater awareness of this manifestation may facilitate

more prompt diagnosis and treatment. While excision of the lesion may be curative, institution

of aggressive immunosuppressive therapy appears to be associated with favorable outcomes

and may obviate the need for surgery.

14

Pachymeningitis an under-recognised manifestation of rheumatic diseases

Molloy ES, Calabrese LH, Embi P, Carey JJ.

Department of Rheumatic and Immunologic Diseases, A50, The Cleveland Clinic Foundation,

9500 Euclid Avenue, Cleveland, OH 44195, USA.

Pachymeningitis is a rare complication of several rheumatic diseases, historically associated

with a high mortality and often not diagnosed until post-mortem.

AIM: To describe the clinical features, treatment and clinical course of five cases of

pachymeningitis.

METHOD

Retrospective case series.

RESULTS

2 subjects had features of pachymeningitis at the time of diagnosis (1 rheumatoid arthritis,

1Wegener’s granulomatosis), and 3 developed this phenomenon as a late complication of their

disease (1 each of rheumatoid arthritis, Wegener’s granulomatosis and sarcoidosis). All had

pachymeningitis on imaging and negative CSF cultures. 4 had pathologic confirmation, 1 did

not have a biopsy as the clinical presentation was confirmatory. All were treated with

corticosteroids and additional DMARDs (3 cyclophosphamide). All patients improved with

treatment and remain alive at follow-up: range 5 to 31 months. 1 had recurrent disease, 1 is

finishing a course of intravenous cyclophosphamide. 3 remain in remission. Details of each

case will be presented including images of MRI scans and pathology.

CONCLUSIONS

These cases highlight the clinical features and course of pachymeningitis, a rare manifestation

of some rheumatic diseases. Presentation may occur at various stages in the disease course and

is often insidious in nature. Diagnosis is based on typical MRI, CSF and/or pathologic findings,

following the exclusion of other aetiologies such as infection and malignancy. Early and

aggressive treatment is warranted, which may ameliorate the disease and improve prognosis.

Pachymeningitis should be considered in the differential diagnosis of rheumatology patients

presenting with headache or other neurologic symptoms.

15

II. Microcrystalline arthritis

BCP crystals increase prostacyclin (PGI2) production and upregulate the PGI2

receptor in osteoarthritic synovial fibroblasts; potential effects on mPGES1 and

MMP-13 *ES Molloy

1 MB MRCPI, MP Morgan

1 PhD, B McDonnell

1 BA (Mod), J O’Byrne

2 MD

FRCSI, GM McCarthy1-3

MD FRCPI 1Department of Molecular and Cellular Therapeutics, Royal College of Surgeons of Ireland,

123 St Stephen’s Green, Dublin 2, Ireland; 2National Orthopaedic Hospital, Cappagh, Dublin

15, Ireland; 3Division of Rheumatology, Mater Misericordiae University Hospital, Eccles St,

Dublin 7, Ireland

Objective: To investigate the potential involvement of prostacyclin (PGI2) in basic calcium

phosphate (BCP) crystal-induced responses in osteoarthritic (OA) synovial fibroblasts (OASF).

Methods: OASF grown in culture were stimulated with BCP crystals. PGI2 production was

measured by enzyme immunoassay. The relative dependence of PGI2 production on COX-1

and COX-2 was assessed using SC-560 (a COX-1 selective inhibitor) and SC-236 (a COX-2

selective inhibitor). Expression of mRNA transcripts was assessed by real-time polymerase

chain reaction (PCR). Protein production was measured by western blot. The effects of

iloprost, a PGI2 analogue, on expression of genes implicated in OA such as microsomal

prostaglandin E2 synthase 1 (mPGES1) and matrix metalloproteinases (MMPs) was also

studied. Since it has been reported that farnesylation of the IP receptor is required for its

efficient intracellular signaling, FPT inhibitor II, a farnesyl transferase inhibitor, was used to

antagonize iloprost-induced responses.

Results: BCP crystal stimulation led to a 5-fold increase in PGI2 production in OASF

compared to untreated cells. This induction was attenuated by selective cyclooxygenase

(COX)-2 and COX-1 inhibition (with SC-236 and SC-560 respectively) at 4 and 32 hours

respectively. PGI2 synthase and PGI2 receptor transcripts were constitutively expressed in

OASF. BCP crystals upregulated PGI2 (IP) receptor expression two-fold. While iloprost

diminished BCP crystal-stimulated PGI2 receptor upregulation, this inhibitory effect of iloprost

was blocked by FPT inhibitor II. In addition, iloprost upregulated mPGES1 but downregulated

MMP-13 expression in BCP crystal-stimulated OASF, effects that were not influenced by the

FPT inhibitor II. The effects of iloprost on mPGES1 in BCP crystal-treated OASF were

confirmed by western blotting.

