By Dr. Geoffrey Modest:
An article just came out comparing total knee replacement surgery (TKR) with nonsurgical management for knee osteoarthritis (OA) –see N Engl J Med 2015;373:1597, along with an editorial (N Engl J Med 2015;373:1668). 670,000 TKRs were done annually in the US in 2012 (with dramatic increase over time) and at a cost of $36.1 billion. This is the first RCT comparing TKR to nonsurgical management, and follows on the heels (?knees) of a study showing that arthroscopic surgery was no better than sham surgery or exercise, though had more adverse events (see https://stg-blogs.bmj.com/bmjebmspotlight/2015/08/05/primary-care-corner-with-geoffrey-modest-md-arthroscopic-surgery-for-knee-oa/ ).
Details of the study:
- 100 patients (62% female; mean 66 yo; BMI 32; Kellgren-Lawrence score of 4 in 46%, 3 in 42% — and score of >=2 being definite OA and 4 being the most severe; KOOS4 score of 48 — Knee Injury and Osteroarthritis Outcome Score with scale 0-100 and includes the 4 fields of pain, symptoms, Activities of Daily Living (ADL), and quality of life; 62% used regular pain meds; and timed up-and-go test measuring time to get up from a chair, walk 10 feet, return and sit down was 9 seconds)
- Randomized to:
- Unilateral TKR followed by 12 weeks of nonsurgical treatment
- Just nonsurgical treatment (delivered by physiotherapists and dietitians, consisting of: exercise to improve functional alignment and stability of the legs; education by two 1-hour sessions on OA, treatments and self-help strategies; dietary advice for those with BMI>25 to attempt 5% weight loss; use of insoles, individually fitted with medial arch support; and pain meds with acetaminophen 1gm qid, ibuprofen 400 tid and pantoprazole 20mg daily.)
- Results (at 12-month follow-up):
- 13 (26%) patients in the nonsurgical group underwent TKR during the 12 months of the study (mean of 6.9 months)
- Comparing TKR to nonsurgical group
- Less weight loss (increase of 0.1 kg vs loss of 2.6 kg)
- Timed up-and-go test decreased 2.4 vs 1.2 seconds
- KOOS subscale score on pain improved by 34.8 vs 17.2 points (KOOS improvement of 15 points is significant)
- KOOS quality of life score improved 38.2 vs 17.8
- KOOS ADL score improved by 30.0 vs 17.6
- And, overall intention-to-treat analysis: KOOS4score improved by 32.5 vs 16.0 points
- The TKR group had more serious adverse events (24 vs 6): 3 DVTs, 1 femoral fracture, 1 deep infection, 3 with stiffness requiring manipulation under anesthesia.
For recommendations regarding nonpharmacologic management of knee (and hip) OA, EULAR (European League Against Rheumatism, which it turns out is not a medieval cult) came out with a series of recommendations, many similar to what was done in this study (see knee arthritis nonpharm recs AnnRheumDis2013 in dropbox, or Ann Rheum Dis 2013;72:1125–1135).
So, a few issues:
- TKR did improve pain more than the nonsurgical management, though it is important to note that 68% in the nonsurgical group (vs 85% with TKR) did have significant improvement in pain scores (>15%) after 1 year
- TKR does not always work: about 20% of patients have significant pain 6 months later in other studies
- TKR clearly and rather expectedly does have significant adverse events in other studies: overall around a 0.5-1% mortality within 90 days; up to 1% risk of VTE, infections, periprosthetic fractures
- Unclear if these results can be extrapolated to mean sustained benefits over time for nonsurgical therapy
- My sense, as with many of these types of decisions, is that it is really important to individualize the decisions: how much functional impairment is caused by the knee pain? (I would refer the patient earlier for surgery if the pain is debilitating and interferes significantly with important functions). Is the patient willing/able to adhere to a rigorous nonsurgical program (though, rehab after TKR is also quite rigorous and difficult, and I have had a couple of patients who were worse off after surgery because they could not do the rehab and developed contractures/unable to walk). But given the individual variabilities of responses, to me this study reinforces the benefits of suggesting a multidisciplinary nonsurgical approach first, including PT, nutrition counseling if overweight, and podiatric assessment. In fact, the results of this study fit in nicely with my recent blog on hip OA, finding, that there was significant clinical improvement with empathy and reassurance (as much as with PT).