Two Treatments That Don’t Work for Osteoarthritis

WASHINGTON — If you’re looking for nonsurgical osteoarthritis (OA) treatments with fewer side effects than ordinary pain relievers, two randomized trials presented here with negative results should at least narrow your search.

In one, colchicine proved to be no better than placebo for pain relief or patient-reported functional improvement in OA of the knee. The other, testing a cocktail of herbal supplements that other studies had suggested are beneficial, likewise found no advantage over placebo in hand OA.

Both studies were presented at the American College of Rheumatology (ACR) annual meeting.

No medications now available for OA do more than relieve pain and other symptoms. Medications such as non-steroidal anti-inflammatory drugs (NSAIDs) and opioids come with significant side effects with chronic use, and injectables such as steroids and hyaluronic acid must be repeated periodically with diminishing effectiveness. In no case do they stop the underlying disease process.

Colchicine Trial

Understanding of that process has evolved over time, said Jonathan Samuels, MD, of NYU Langone Health in New York City, in an ACR presentation. Until recently it was believed to be simply a degenerative process in which joint components break down with age. But it now appears to be “a disease of low-grade chronic inflammation,” he said. Progression comes with effusions and synovitis, and various immune system components such as leukocytes show activation and cytokine release increases — as seen in rheumatic joint diseases. Thus, agents that block these processes ought to prove helpful in OA.

With this picture in mind, his group believed that colchicine would be worth a trial. The drug had been in use for a long time, and has a variety of immunomodulatory effects. Samuels listed inhibition of macrophages and neutrophils and of leukocyte-endothelium adhesion, as well as reduced expression of key interleukin species. Colchicine “has many potential targets in knee OA,” he said.

He noted that the current CLOAK trial was not the first to test the drug in OA: The 2018 OLKOA study also failed to show a benefit. But patients in that study were allowed to use NSAIDs, thus becoming a potential confounding influence, Samuels said.

CLOAK enrolled 120 patients, randomizing them in equal numbers to daily colchicine or placebo for 12 weeks. Patients needed to be older than age 40, not severely obese, with knee OA rated as grade 2 or 3 on the Kellgren-Lawrence scale. They also had to promise not to use NSAIDs, intra-articular injections, supplements, acupuncture, or any other new pain medications. Patients also diagnosed with other rheumatic diseases were excluded, as were those with diabetes or renal or hepatic impairment.

Mean patient age was 67 and about one-third were men. Patient-reported average knee pain stood at about 6 on a 10-point scale at baseline. A fluke of the randomization was that nearly half of the placebo group had effusions, versus just 32% of those assigned to colchicine, but otherwise the groups were well-matched, Samuels said.

That probably made no difference. Every outcome measure showed that both groups improved nearly equally — indeed, the placebo group enjoyed numerical advantages for most.

The analysis focused on change from baseline in Knee Injury and OA Outcome Score. The placebo group showed a decrease of about 1.4 points, compared with 1.0 in the colchicine group. Mean effusion size also shrank to nearly the same degree in both groups. Subgroup analyses examining only patients completing the full 12 weeks or those with more severe baseline disease by various measures (C-reactive protein level, Kellgren-Lawrence grade, baseline pain, etc.) also failed to identify a group benefitting more from colchicine than placebo. Even within the colchicine group, patients with more severe disease showed no greater improvement than those with milder symptoms.

Supplement Trial

Similar disappointment beset the herbal supplement study reported by Xiaoqian Liu, MD, of the University of Sydney, in patients with OA of the hand.

A wide range of herbal products have been touted over time as effective in hand OA, but studies — albeit of low quality, Liu emphasized — had suggested that four particular ones were better than the rest. These are:

• Curcumin (a component of the spice turmeric)
• Pine bark extract
• Methylsulfonylmethane (MSM), a sulfur compound included in many over-the-counter nutritional supplements
• Boswellia serrata extract, also known as Indian frankincense

For the trial, these four substances were formulated into two separate capsules than patients took multiple times daily. Daily doses were 168 mg for curcumin, 100 mg for pine bark extract, 1,500 mg for MSM, and 250 mg for B. serrata extract. The primary outcome was change from baseline in a 100-point, patient-reported hand pain scale. Some functional assessments were conducted as well.

Much of the study was conducted online, with only occasional in-person contact (such as when otherwise eligible patients had no recent hand x-ray, and thus had one performed in the clinic). Capsules were sent by mail and patients reported their pain and functional capabilities through their devices. This was among the limitations Liu cited, in that “a good level of technology literacy and skill was required.”

The study enrolled 106 patients in Australia who were at least age 40, and who rated their baseline pain at ≥40 and ≤90 on a visual analogue (VAS) scale, and with Kellgren-Lawrence grade of at least 2. Mean age was 66 and a large majority were women. Overweight was common but obesity was not. Baseline hand pain averaged 60 on the 100-point VAS scale.

As in the colchicine trial, both the active-treatment and placebo group showed similar degrees of improvement in pain after 12 weeks, with a numerical advantage for placebo (-8.6 points for the supplement cocktail and -14.6 points with placebo) that fell short of statistical significance.

Scores on other measures including the Functional Index of Hand OA, the Patient’s Global Assessment, and a quality-of-life scale, all improved to similar degrees in the two study arms.

No major safety issues were seen in either study, but the supplement capsules weren’t tolerated by all patients — four assigned to the active treatment quit prematurely because of gastrointestinal symptoms. Diarrhea-like symptoms were also slightly more common with colchicine than placebo in that trial.

Limitations

Both studies also had to overcome hurdles related to the COVID-19 pandemic that might have influenced the results. In the colchicine trial, Samuels said, the drug supplier lost access to the original 0.8-mg dose and had to substitute a dose of 0.6 mg partway through. Liu mentioned that the pandemic had interfered to some degree, but there had also been a major wildfire affecting many participants; these events in tandem might have had both mental and physical consequences for patients, she said.

As well, both studies faced the same major limitation: they only ran for 12 weeks. It’s possible that longer treatment and follow-up periods would show a true benefit for the respective treatments. That was particularly true for the colchicine study: Samuels and colleagues hypothesized that the drug would inhibit pathogenic immune processes related to disease progression, but no such effect could be expected in 12 weeks.