Conclusions: These data show for the first time that BCP crystals increase PGI2 production

and upregulate expression of the IP receptor in OASF. The potential of PGI2 to influence BCP

crystal-stimulated responses was supported by the effects of iloprost on the expression of the

IP receptor, mPGES1 and MMP-13. These data demonstrate the potential involvement of PGI2

in BCP crystal-associated OA and suggest that inhibition of PG synthesis with non-steroidal

anti-inflammatory drugs may have both deleterious and beneficial effects in BCP crystal-

associated OA.

16

Mechanism of Cyclooxygenase Induction by Basic Calcium Phosphate Crystals

in Osteoarthritic Synovial Fibroblasts ES Molloy

1,4, MP Morgan

1, G Doherty

2, J O'Byrne

3, DJ Fitzgerald

2, GM McCarthy

1,2,4.

1Department of Clinical Pharmacology, Royal College of Surgeons in Ireland,

2Conway

Institute, University College Dublin, 3Department of Orthopaedic Surgery and

4Department of

Rheumatology, Mater Misericordiae University Hospital, Dublin, Ireland.

Introduction: Intraarticular basic calcium phosphate (BCP) crystal deposition has been

implicated in the pathogenesis of osteoarthritis (OA). Prostaglandin E2 (PGE2) may contribute

to cartilage degradation in OA. We have previously demonstrated that BCP crystals upregulate

mRNA expression and protein production of cyclooxygenase (COX) -1 and -2, both of which

contribute to the observed PG production in OA synovial fibroblasts (OASF). In addition, we

have previously shown that BCP crystals upregulate interleukin-1β (IL1β) expression in

human fibroblasts. The aim of this work was to explore the mechanism of COX upregulation

by BCP crystals.

Methods: OASF were obtained at joint replacement. All experiments were performed on cells

cultured in monolayer. mRNA expression was determined by quantative real-time PCR. OASF

were pre-treated with the following signal transduction pathway inhibitors: SB203580 (p38

MAP kinase inhibitor), UO126 (ERK1/2 pathway inhibitor), wortmannin (PI3 kinase

inhibitor), bisindolylmaleimide I (protein kinase C (PKC) inhibitor) and SP600125 (JNK

inhibitor). Interleukin-1 receptor antagonist (IL1ra) was used to explore the potential

involvement of IL1 in upregulation of COX by BCP crystals. The role of intra-lysosomal

crystal dissolution was assessed using bafilomycin A1, a proton pump inhibitor. The effect of

cycloheximide, an inhibitor of protein synthesis, was also examined.

Results: BCP crystal-induced COX-1 expression was significantly inhibited by pre-treatment

with either UO126 or bisindolylmaleimide I. SP600125 had a partial inhibitory effect. The

effects on BCP crystal-induced COX-1 upregulation of a panel of PKC inhibitors suggested

that PKCµ was the PKC isoform primarily involved. BCP crystal induction of COX-1 but not

COX-2 mRNA expression was inhibited by bafilomycin A1. Inhibition of new protein

translation using cycloheximide also prevented BCP induction of COX-1 but not COX-2

mRNA expression. There was no demonstrable effect of IL1ra on BCP crystal induced COX-1

or COX-2 expression.

Conclusions: Induction of COX-1 by BCP crystals involves the PKC and ERK1/2 signal

transduction pathways. Intra-lysosomal crystal dissolution and new protein synthesis are also

necessary for BCP crystal induced COX-1 but not COX-2 upregulation, suggesting that the

induction of COX-1 and COX-2 by BCP crystals occurs via distinct and separate mechanisms.

Involvement of ΙL-1β in BCP crystal-induced COX expression could not be confirmed.

17

Basic Calcium Phosphate Crystals Upregulate MMP-13 in Human Osteoarthritic

Synovial Fibroblasts: Effect of PGE2

ES Molloy1,4

, MP Morgan1, G Doherty

2, J O'Byrne

3, DJ Fitzgerald

2,

GM McCarthy1,2,4

. 1Department of Clinical Pharmacology, Royal College of Surgeons in Ireland,

2Conway

Institute, University College Dublin, 3Department of Orthopaedic Surgery and

4Department of

Rheumatology, Mater Misericordiae University Hospital, Dublin, Ireland.

Basic calcium phosphate (BCP) crystals are present in 30-60% of synovial fluids of patients

with osteoarthritis (OA). BCP crystal induction of fibroblast proliferation, and matrix

metalloproteinase (MMP), prostaglandin and nitric oxide production may contribute to the

pathogenesis of OA. The collagenase group of MMPs include MMP1, 3, and 13; MMP-13 is

particularly implicated in matrix degradation in OA. While the mechanisms of BCP crystal-

induced MMP 1 and MMP-3 production have been characterised, the induction of MMP 13

from human osteoarthritic synovial fibroblasts (OASF) stimulated by BCP crystals has been

less well studied.

