Efficacy of Doxycycline Treatment for Knee Osteoarthritis

Preliminary evidence from experimental animal models has suggested that doxycycline may be effective in preventing or slowing the progression of osteoarthritis (OA), perhaps through its ability to modulate the activity of cartilage-degrading matrix metalloproteases (MMPs). Despite this encouraging animal evidence, no studies have investigated doxycycline in humans with, or at risk for, OA. Here, Brandt et al (Arthritis Rheum 2005;52(7): 2015) examine the potential for doxycycline to slow the progression of unilateral knee OA in obese women.


Obese women 45 to 64 years of age with radiographic evidence of moderate (K/L grade 2 or 3) unilateral knee OA were randomized to receive doxycycline 100 mg twice daily or matched placebo for 30 months. Adherence was assessed via electronic monitors affixed to the study drug containers. The primary outcome measure was the change in joint space width (JSW) in the medial tibiofemoral compartment for both the index and contralateral knees. JSW was measured via fluoroscopically positioned semi-flexed radiographs of the knees, obtained at baseline, 16 and 30 months. Supine lateral knee radiographs and skyline (patellofemoral) radiographs were also obtained at these time points.

Continuation of all prescription and/or over-the-counter medications was permitted throughout the trial. However, NSAIDs and analgesics were required to be washed out prior to each 6 month clinical assessment. 6 month clinical assessments included patient-reported pain and disability in both the index and contralateral knees, including the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) questionnaire and rating of knee pain after a 50-foot walk test.

Results: 431 subjects were randomized to receive either doxycycline (n=218) or placebo (n=213). 81% of randomized subjects were Caucasian with a mean age of 55 years and a mean BMI of 36.7 kg/m2. Baseline knee pain scores were low, with a mean WOMAC pain score of 10.8 (scale 5-25). 60% of subjects were taking NSAIDs, 30% acetaminophen, and only 7% narcotic analgesics for knee pain. 26% of subjects were not taking any analgesics for knee pain. 24% of patients randomized to either group took glucosamine (with or without chondroitin) at some point in the trial. Demographic variables, knee pain and function scores, and analgesic medications used were balanced between the doxycycline and placebo groups.

307 subjects (71%) completed 30 months of study medication. Subjects who discontinued early had significantly higher mean baseline knee pain and function scores. 60% of the patients who discontinued early returned for 30-month radiographic assessments, which were analyzed with those who completed the protocol with adjustment for differences in periods of therapeutic coverage.

Index Knee Contralateral Knee
Baseline JSW
3.64+1.15 3.91+0.90
3.61+1.19 3.96+0.86
P value 0.74 0.554
Mean change in JSW
at 16 mos

0.15+0.42 0.23+0.45
0.24+0.54 0.23+0.59
P value 0.027 0.504
Mean change in JSW
at 30 mos

0.30+6.0 0.47+0.70
0.45+0.70 0.41+0.67
P value 0.017 0.317

Doxycycline appeared to have no effect on progression of joint space narrowing in the contralateral knee, even in subjects with the smallest difference in JSW at baseline between the index knee and contralateral knee.

Subjects in the placebo group were significantly more likely to report > 20% increase in pain in the index knee between 6 month visits compared to the doxycycline group (mean inter-visit increase in WOMAC pain score >20%: 30.2% vs. 23.8%, respectively; p=0.004). However, doxycycline had no significant effect on inter-visit knee pain in the contralateral knee compared to placebo. The use of NSAIDs, acetaminophen, other analgesics, glucosamine, and vitamins C and D was balanced between groups and appeared to have little direct impact on JSW in either group. Treatment related adverse events were uncommon, and rarely led to discontinuation of the study medication.


In obese women with moderate unilateral knee OA, daily oral doxycycline treatment was associated with a significantly lower rate of joint space narrowing compared to placebo. The effect was evident at 16 months of treatment and was sustained to at least 30 months. Doxycycline was not associated with a decrease in the rate of joint space narrowing in the contralateral knee.

Editorial Comments:

To date, there has been little convincing evidence from randomized clinical trials to support the viability of any pharmacotherapeutic as an effective DMOAD (disease modifying osteoarthritis drug). Accordingly, the results presented here are both novel and convincing in demonstrating the ability of doxycycline to slow radiographic progression in the index knee.

It is somewhat confusing, however, why doxycycline would have no effect on the contralateral knee even in patients with very similar JSW (and presumably similar knee cartilage volumes) between index and contralateral knees. Some of the discrepancy may be related to the population studied; unilateral moderate knee OA with minimal or no symmetry in the opposite knee. This group may contain a disproportionate number of subjects with asymmetric unrecognized ligamentous/meniscal injuries or unbalanced biomechanical forces across the index knee and may not generalize to other OA populations. In addition, a threshold level of disease activity may be required for doxycycline to be effective.

Again, this is somewhat confusing as mechanistically one would expect MMPs to be active in the earliest phases of cartilage breakdown in OA and that MMP inhibition with doxycycline would be most effective at preventing knee OA in assymptomatic individuals rather than showing the greatest effect in moderate disease.

These points of contention indicate that additional investigation is required to establish the appropriate role of doxycycline in clinical OA management. In addition, the optimal duration of doxycycline therapy remains to be established. Presumably, continuous therapy would be required considering that local MMP production would likely resume if suppression with doxycycline were to abate. However, if proven effective and generalizable, the overall safety and tolerability of doxycycline make it an attractive agent in a disorder that is severely lacking in effective therapies.

Arthritis Center

About Arthritis Center

Founded in 1998, the Arthritis Center at Johns Hopkins is dedicated to providing quality education to patients and healthcare providers alike.