Faculty Bibliography

Purpose: Obesity is a modifiable risk factor of knee osteoarthritis (KOA). While diet, exercise and other conservative treatments can have limited and often transient beneficial effects, an alternative strategy would target weight loss via surgery to delay or avoid joint replacement. Some retrospective data, including a study from our group, have in fact shown sustained improvement in KOA pain after bariatric surgery. We initiated a prospective study to evaluate painful KOA in the obese population, and track whether weight loss after bariatric surgery affects KOA-related pain and physical function. Methods: We screened consecutive patients prior to laparoscopic adjustable gastric banding (LAGB), sleeve gastrectomy, or gastric bypass (RYGB), at NYU Langone Medical Center and Bellevue Hospital Center. Patients age >21 with knee pain for >1 month and a visual analog scale pain score >30mm were enrolled, excluding those with lupus, rheumatoid arthritis, psoriatic arthritis, or psoriasis. Baseline pre-op assessments included x-rays for OA severity by Kellgren-Lawrence (KL) grade, the Knee Injury and Osteoarthritis Outcome Score (KOOS), and the Western Ontario McMasters Universities Osteoarthritis Index (WOMAC) with a Likert scale calculated from the KOOS. Patients were consented for optional tissue collection (blood, urine and intra-operative adipose samples) for future biomarker analysis. They are (still) completing the questionnaires and being measured for BMI and % excess weight loss (%EWL) at 1,3,6 and 12 month post-op intervals. Results: In total, we screened 537 patients planning to have bariatric surgery, found that 309 (58%) of them reported knee pain - and enrolled 176who met criteria and consented for the study. Our cohort is 89.7% female, with a mean BMI of 43.6 kg/m2+/-7 (31.6-60.6), a mean age of 42.4 +/-11 (18-73), and radiographic severity as follows: KL0=43 (25%), KL1=34 (19%), KL2=38 (22%), KL3=34 (19%), KL4=27 (15%). The mean pre-op KOOS scores were 45.4 for pain and 46.0 for ADL (0=worst, 100=best), the mean pre-op WOMAC pain score (Likert scale) was 11 (0=best, 20=worst), and the mean overall WOMAC index was 52 (0=best, 96=worst). Before surgery, a higher KL correlated with symptoms; mean KOOS pain was 53.2, 48.1 and 36.7 for KL0, KL1-2, and KL3-4 (p=0.00002 for KL0 vs KL3-4, and p=0.0005 for KL1-2 vs KL3-4), with similar trends across other KOOS and WOMAC scores. Higher BMI also trended with worse pre-op knee symptoms, as the tertiles with the lowest and highest BMIs (31-39 and 46-61) had mean KOOS pain scores of 46.8 and 43.7 (p=0.37). While 23 ultimately decided against weight loss surgery, we are collecting post-operative data on the 153 patients (40 RYGB=26%, 93 sleeve=61%, 20 LAGB=13%). Improvement in average KOOS and WOMAC scores over baseline has been observed at all intervals (67, 71, 65, and 42 responses at 1,3,6,12 month visits), with more improvement farther after surgery. At 6 months post-op, mean KOOS scores available thus far improved 29 points for pain, with mean WOMAC pain and index improving by 6 and 22 points. The %EWL correlated with knee symptoms at each interval and for all followups combined, as the smallest and largest %EWL quartiles (4-29%, 54-92%) showed mean improvements of 18 and 31 points (p=0.03) in KOOS pain - mirrored across KOOS and WOMAC scores. RYGB and sleeve yielded higher %EWL than LAGB (44%, 43% vs. 37%) across all intervals, and greater improvement in mean KOOS and WOMAC scores (e.g. mean KOOS pain increased by 28, 29 and 8). Neither presence nor severity of KOA severity affected knee pain improvement from weight loss. Conclusions: These data suggest that bariatric surgery improves patients' KOA pain proportional to percent excess weight loss, with durability over time. RYGB and sleeve gastrectomy have more impact on knee symptoms than LAGB. While patients with worse KL grades report more baseline pain and disability, as expected, x-ray severity did not impact the response to surgical weight loss

