ATLANTA -- Forget glucosamine. Oral NSAIDs are fine. Hyaluronic acid injections, meh. Just say no to platelet-rich plasma and stem cell injections. Definitely recommend tai chi. Opioids? Let's talk.
That's a sample of new guidelines on hip, knee, and hand osteoarthritis (OA) management from the American College of Rheumatology (ACR), now being finalized but previewed here at the group's annual meeting.
Scheduled for release next month after journal editors give their final blessing, the guidelines represent a major update from the ACR's .
Leaders of the update effort stressed that while they relied heavily on hard evidence in the published literature, they weren't bound by it. So in some cases, a lack of evidence for benefit from a particular therapy could mean the unpublished draft recommends strongly against it, or that it remains neutral or even offers cautious endorsement.
Often, that decision was influenced by a treatment's cost, potential harms, and input from patients who participated in the guideline development.
As with most modern guidelines, the effort , resulting in a report completed in the summer of 2018. A patient panel was convened at that point to provide that perspective, two members of which were on the voting panel as well.
The process was structured through the so-called PICO framework: Patient/Intervention/Comparator/Outcome. Sharon Kolasinski, MD, of the University of Pennsylvania, who walked ACR meeting attendees through the draft, gave the following example of how PICO works: In Patients with knee OA, what are the benefits and harms of Interventions with NSAIDs, Compared to no treatment, for an Outcome of pain measured on the standard WOMAC scale? This process was followed for 44 interventions assessed during the guideline update.
Treatments were rated into four categories: strong recommendations for or against, or "conditional" recommendations for or against. Therapies got strong recommendations when there was good quality evidence that was convincing in one direction or the other, the recommendation would apply to nearly all patients, and when the panel anticipated that few patients would go against the recommendation. Conditional recommendations were made when the balance of benefits and harms was close, or the evidence quality was poor; the recommendations apply to most patients but many might reasonably reject them -- and shared decision-making is essential.
The guideline also gives separate recommendations for OA in the knee, hip, and hand, although in many cases they don't differ.
Kolasinski helpfully spelled out where the new draft differs from the 2012 version. Some recommendations that were previously conditional are now strong, she said, thanks to the accumulation of additional data. These include tai chi (for knee and hip OA), "self-efficacy" and "self-management," topical and oral NSAIDs (knee and hand), and intraarticular steroid injections for the hip and knee.
Also upgraded were balance exercise and duloxetine (Cymbalta), which got no recommendation in 2012 but now get "conditionally-for" status.
One reversal from 2012 involves topical capsaicin for knee OA. Then, the recommendation was conditionally against it, but the new guideline now gives it a conditional endorsement. It remains disrecommended for hand OA, though.
And another reversal appears likely to be a flashpoint for rheumatologists and patients. The 2012 guideline conditionally recommended hyaluronic acid injections in knee OA, but now it will get a conditional thumbs-down, and for hand OA as well.
Kolasinski and co-presenter Tuhina Neogi, MD, of Boston University, both acknowledged that this will surprise and perhaps rankle many rheumatologists who believe hyaluronic acid is effective and safe. But Neogi said the hard evidence that these injections genuinely help is weak -- that the studies showing the most benefit were also the smallest and with the most bias, whereas larger and better-designed studies mostly showed little to no benefit.
One audience member nevertheless warned that many clinicians will not take kindly to a recommendation against a therapy they have been using successfully and that "has survived in the marketplace for 20-30 years," though he conceded that he frequently offers treatments whose benefits may stem primarily from a placebo effect.
And the new guideline greatly expands the roster of interventions it recommends strongly against. These include:
- Transcutaneous electrical nerve stimulation (hip and knee)
- Glucosamine
- Chondroitin (hip and knee)
- Hyaluronic acid injections (hip)
- Hydroxychloroquine
- Methotrexate
- Platelet-rich plasma injections (hip, knee)
- Stem cell injections (hip, knee)
- Tumor necrosis factor inhibitors
- Interleukin-1 receptor antagonists
Opioids other than tramadol receive a "conditional against" grade in the new guideline. Neogi explained that clinicians on the panel were ready to recommend strongly against opioids, because of evidence indicating they aren't very effective against chronic OA pain and come with well-publicized risks, but the patient panel urged that they be kept "as an option" for patients who don't respond to any other treatment. (The same was true for acetaminophen, for the same reasons, Neogi said.)
Tramadol, on the other hand, got a "conditional-for" recommendation for OA generally, on the strength of its relative safety.
Another feature of the new guideline is an increased emphasis on what is now called "educational, behavioral, psychosocial, mind-body, physical approaches." These were formerly called (more efficiently, perhaps) "nonpharmacological" interventions, but as Neogi explained, the guideline panel thought that word was taken by patients to mean "no treatment," which is not at all the case.
In fact, the guideline puts it at the top of first-line approaches for OA of all three joints, with exercise in various forms as most critical, along with strong patient engagement in all aspects of their OA treatment.
And the guideline emphasizes what it calls, somewhat euphemistically, "contextual effects," to some degree a code word for the placebo effects. Panel leaders noted that how interventions are framed for patients makes a big difference in whether they agree to them and whether patients benefit from them. A treatment the clinician recommends with enthusiasm typically yields better results than one presented neutrally, a hallmark of the placebo effect, although that term was barely spoken during the 90-minute presentation and discussion.