Methods: OASF were obtained at joint replacement. All experiments were performed on cells

cultured in monolayer. mRNA expression was determined by quantative real-time PCR. OASF

were pre-treated with the following signal transduction pathway inhibitors: SB203580 (p38

MAPK inhibitor), UO126 (ERK1/2 pathway inhibitor), wortmannin (PI3 kinase inhibitor),

bisindolylmaleimide I (protein kinase C (PKC) inhibitor) and SP600125 (c-jun N-terminal

kinase inhibitor). Interleukin-1 receptor antagonist (IL1ra) was used to explore the potential

involvement of IL1 in upregulation of MMPs by BCP crystals. The role of intralysosomal

crystal dissolution was assessed using bafilomycin A1, a proton pump inhibitor. The effect of

cycloheximide, an inhibitor of protein synthesis, was also examined. The potential role of

PGE2 in BCP crystal-induced MMP expression was assessed by adding PGE2, with or without

L-161982, a selective EP4 receptor antagonist.

Results: BCP crystal upregulation of MMP-13 mRNA expression was maximal at 32 hours.

This induction was abrogated by pre-treatment with SB203580, a p38 MAP kinase inhibitor or

UO126, an ERK1/2 pathway inhibitor. BCP crystal induction of MMP-13 mRNA expression

was inhibited by bafilomycin A1. Inhibition of new protein translation using cycloheximide

also prevented BCP crystal induction of MMP-13 mRNA expression. Addition of PGE2

decreased, increased and did not significantly influence BCP crystal-induced MMP-13, MMP-

3 and MMP-1 mRNA expression respectively. These effects were only partially suppressed by

pre-treatment with L-161982. There was no demonstrable effect of IL1ra on BCP crystal

induced MMP expression.

Conclusions: These data suggest that induction of MMP-13 by BCP crystals involves the p38

and ERK1/2 signal transduction pathways. Intra-lysosomal crystal dissolution and new protein

synthesis are also required. PGE2 may also differentially influence BCP crystal-induced MMP

mRNA expression. A role for ΙL1β in BCP crystal-induced MMP expression could not be

confirmed.

18

III. Scleroderma and Raynaud's vasospasm

Evaluation of Digital Microcirculation in Systemic Sclerosis and Raynaud’s

Disease using Laser Doppler Imaging Feyrouz Al-Ashkar, James Bena, Kay Stelmach, Soumya Chatterjee. Cleveland Cinic

Foundation

Purpose: Evaluation of digital microvascular flow (DMF) and assessment of therapeutic

responses in scleroderma (SSc) and in Raynaud’s disease (RD) have been largely clinical. We

lack a standardized quantifiable objective measure of DMF.

The aims of the study were to determine whether measurement of DMF by laser Doppler

perfusion imaging (LDI) can: (1) estimate the reproducibility of mean responses within SSc,

RD and healthy controls (HC) at basal conditions (25°C), and then after 'cold challenge (CC)'

(10°C, 2 min), and after 'warm challenge (WC)' (43°C, 2 min), (2) distinguish between the 3

groups by evaluating the differences in mean response, and (3) estimate the variability within

each group.

Methods: 15 subjects in each group had measurements by LDI at baseline and then after CC

and WC. Participants were asked to refrain from vasoactive agents prior to testing. The

dorsums of the 2nd to 5th fingers were scanned and regions of interest at the fingertips were

identified. The geometric mean perfusion level was used to summarize each finger region,

since it better reflected the central tendency of the data. Repeated measure linear models were

used to estimate the variability in mean measurements within subjects and to identify

differences between groups under basal conditions, and then after CC and WC. F-tests were

used to compare groups overall, and t-tests were used to make paired comparisons. A P-value

of 0.05 was considered significant.

Results: Data on 43 subjects were analyzed. Two SSc patients were excluded due to image

artifacts. The standard deviation of overall mean perfusion values in all 4 fingertips was 0.12

(volt) under basal conditions, and ranged between 0.27 and 0.35 for single fingers. The table

shows the observed mean perfusion levels and changes by group. RD patients had the lowest

perfusion levels under all conditions, while HC had the highest levels. A borderline difference

was observed among groups under basal conditions (p = 0.055). RD patients had a significantly

larger difference between WC and CC mean perfusion levels than HC subjects (p = 0.009). No

differences were observed between SSc and the other groups.

Conclusion: RD patients had lower basal perfusion levels, and showed more change with CC

and WC than the control subjects. Differences between SSc and HC subjects were not

significant, possibly due to the small number of subjects. Thus, Laser Doppler imaging may

provide an objective measurement of DMF at baseline in patients with SSc and RD. In

planning future research, we can use the results from this pilot study to estimate group

differences and variability. Further research with a larger patient population is necessary for

evaluation of DMF in SSc patients.

19

IV. Epidemiology and Outcome

US Rheumatologist Supply and Demand : 2005-2006 Workforce Study Chad Deal

1, Walter G. Barr

2, Tim Harrington

3, Roderick Hooker

4, Paul Hogan

5, Ellen

Bouchery5, Neal Birnbaum

6, for the ACR Workforce Subcommittee COTW

7.