Purpose: We and others have demonstrated low grade inflammation exists in OA joint tissues, where it may contribute to disease pathogenesis. In the current studies we assessed whether inflammatory events occurring within joint tissues were reported in the peripheral blood leukocytes (PBLs) of patients with symptomatic knee OA (SKOA). Methods: PBL inflammatory gene expression (IL-1, TNFalpha, COX-2) was assessed in two independent cohorts of patients with SKOA, and a cohort of healthy control subjects: 1) 111 patients with tibiofemoral medial OA and 21 healthy volunteers from the NYUHJD Cohort, and 2) 200 patients from the OAI progression cohort who had "high quality radiographs", at both baseline and 24 months, and had KL2 or 3 in the signal knee at baseline. Radiographic progression was defined as narrowing of medial joint space width (JSW) in the signal knee between baseline and 24-months in each cohort. Radiographic progressors were defined as subjects who had JSN >0.0, 0.2 and 0.5mm over 24 months. For measuring predictive performance, we used the area under the curve (AUC) of a receiver operating characteristics (ROC). OAI SKOA subjects were dichotomized as radiographic non-progressors (JSN <0.0 mm) and progressors (JSN>0.0mm) for association studies. Results: Elevated PBL expression of IL-1, TNFalpha or COX-2 identified SKOA patients who were "fast progressors" (mean JSN 0= 0.71, 0.75 and 0.71 mm / 24 months, respectively) compared to patients with levels below the median. In a multivariable model, anthropometric traits alone (BMI, gender, age) did not predict progression, whereas addition of PBL gene expressions improved prediction of fast progressors (JSN>0.5mm). We next examined inflammatory gene expression in PBLs of radiographic progressors in the OAI cohort. Similar to the NYUHJD cohort, elevated expression of IL-1beta, TNFalpha and COX-2 mRNA distinguished radiographic progressors from non-progressors (Table 1). PBL IL-1beta expression found to be strongest predictor of all three radiographic progressors. In multivariate models that combine all three markers did not improve upon IL-1beta predictivity. We thus conclude that either the signal in TNFalpha and Cox-2 is already subsumed by IL-1beta and/or that it is not easy to capture the non-overlapping signals without increasing the sample size (i.e., fitting a stronger multivariate predictor will require more sample size). Conclusions: We identified, and confirmed in two cohorts, increased inflammatory gene expression (IL-1, TNFalpha or COX-2) by PBLs that predict radiographic progression in patients with SKOA. The data indicate that inflammatory events within joint tissues of patients with SKOA are reported in the peripheral blood. These PBL transcriptome signals of local joint inflammation merit further study as potential biomarkers for OA disease progression. (Table Presented)

BACKGROUND: Gout and osteoarthritis (OA) are the most prevalent arthritides, but their relationship is neither well established nor well understood. OBJECTIVES: We assessed whether a diagnosis of gout or asymptomatic hyperuricemia (AH) is associated with increased prevalence/severity of knee OA. METHODS: One hundred nineteen male patients aged 55 to 85 years were sequentially enrolled from the primary care clinics of an urban Veterans Affairs hospital, assessed and categorized into 3 groups: gout (American College of Rheumatology Classification Criteria), AH (serum urate >/=6.8 mg/dL, no gout), and control (serum urate <6.8 mg/dL, no gout). Twenty-five patients from each group subsequently underwent formal assessment of knee OA presence and severity (American College of Rheumatology Clinical/Radiographic Criteria, Kellgren-Lawrence grade). Musculoskeletal ultrasound was used to detect monosodium urate deposition at the knees and first metatarsophalangeal joints. RESULTS: The study showed 68.0% of gout, 52.0% of AH, and 28.0% of age-matched control subjects had knee OA (gout vs control, P = 0.017). Odds ratio for knee OA in gout versus control subjects was 5.46 prior to and 3.80 after adjusting for body mass index. Gout subjects also had higher Kellgren-Lawrence grades than did the control subjects (P = 0.001). Subjects with sonographically detected monosodium urate crystal deposition on cartilage were more likely to have OA than those without (60.0 vs 27.5%, P = 0.037), with crystal deposition at the first metatarsophalangeal joints correlating most closely with OA knee involvement. CONCLUSIONS: Knee OA was more prevalent in gout patients versus control subjects and intermediate in AH. Knee OA was more severe in gout patients versus control subjects.

