Osteoarthritis Highlights

Allan Gelber, M.D.

Abstract #1942: The Joints on Glucosamine (JOG) Study: A Randomized, Double-Blind, Placebo-Controlled Tiral to Assess the Structural Benefit of Glucosamine in Knee Osteoarthritis Based on 3T MRI.

Authors:

CK Kwoh, FW Roemer, MJ Hannon, CE Moore,  JM Jakicic, A Guermazi, SM Green and RM Boudreau, University of Pittsburgh, VA Pittsburgh Healthcare System, Pittsburgh, PA, Boston University School of Medicine, Boston, MA, University of Pittsburgh, Pittsburgh, PA, Texas Woman's University, Houston, TX

Background:

Prior studies which have addressed the structural benefit of glucosamine for knee osteoarthritis (KOA), as assessed by radiographs, have yielded conflicting results.  The aim of the present study was to determine the short-term efficacy of glucosamine, as assessed by MR imaging, based upon decreased worsening of structural lesions in KOA.

Methods:

The authors conducted the Joints on Glucosamine (JOG) study, a 24-week double-blind, placebo-controlled trial in which a 1500 mg dose of glucosamine hydrochloride once daily, in beverage form, was compared to placebo. The primary outcome was a diminished rate of knee cartilage damage. All study participants had mild-moderate knee pain and underwent 3T MR imaging of both knees, at baseline and at 24 weeks of follow-up. The presence of bone marrow lesions (BMLs) and synovitis in each knee were scored according to the WORMS system. Baseline fixed-flexion knee x-rays were read for Kellgren-Lawrence (K-L) grade and radiographic features of osteoarthritis, such as joint space narrowing using the OARSI grading system. A biomarker of cartilage synthesis, in the form of urinary excretion of C-terminal cross linking telopeptide of type II collagen (uCTX-II) was also assessed. An intent-to-treat analysis with last event carried forward was performed. Negative binomial regression for cartilage defects and multivariate logistic regression for worsening BML, taking into account clustering of knees within an individual, and adjusted for gender, age, BMI, baseline WOMAC, and K-L grade was conducted.

Results:

There were a total of 201 participants who participated in the JOG Study. The mean age of the participants was 52 years, among whom 49% were women. In terms of the primary outcome measure, the probability of worsening cartilage defects in the glucosamine group was not quantitatively different from that observed in the control group [odds ratio (OR) 0.9; 95% CI 0.55-1.58]. With regard to the outcome of bone marrow lesions, there was a trend toward less worsening of these lesions in the glucosamine compared to the placebo group [OR 0.73; 95% CI 0.50-1.07]. Interestingly, there was no definitive evidence that glucosamine improved the biomarker, uCTX-II over placebo (beta = -0.10, 95%CI -0.21-.001). Also of inteset is that the drop-out rate was much greater, at 16.5% in the placebo group vs. 5.1% in the glucosamine group, p < 0.01).
Conclusions: This short-term study provided no evidence that glucosamine results in structural benefit in knee OA as assessed using 3T MRI or uCTX.

Editorial:

There has been substantial interest over the last decade in both the medical and lay communities, as to what therapeutic benefit might be achieved from the use of glucosamine supplements in the management of knee osteoarthritis. In this context, prior studies have examined the potential structural benefit of glucosamine therapy using knee radiographs (including digital x-rays) as the outcome of interest. No clear structural improvement, by plain film radiography, was achieved in a prior large, multicenter, NIH-sponsored placebo-controlled trial. In the present study, MR imaging of the knee, instead, was used as the outcome measure. This measure is of much interest, inasmuch as MR imaging may be a more sensitive, and hence, potentially, an earlier measure of osteoarthritis development (and/or progression) that traditional x-rays. Importantly, in this study, a structural benefit by MR imaging was not demonstrated. What the abstract does not address, however, is whether the combination of glucosamine with chondroitin supplementation might achieve a structural benefit, by MR imaging, over placebo therapy. An important strength is the interpretation of the MR images by readers without knowledge of treatment assignment group.

Abstract #629: The Relationship between Weight Maintenance and Incident Radiographic Knee osteoarthritis: The Johnston County Osteoarthritis Project.

Authors:

LM Abbate, J Stevens, TA Schwartz, LF Callahan, JB Renner, CG Helmick and JM Jordan, The University of North Carolina, Chapel Hill, NC and  Centers for Disease Control and Prevention, Atlanta, GA

Background: 

The purpose of this study was to determine whether maintenance of stable body weight is related to a reduction in risk of incident radiographic knee osteoarthritis (rKOA).  Moreover, the authors focused on weight maintenance because of the recognition that weight loss is a difficult goal to achieve. Notably,  were adults who maintain a steady weight to experience a diminished risk of OA, then substantial gain could be achieved at the population level, were more adults to maintain, rather than gain, weight during their lifetime.

Methods:   

The authors examined data from the Johnston County Osteoarthritis Project, a longitudinal study of African-Americans and Whites living in Johnston County, North Carolina. The study participants are aged 45 years and older and were first evaluated at T0, between 1990-1998, then reassessed in follow-up at  T1,  between 1999-2003. There were a total of 1,480 participants in the study.  Weight change was defined as a change from the initial weight, and was coded as a 5-level variable with categories defined as:   ≥5% loss, >3 to <5% loss, ± 3%, >3 to <5% gain, and ≥5% gain.  Indicator variables were used to  contrast between ≥5% loss (weight loss), ± 3% (weight maintenance), and ≥5% (weight gain) with weight gain as the referent. The outcome of interest, rKOA, was defined as Kellgren-Lawrence (K-L) grade of 0 or 1 at T0 and K-L ≥ 2 at T1.   Proportional hazards models with adjustment for the correlation between knees were used to calculate hazard ratios and 95% confidence intervals for the association between weight change and incident rKOA.  All models were adjusted for age, race, sex, height, and the mean of weights from T0 and T1. 