1Cleveland Clinic Foundation, Cleveland, OH;

2Northwestern University, Chicago, IL;

3University of Wisconsin, Madison, WI;

4University of Texas Southwestern, Dallas, TX;

5Lewin Group, Falls Church, VA;

6Pacific Rheumatology, San Francisco, CA;

7ACR, Atlanta

Purpose: A decline in US rheumatologists is anticipated when retirement rates will exceed

completed fellowships. Demand for rheumatology services is increasing due to an expanding

and aging US population, new technologies and treatments.

A workforce (WF) study by the Lewin Group in 2005-06 investigated this supply and demand

interface, the ACR Subcommittee on Workforce (Committee on Training and Workforce)

served an advisory role.

Methods: The WF study included a literature review, analyses of national databases, and a

survey sent to a random sample of 4946 adult rheumatologists (n=1683) and all 218 pediatric

rheumatologists. Survey components included work efforts, productivity, measures of excess

demand, retirement plans, sources of income, job satisfaction, and conditions treated. Pediatric,

age <40, academic, and rural rheumatologists were over-sampled. Results were used to develop

a computer model of supply and demand (S/D) that provides sensitivity analyses of key

components including number of fellows trained, retirement rates, physician assistant/nurse

practitioners (NP/PA's), hours worked, gender/age effects, lifestyle changes and changes in

GDP. A logistical regression identifying statistical differences between the groups was set at

the 95% confidence level.

Results: 1,683 rheumatologists were surveyed, 627 responded (return = 37%). 70% were male

(median age 53), median age of adult and pediatric physicians was 51 and 47. Overall there are

17.7 and 0.7 adult and pediatric rheumatologists per million people. There were 179 first-year

fellowships offered in 2005, with an occupancy rate of 91% and 95% completion rate. At least

35% were filled by international medical graduates of whom 80% remain in the US. The

number of annual visits averaged 2,624 for female and 3,548 for male rheumatologists. The

model predicts a shortfall of supply to demand (assuming S=D in 2005) of 8, 21, 27, and 56%

in 2010, 2015, 2020, and 2025.

Conclusions: At the time of this survey in 2005, the majority of rheumatologists were in their

most productive years. As a younger, increasingly female cohort follows, we predict a decline

in annual visit productivity based on the model and a shortfall of 2,600 rheumatologists by the

year 2025. Solutions could include:1) increasing fellowship positions, 2) increasing NP/PA's,

3) improving practice efficiency. The incremental addition of 188 first year fellow positions

(n=367 total first year positions) by 2021 would be required to equalize supply and demand.

The committee recommends that the ACR develop strategies that address the predicted

imbalance and use the computer model in evaluating the impact of proposed solutions.

20

A Practical Quality Reporting System for Hip and Knee Arthroplasty Procedures

American Academy of Orthopaedic Surgeons

Elaine Husni, Mary Lograsso, Dawn M Gerz, Charles Emmerich, Julie Thornton, Sarah

Worley, George F Muschler,

Background: This effective system for quality assessment following arthroplasty procedures

imposes no change in practice and eliminates provider bias and burden. Institutions are

increasingly obligated to implement proactive continuous quality assessment (QA) programs.

Addition of personnel or adding burden to busy providers is impractical, particularly in large

institutions with multiple providers and locations. This paper describes an effective institution-

wide arthroplasty QA system.

Methods: Between 12/2000 and 3/2004, 2,362 patients were scheduled for unilateral hip or

knee arthroplasty (16 surgeons, 3 locations). One month before and 10 months after surgery,

patients were mailed 8 demographic questions, the Charlson co-morbidity index, the SF-36

instrument, a cover letter and a stamped return envelope.

Results: 1,021 returned pre-op forms (43%). 679 returned post-op forms (29%). 619 patients

(26%) returned both, including 257 hips, 232 primary knees, 71 revision hips and 59 revision

knees. No differences were found between providers or treatment centers with respect to

change in Physical Composite Score (PCS). Improvement in PCS was associated with THA,

primary procedures, lower pre-op physical function, role function, bodily pain, Charlson score

and PCS, and with higher pre-op general health, role emotional and mental composite scores.

No association was found with respect to body mass index or gender,

Conclusion: This system is effective in comparing provider and care centers at acceptable cost.

It imposes no change on practice patterns, eliminates provider bias and burden, and allows

assessment of practice variation and clinical predictors of outcome. It can be implemented in

large institutions with multiple delivery sites and even on a national scale across IT systems.