Objective: To analyze discordance between global estimates by patients (PATGL) and their physicians (DOCGL) according to demographic and self-report variables on a multidimensional health assessment questionnaire (MDHAQ), in patients with many rheumatic diseases seen in usual care. Methods: Each patient completes an MDHAQ at each visit, which includes scores for physical function, pain and PATGL, each found on the HAQ, and scores for sleep quality, anxiety, depression, self-report joint count and fatigue, which are not found on the HAQ. A random visit of 980 patients with any rheumatic diagnosis was analyzed in 3 categories: PATGL=DOCGL (within 2/10 units); PATGL>DOCGL (by >/=2/10 units); DOCGL>PATGL (by >/=2/10 units), using descriptive statistics and multinomial logistic regression models. Results: Patients included 145 with rheumatoid arthritis, 57 systemic lupus erythematosus, 173 osteoarthritis, 348 other inflammatory, and 257 other non-inflammatory diseases. Overall, PATGL=DOCGL in 509 (52%), PATGL>DOCGL in 371 (38%) and DOCGL>PATGL in 100 (10%). PATGL>DOCGL was associated significantly with older age, female gender, low formal education, Hispanic ethnicity, not working, high MDHAQ physical function and pain, and high scores for fatigue, poor sleep, anxiety, depression, and self-report joint count, not available on the HAQ. Pain and fatigue were significant in a final multinomial logistic regression; the other variables may raise awareness of discordance to clinicians. Conclusions: Global estimates of patients indicated significantly poorer status than estimates of their physicians in 38% of 980 patients with rheumatic conditions, and were associated with demographic and MDHAQ scores, 5 of which are not available on the HAQ. (c) 2013 American College of Rheumatology.

BACKGROUND: Age, gender and genetic predisposition are major intrinsic risk factors for osteoarthritis (OA). Iron increases are associated with age and gene mutation. In the present study, we examined whether serum ferritin, an indicator of total body iron stores, correlates with clinical features in patients with OA, and whether the hemochromatosis Fe (HFE) gene mutation plays a role. METHODS: In a 2-year longitudinal observational study, 127 patients with knee OA and 20 healthy individuals (controls) were enrolled. All patients underwent standardized weight-bearing fixed-flexion posteroanterior knee radiographs. Peripheral blood samples were analyzed for serum ferritin, and genotyped for HFE using allelic discrimination methods. RESULTS: Higher levels of serum ferritin were found in patients older than 56 years (P =0.0186) and males (P =0.0006), with a trend toward higher ferritin in patients with OA. HFE gene mutation carriers were more prevalent among patients with OA than among healthy controls. When stratified further by gender, we found that male patients with OA had higher levels of serum ferritin than male control subjects [odds ratio = 4.18 (limits of 95% confidence interval: 0.86-27.69, P = 0.048)]. Analyses of radiographic data indicated that higher ferritin was associated with narrower joint space width at baseline (P = 0.032) in male patients. Additionally, among men, risk prediction of radiographic severity [Kellgren-Lawrence (KL) grade >2)] in the higher ferritin group was almost five times that of the lower ferritin group (odds ratio = 4.74, P = 0.023). CONCLUSION: Our data suggest that increased ferritin levels are associated with symptomatic knee OA in males. This finding needs to be validated in a larger cohort of patients.

Objective. To examine the effect of rilonacept on the health-related quality of life (HRQoL) in patients with poorly controlled familial Mediterranean fever (FMF). Methods. As part of a randomized, double-blinded trial comparing rilonacept and placebo for the treatment of FMF, patients/parents completed the modified Child Health Questionnaire (CHQ) at baseline, and at the start and end of each of 4 treatment courses, 2 each with rilonacept and placebo. Results. Fourteen subjects were randomized; mean age was 24.4 +/- 11.8 years. At baseline the physical HRQoL score was significantly less (24.2 +/- 49.5) but the psychosocial score was similar to the population norm (49.5 +/- 10.0). There were significant improvements in most HRQoL concepts after rilonacept but not placebo. Significant differences between rilonacept and placebo were found in the physical (33.7 +/- 16.4 versus 23.7 +/- 14.5, P = 0.021) but not psychosocial scores (51.4 +/- 10.3 versus 49.8 +/- 12.4, P = 0.42). The physical HRQoL was significantly impacted by the treatment effect and patient global assessment. Conclusion. Treatment with rilonacept had a beneficial effect on the physical HRQoL in patients with poorly controlled FMF and was also significantly related to the patient global assessment. This trial is registered with ClinicalTrials.gov Identifier NCT00582907.