Results:

Among 1,480 individuals, 63.2% were female and 25.9% were African-American. The mean age and BMI of the study participants was 59.4 years and 28.6 kg/m2, respectively.  During a mean follow-up period of 5.9 years, rKOA developed in 415 (14.9%) of 2,788 knees. Of note, 31.4% of the cohort gained weight compared to 32.% who maintained weight and 16.7% who lost weight.  Furthermore, those who maintained stable body weight were no less likely to develop incident rKOA [HR=1.02 (95% CI=0.77, 1.35)] than those who gained weight. However, those study participants who lost weight experienced a reduction in risk [0.71 (0.49, 1.01)] of incident rKOA, of borderline statistical significance. 

Conclusions:    

The authors conclude that weight loss, but not weight maintenance, may be an effective strategy to reduce the risk of incident rKOA. 

Editorial:

The association of body weight, particularly excess body weight, has been a longstanding area of focus in the epidemiologic literature in osteoarthritis, particularly of weight bearing joints (e.g. knee and hip). Moreover, a seminal article from the Framingham Osteoarthritis Study in the mid-1980s demonstrates that weight loss is associated with a ~50% reduction in incidence of knee osteoarthritis. In the present study from the Johnston County Osteoarthritis Project, the authors further investigate the association of weight loss with incident knee OA and also examine the association of weight stability versus weight gain in this context.

Importantly, the study hypothesis was negated, namely that maintenance of stable body weight was not related to a decrease in risk of subsequent knee osteoarthritis compared to those who gained weight. A further comment regarding the study findings is that although the risk of incident knee OA was reduced in those who lost weight, this finding was only of borderline statistical significance.

Abstract #632: Pain in One Knee Increases the Risk of Tibiofemoral Osteoarthritis in the Contralateral Knee: The MOST Study.

Authors:

J Niu, DT Felson, M Nevitt, P Aliabadi, C Lewis, B Sack, J Torner and Y Zhang, Boston University School of Medicine, Boston, MA, UCSF, San Francisco, CA, UAB, Birmingham, AL,  and UIowa, Iowa City, IA

Purpose: 

Several studies have reported that painful non-osteoarthritic knees have an increased risk of radiographic tibiofemoral osteoarthritis (TF OA). As gait usually changes when a person develops knee pain, the authors expected the risk of TF OA would also increase in the contralateral non-painful knee due to unloading of the painful side. This research question was investigated by examining the relationship of knee pain on one side of the body to the risk of developing TF OA on the contralateral side in the Multicenter Osteoarthritis (MOST) Study.

Method:

The investigators examined data from the MOST Study, a NIH-funded longitudinal observational study of subjects who have or are at high risk to develop knee OA. The outcome of interest was radiographic evidence of knee OA. OA status was ascertained at baseline and at 30 months of follow-up using PA & lateral radiographs of each knee. Among knees without whole knee OA (K/L grade <2 and no patellofemoral (PF) OA) at baseline, a knee was defined as having incident TF OA if its K/L grade was ≥2 at 30-month or had undergone knee replacement during follow-up period. Frequent knee pain (FKP, pain on most days of the past 30 days) and severity of knee pain (SKP, WOMAC knee pain subscale, categorized as 0, 1-5, 6-20) were collected at baseline. The authors divided knees into 4 groups based on which side was affected by FKP, and 9 groups based on the side of their SKP. The relation of FKP and SKP categories to the risk of TF OA was examined using logistic regression, with adjustment for age, sex, BMI, history of knee injury and surgery, depression (CES-D≥16), whole knee OA status of the contralateral side, at baseline, while accounting for the correlation between two knees. Results: There were a total of 3094 knees in the study without whole knee OA at baseline, among 1844 subjects who had a mean age of 61.3 years, a mean BMI of  29.6 kg/m2, and who were 57.2% female. Among this group, there were 191 incident cases of TF OA over the 30-month period of follow-up. Compared with those without FKP on both sides, the risk of developing TF OA were substantial (odds ratio of 4.7 (95% CI: 2.9, 7.4) in the ipsilateral knee and elevated in the contralateral knee as well for incident TFOA 2.1 (95% CI: 1.3, 3.2) among persons with unilateral knee pain. The risk of RKOA was also elevated among persons with bilateral knee pain (OR 2.7; 95% CI 1.8, 4.1. A similar pattern emerged when knee pain was quantified based on three levels of severity. In this manner, successively higher levels of pain in the one knee were related to a successively greater increase in risk of developing knee OA in the contralateral knee. Lastly, the findings were essentially unchanged when the analyses were restricted to the knees without any OA feature at baseline.

Conclusion: 

The authors conclude that knees with pain at baseline had a higher risk of developing TF OA during the follow-up period than those study participants without knee pain. Importantly, the presence of pain in one knee was found to predict the development of incident osteoarthritis in the contralateral knee. 

Editorial Comment:

The findings are quite instructive in providing insight about a simple measure of disease activity, namely pain, rather than on a novel biomarker, radiographic marker or detailed patient questionnaire instrument. Simply the presence of pain, in a single knee, was related to an increase in risk of developing both ipsilateral and contralateral radiographic evidence of knee OA. The authors suggest this association may be attributable to a biomechanical change in knee loadingonce unilateral knee OA develops. An alternate explanation is that that the presence of knee pain may be a more sensitive and predictive measure of pending knee OA. The symptom of pain may be present even before overt radiographic evidence of knee OA develops.

AddThis Social Bookmark Button