21

Early Outcome Assessment Comparison of Unilateral

And Bilateral Primary Total Knee Arthroplasty

George F Muschler, Elaine Husni, Mary Lograsso, Viktor Erik Krebs, Lester Stuart Borden,

Julie Thornton, Sarah Worley

Objective: Bilateral total knee arthroplasty (TKA) is an option available to patients and

surgeons in the seeing of severe degenerative arthritis of the knee. This decision balances the

desire for a single procedure against the potential for increased short term risk and discomfort.

Disease severity, functional limitations, and the burden of co-morbid disease are expected to

influence this decision.

Methods: A prospective arthroplasty outcomes registry was used to test the hypothesis that

patients undergoing bilateral vs. unilateral TKA differ with respect to pre-operative health

perception and co-morbidity.

Results: Baseline and one year outcome was assessed in 232 unilateral and 129 bilateral TKAs

performed between 12/1/2000 and 3/24/2004. Patients reported co-morbidity, and SF-36

scores pre-op and one year post-op were compared. Patients undergoing bilateral TKA

reported: significantly higher pre-op general health perception (p<0.-001) but no difference in

pre-op physical function, role function, pain, vitality, social or role function mental health,

physical composite score, Charlson co-morbidity index, body mass index, or gender. Bilateral

patients also reported a greater change in physical composite score one year following surgery

(median change 12 vs. 9, p<0.001).

Conclusion: Perception of general health appears to be a significant factor in the decision to

undergo bilateral TKA, independent of objective co-morbidity. Patients with this perception

may be more likely to seek out and accept the increased short term risk associated with

bilateral TKA. This pre-operative perception may also predict a higher level of expectation

contributing to the improved overall outcome.

22

V. Psoriatic arthritis

Development and evaluation of PASE: A self-administered Psoriatic Arthritis

Screening and Evaluation tool Abrar A. Qureshi, MD, MPH, Darel S. Cohen, MD, MPH, Elinor Mody, MD, and

M. Elaine Husni, MD, MPH

The incidence and prevalence of psoriatic arthritis is largely unknown at the population level.

Psoriatic arthritis can result in disability and deformity if not diagnosed and treated

appropriately. However, the diagnosis of psoriatic arthritis is difficult as there are no well-

accepted diagnostic criteria and evaluation available. We have developed a self-administered

questionnaire that dermatologists can use to screen psoriasis patients for evidence of psoriatic

arthritis. This will expedite and optimize a rheumatology referral.

The PASE questionnaire (PASE – Psoriatic Arthritis Screening and Evaluation) contains a total

of 15 five-choice items and two subscales for symptoms and function. PASE minimum score

was 15 and maximum score 75. Forty participants were screened with PASE after institutional

IRB approval. All participants responded to the questionnaire prior to intervention or systemic

therapy. Seventeen participants were diagnosed with psoriatic arthritis plus psoriasis and 23

with psoriasis alone by a dermatologist and rheumatologist working in collaboration. Average

age was 59 years and 19 were female. In the psoriatic arthritis group, total median score was

51 (25th

, 75th

percentiles = 47, 63). In the psoriasis alone group, total median score was 37

(25th

, 75th

percentiles = 30, 47), p = 0.003. Sensitivity was 0.76 and specificity was 0.70 for

the total score. Similarly, median function score for psoriatic arthritis was 26 (25th

, 75th

percentiles = 23, 31) and median function score for psoriasis alone was 18 (25th

, 75th

percentiles = 16, 22), p = 0.001; sensitivity was 0.88 and specificity was 0.78. Median

symptom score for psoriatic arthritis was 24 (25th

, 75th

percentiles = 23, 30) and median

symptom score for psoriasis alone was 21 (25th

, 75th

percentiles = 14, 24), p = 0.001;

sensitivity was 0.76 and specificity was 0.65.

The PASE questionnaire is a valid screening tool for detection of psoriatic arthritis among

patients with psoriasis. There is a need to identify patients with PsA so that timely therapy can

be initiated and long term disability avoided. Further work is on-going to test PASE for

internal consistency and test-retest reliability in a larger sample of patients with psoriasis and

psoriatic arthritis.

23

Diagnosis of arthritis in psoriasis patients presenting with joint pain to a

dermatology-rheumatology clinic Elinor A. Mody, MD, M. Elaine Husni, MD, MPH and Abrar A. Qureshi, MD, MPH

Psoriatic arthritis (PsA) is an inflammatory spondyloarthropathy associated with psoriasis. The

epidemiology of psoriatic arthritis is largely unknown as the diagnosis is difficult to make.

Currently there are no formally accepted diagnostic criteria for PsA in adults. When untreated,

PsA can be progressive in some patients leading to permanent joint destruction and disability.

