More than 200 patients with osteoarthritis (OA) of the knee got no more benefit from intra-articular corticosteroid injections than from lidocaine shots in a randomised cross-over trial, and neither treatment was all that helpful, a researcher reported at the American College of Rheumatology’s annual meeting last week.
Medpage reports that on average, participants in the 221-patient trial saw improvements of only a few points on the Knee Osteoarthritis Outcome Score (KOOS) system as well as with the Patient-Reported Outcomes Measurement Information System (PROMIS), irrespective of which agent was given, according to Joshua Baker, MD, MSCE, of the University of Pennsylvania in Philadelphia.
With a 10-point threshold for KOOS prespecified as the minimal clinically important difference, his group judged that neither steroids nor lidocaine made any real impact on participants’ symptoms.
Some numerical advantages were seen for the steroid shots, Baker noted. But leaving statistical significance aside and treating them as real differences, calculations of the number needed to treat to achieve a clinically important improvement yielded values of 22 when steroids were given as the first injection, and 24 when they were administered after a lidocaine shot.
Exclusion criteria were relatively minimal, leaving the sample almost like an “all comers” trial for patients with knee OA who might receive steroid shots in normal practice.
Thus, Baker said, the results provided “no evidence that corticosteroids are effective in this real-world population”.
And even if it were possible to identify those patients more likely to see improvement from this treatment, “the benefits are still likely to be small”, he said.
This is just the latest salvo in a long-running dispute over the true effectiveness of steroid injections for knee OA, and it’s certainly not the first to show little to no benefit when compared with placebo or something close to it (like lidocaine).
True, some studies have found benefits. One reported earlier this year that a single steroid injection seemed to reduce patients’ need for painkillers for many months afterward; another from 2023 showed that a polymer-coated formulation of fluticasone helped 50% of patients see meaningful relief lasting 14 weeks.
But others have indicated that steroid shots can be harmful, and hardly any research has demonstrated pain relief lasting more than a few months.
Baker observed, too, that results from placebo-controlled trials, which typically select participants carefully, don’t always translate to routine practice with patients who come in with comorbidities and other complicating factors.
For the current study, Baker and colleagues actually conducted two studies with the same 221-patient sample, conducted at four Veterans Affairs clinical sites across the US.
Participants were randomised to receive incentives in a game-based format to increase their daily step counts, or just to wear a Fitbit with no special goal. Those results were to be reported separately.
These same individuals also were given two knee injections, one with 40 mg methylprednisone and the other with lidocaine, in random order.
Results after each shot were taken 12 weeks later, and there was a month-long washout before the second injection. Patients, treating clinicians, and evaluators were all blinded to which injections were given.
Baker said there was no indication that participants’ assignments in the incentive trial influenced responses to the injections.
Inclusion criteria were ages 40-80, X-ray findings indicating a Kellgren-Lawrence grade of at least 1 for knee OA, and the treating clinician’s opinion that steroid injections were warranted. Participants could have received them previously.
The chief exclusions were for contraindications to intra-articular injections, a recent acute increase in symptom severity, severe mobility problems, or evidence of inflammatory or crystalline arthritis. They also needed to have a smartphone, as participants recorded and reported their symptom severity by themselves remotely.
Mean overall KOOS score at baseline was 43. Baseline PROMIS scores averaged about 44 for pain intensity, 53 for pain behaviour, and 55 for pain interference. Mean patient age was about 64 and 16% were women. Body mass index values averaged about 33.
More than three-quarters had previous knee injections.
The maximum improvement in mean KOOS score came at week 4, both for steroids and for lidocaine. Changes averaged 3-4 points with steroids versus about 1 point with lidocaine.
By week 12, KOOS scores had returned close to baseline levels with both treatments. Baker and colleagues calculated a mean differential effect from the steroids as the first injection, relative to lidocaine, with a Cohen’s d value of 0.10 (P=0.23). The same comparison for the second injection produced a d value of 0.02 (P=0.29).
Results were very similar for the PROMIS subscales, with slightly greater numerical benefit from steroids, but with differences of only a point or two relative to lidocaine. With baseline values in the 50s, these differences were not clinically meaningful.
Baker’s group also looked at how subgroups fared. Women seemed to respond better to lidocaine than to the steroid, whereas the opposite was true for men (although again, the absolute benefit was small).
Participants with more severe symptoms, as reflected by lower baseline KOOS scores and Kellgren-Lawrence grades of 2 or higher, also seemed to get more relief from steroids than did those with less severity.
Limitations listed by Baker included that a higher steroid dose, or restricting the sample to patients with at least moderate symptom severity, could have altered the findings.
As well, since the study was conducted among military veterans, it might not be readily generalisable to the general knee OA population.
See more from MedicalBrief archives:
Study finds two treatments osteoarthritis ineffective
Excessive knee jabs linked to industry's marketing payouts
Low-dose radiation re-emerging as osteoarthritis therapy
Osteoarthritis to affect 1bn people by 2050, predicts global study