Consecutive patients presenting to the Center for Skin and Related Musculoskeletal Diseases at

Brigham and Women’s Hospital with psoriasis and joint pain were included in this study. All

71 patients were simultaneously evaluated by dermatology and rheumatology and all had a

diagnosis of psoriasis. The average age was 55 years and 52% were female. About 41%

(29/71) of patients were diagnosed with PsA alone on the basis of clinical history, physical

examination and radiographs. A further 27% (19/71) were diagnosed with PsA and

osteoarthritis (diagnosis based on radiographs). Hence, about 55% (39/71) were diagnosed with

PsA overall in this population. A total of 14% patients reporting joint pain could not be

diagnosed with any form of arthritis. The mean age of patients in the PsA group was 49 years

and non-PsA group was 57 years (P=0.01). Family history of psoriasis or PsA in a first- or

second-degree relative was reported by 7% of participants. Radiograph data was available for

51 participants and 39% (20/51) had erosions present. Data on Schober’s test was available for

52 participants, was abnormal in 24% (16/52) and 88% with an abnormal Schober’s were

diagnosed with PsA (p=0.001). About 40% (28/70) of participants reported morning stiffness

of greater than 1 hour and 82% with morning stiffness were diagnosed with PsA (p=0.001).

Psoriatic arthritis may be more common among psoriasis patients than previously reported. A

large proportion of patients may have concomitant PsA and osteoarthritis making it more

difficult to diagnosis and treat joint symptoms in psoriasis patients. Care for patients with

psoriasis and PsA can be optimized in a multi-disciplinary environment.

24

VI. Metabolic Bone Disease Equivalence of Technology Generated T-scores and Z-scores in Young Adult Bone

Densitometry.

J.J.Carey1, M.F. Delaney

1, B. Richmond

1,3, B.A. Cromer

4, T.E. Love

5, S. Lewis

5, C.A.

Thomas5, P.D. Miller

6, A.A. Licata

1,2,

Affiliations: 1The Center for Osteoporosis and Metabolic Bone Disease, A50,

2Department of Endocrinology,

3Department of Radiology

The Cleveland Clinic Foundation, Cleveland, Ohio 44195. 4Department of Pediatrics, Division of Adolescent Medicine, BG 4,

5Center for Health Care Research & Policy, Rammelkamp Research & Educ. Bldg.,

MetroHealth Medical Center, Cleveland, OH 44109-1998 6Colorado Center for Bone Research, 3190 South Wadsworth, #250, Lakewood, CO 80227..

ASBMR 2006 Meeting Abstract The purpose of this study was to assess T-score and Z-score equivalence of central dual-energy

x-ray absorptiometry (DXA) generated in Young Adults (ages 20-50 years). Recently Z-scores

have been recommended for diagnosis of low bone mass in young men and premenopausal

women. Little difference between T-scores and Z-scores is expected in this population,

particularly young Caucasian women. Unlike T-scores, rigorous scientific publications to

support such recommendations are lacking. Additionally consensus Z-score definitions vary,

and methods for Z-score calculation differ within and between technologies. A cross-sectional

design was used for this study using a convenience sample of young adults scanned (Hologic

or Lunar) at the Cleveland Clinic Foundation (CCF) between 2000 and 2006. Paired

comparison tests and simple linear regression models were used for preliminary analyses;

because of their limitations in assessing reliability, more complex analyses including Deming

regression, graphical plots including Altman-Bland and mountain were also used to assess

differences, and bias, between T-scores and Z-scores. 4275 predominantly female Caucasians

subjects had ≥1 scans available for analysis. Overall, significant differences were seen between

DXA-generated T-scores and Z-scores at the spine and hip with both technologies: maximum

range: 3.5 (standard deviations), p-value <0.001. Stratification by gender, ethnicity, skeletal

site and age did not significantly alter the results. Caucasian women aged 20-30 years differed

significantly at the spine and hip: p-value <0.05, maximum range 2.5. Using Z-scores instead

of T-scores resulted in significant alterations in DXA diagnosis using either 1994 W.H.O. or

2005 I.S.C.D. criteria; kappa ranged from 0.573-0.808, McNemar’s p-value <0.001.

Considerable differences exist between Z-scores and T-scores in young adults which may

result in significant ascertainment bias if Z-scores are used for DXA diagnosis of osteoporosis.

Altman-Bland Plot of 20-30 year-old Caucasian Women Lumbar Spine T-scores and Z-scores:

difference V Mean.

25

-4 -2 0 2 4

AVERAGE of hipt and hipz

1.0

0.5

0.0

-0.5

-1.0

hip

t -

hip

z

Mean

-0.06

-1.96 SD

-0.76

+1.96 SD

0.65

26

Inappropriate PTH Response to Severe Vitamin D Deficiency in Patients with Chronic

Liver Disease. M. E. Schulte*

1, A. A. Licata

2, J. J. Carey

3, M. F. Delaney

3.

1Medicine, Cleveland Clinic

Foundation, Cleveland, OH, USA, 2Endocrinology, Cleveland Clinic Foundation, Cleveland,

OH, USA, 3Rheumatology, Cleveland Clinic Foundation, Cleveland, OH, USA.

The purpose of this study was to evaluate the calcium-vitamin D-PTH axis in patients with

chronic liver disease.

Vitamin D (VitD) deficiency is associated with a compensatory increase in parathyroid

hormone release to maintain serum calcium homeostasis. This secondary

hyperparathyroidism is associated with decreased calcium absorption, increased efflux of

calcium from bone, and increased risk for low bone mass. Reduced bone formation in

chronic liver disease has been ascribed to several factors including vitamin D deficiency,

osteomalacia, hypogonadism and excessive alcohol consumption. We noted an impaired PTH

response to VitD deficiency in patients with chronic liver disease.

To evaluate this we performed a cross-sectional study. A cohort of subjects undergoing

evaluation for liver transplantation seen at the Center for Osteoporosis at the Cleveland

Clinic Foundation from 2004 to 2006 were age-matched to 49 controls with vitamin D

deficiency (<20ng/ml) but without clinical or biochemical evidence of liver disease during

the same time period. Subjects with known primary hyperparathyroidism, renal insufficiency

(serum creatinine >1.4mg/dl) or other disorders of calcium metabolism were excluded. 2

sample t-tests were used to compare serum creatinine, corrected serum calcium, vitamin D

levels and serum parathyroid hormone levels.

Of 114 liver disease subjects screened, 12 were excluded due to abnormal creatinine. Of the

remaining 102 subjects, 73 had low vitamin D (<20) and 30 had very low vitamin D (<10).

3 controls were excluded, 2 primary hyperparathyroidism,1 hypocalcemia. We found no

significant difference in corrected serum calcium, creatinine or age between groups.

Although controls had higher mean serum vitamin D levels, they also had significantly

higher serum PTH levels - see table. In patients with chronic liver disease there was no

significant difference in serum PTH levels in those with low Vit D compared with those with

normal Vit D levels. Subjects with advanced chronic liver disease appear to have a

subnormal rise in PTH in the setting of Vit D deficiency. Further studies to evaluate this

phenomenon are needed.

<0.001 -46.53, -22.8971.40 (36.50)36.69 (21.26)Serum PTH

0.004-3.7, -0.714.53 (3.56)12.30 (3.78)Serum 25 vitamin D

0.89-0.22, 0.269.60 (0.48)9.61 (0.68)Corrected Calcium

0.92-0.08, 0.090.80 (0.24)0.80 (0.20)Serum Creatinine

0.80-4.69, 3.6356.7 (14.6)56.2 (8.3)Age

ControlsCirrhotics

p-value95% C.I.Mean (SD)Variable

<0.001 -46.53, -22.8971.40 (36.50)36.69 (21.26)Serum PTH

0.004-3.7, -0.714.53 (3.56)12.30 (3.78)Serum 25 vitamin D

0.89-0.22, 0.269.60 (0.48)9.61 (0.68)Corrected Calcium

0.92-0.08, 0.090.80 (0.24)0.80 (0.20)Serum Creatinine

0.80-4.69, 3.6356.7 (14.6)56.2 (8.3)Age

ControlsCirrhotics

p-value95% C.I.Mean (SD)Variable

27

The impact of body weight on DXA generated Z-scores in Young Adults

Brad Richmond1,2

, John J Carey2,3

, Miriam F Delaney2,3

, Angelo A Licata2,4

, Martha Manilich1

McIntosh, Paul D Miller4,

1Department of Radiology,

2The Center for Osteoporosis and Metabolic Bone Disease,

3Department of Rheumatology &

4Department of Endocrinology, All at The Cleveland Clinic

Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195, U.S.A. 4Colorado Center for Bone

Research, 3190 South Wadsworth, #250, Lakewood, CO 80227-4800

Background:

We have previously shown significant differences exist between DXA-generated Z-scores and

T-scores at the spine, total hip, femoral neck and trochanter in young adults. Stratifying the

analyses by technology, gender, race and age did not alter the results; differences were seen

even in White women aged 20-29 years. Z-scores are sometimes adjusted for body weight, an

important predictor of fracture risk.

Purpose:

To assess the impact of body weight on weight-adjusted DXA-generated Z-scores in young

adults.

Methods

A cross-sectional design using a convenience sample of subjects aged 20-49 years from The

Cleveland Clinic Foundation between 1999 and 2000 was used. All subjects had a DXA scan

on one of two technologies: Hologic or Lunar. Linear Regression models and stratified

bivariate analyses were used to indirectly assess the impact of body weight on DXA-generated

Z-scores at 4 skeletal sites.

Results:

Heavier individuals’ Z-scores were significantly lower than lighter individuals after adjusting

for T-score, though considerable overlap was seen between groups. Univariate linear

regression showed weight was a weak predictor of Z-score (r2

<0.023), but when coupled with

T-scores r2

was >0.99 for all 4 sites.

Conclusions:

Subject’s body weight may account for some of the difference seen between DXA generated

Z-scores and T-scores. However adjustment for weight does not fully explain the difference

between these values in this population.

28

Diagnostic Agreement between DXA generated T-scores and Z-scores in

Young Adults

John J Carey MB2,3

, Miriam F Delaney MD2,3

, Angelo A Licata MD, PhD2,4

,

Martha L Manilla-McIntosh RT1,

,3, Paul D Miller

4, Bradford J Richmond MD

1,2

1Department of Radiology,

2The Center for Osteoporosis and Metabolic Bone Disease,

3Department of Rheumatology &

4Department of Endocrinology at The Cleveland Clinic

Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195, U.S.A. 4Colorado Center for Bone

Research, 3190 South Wadsworth, #250, Lakewood, CO 80227-4800

Purpose:

To assess diagnostic agreement between DXA-generated T-scores and Z-scores in young

adults.

Methods:

This study used a cross-sectional design performed on a cohort on young adults extracted from

a convenience sample of all DXA subjects who had a BMD scan at The Cleveland Clinic

Foundation between 1999 and 2000. All subjects were scanned using one of two technologies:

Hologic or Lunar. Evaluations comparing diagnostic agreement using 1994 W.H.O. and 2005

I.S.C.D. diagnostic criteria were undertaken using McNemar’s test of proportions and Kappa

Statistics using SAS version 9.1.

Results:

4,275 unique subjects met study criteria and were available for analysis. Effective percentage

agreement was good (those with concordant T-score and Z-score diagnosis) range: 80-97%.

Kappa coefficients for agreement between groups (also assesses chance agreement) were good

to excellent: range 0.533-0.808. However significant differences in DXA diagnosis of low

bone mass or osteoporosis were seen using W.H.O. or I.S.C.D. diagnostic criteria for both

technologies when Z-scores were used instead of T-scores, all p values <0.001.

Conclusions:

Using Z-scores instead of T-scores for DXA diagnosis of osteoporosis or low bone mass in

young adults may result in significant ascertainment bias. This is an important consideration

when using DXA in such individuals.

29

Rheumatologists Order Less Testing for Secondary Causes of Osteoporosis than other

Specialties.

John J Carey, Miriam F. Delaney, Anton Mandrov, Angelo A Licata, Lisabeth V. Scalzi,

Holly L Thacker, Andrea Sikon, Chad L Deal.

Background: Prior study shows that physicians’ response to the results of bone mineral

density (BMD) testing is influenced by the format of the report. A more detailed report may

increase physician use and understanding of BMD testing, patient treatment, and testing for

secondary causes of osteoporosis (1). Secondary causes of osteoporosis or low bone mass are

prevalent, but often go unrecognized, and may require specific therapy (3,4,5). At our

institution, a statement in the BMD report to evaluate for secondary causes of osteoporosis is

recommended for all individuals with Z-scores of <-1.5. Few studies have evaluated the effect

of such recommendations, and how response varies by physician specialty.

Methods:

• Retrospective cohort study between August 1st 2002 and July 31

st 2003.

• All patients with a Z-score of <-1.5 at any site scanned on our DXA machine (G.E.

Lunar Prodigy DF 60004) between August 1st 2002 and January 31

st 2003 were

screened using BMD report and Electronic Medical Record (EMR) review.

• Subjects were eligible for study inclusion if:

1. There was such a recommendation on their BMD report;

2. EMR available for review;

3. Such a work-up had not been performed previously.

• I.R.B. approval was obtained for our study.

• All statistical analysis performed using S-Plus version 6.2, Insightful Inc. U.S.A., and

SAS Version 9.1, Cary, N. Carolina, U.S.A.

Results:

• 752 BMD reports were identified which had a Z score of <-1.5 at one site or more.

• 148 met study inclusion criteria

• 48 referred by their Rheumatologist, 47 by their internist and 53 by 13 other specialties.

• Physicians’ responses varied widely, some ordering no investigations, while others

performed limited or extensive laboratory and imaging studies (range 1-17).

• 88 of 148 (59%) subjects had at least 1 investigation performed.

• Rheumatologists were less likely to order further testing that other specialties, P = 0.004.

• The median number of investigations ordered was 7 and did not differ between specialties,

p=0.37.

• The most frequently ordered tests were serum calcium (77), 25 hydroxy vitamin D (68),

phosphorous (67), and intact parathyroid hormone (61).

• 45 patients had abnormal investigations leading to a diagnosis of secondary osteoporosis:

12 Rheumatology, 13 Internist, 20 other.

• The tests with the highest yield were 25 hydroxy vitamin D (47%), 24 hour urine calcium

(22%) and serum PTH levels (13%).

30

Conclusions:

• Physicians’ response to a recommendation to evaluate secondary causes of low bone

mass/osteoporosis on the BMD report varied.

• Secondary causes of low bone mass/osteoporosis are common.

• Rheumatologists were less likely to order ancillary testing than non-rheumatologists.

• Given the prevalence of secondary osteoporosis, such testing may be warranted