Knee osteoarthritis: treatment, GLP-1s and BPC-157

Complete guide: BPC-157 and TB-500: a complete guide to evidence and safety

What is knee osteoarthritis and how many people have it?

Knee osteoarthritis (ICD-10 code M17) is a degenerative, painful disease of the joint that involves the cartilage and also the subchondral bone, the synovium and the surrounding soft tissues. It is the most common form of osteoarthritis: in the Global Burden of Disease analysis, the knee was the most common site².

A meta-analysis of 88 population-based studies with 10,081,952 participants estimated a global prevalence of 16.0% (95% confidence interval, 14.3 to 17.8) in people aged 15 and over and 22.9% in those aged 40 and over, which is about 654.1 million people aged 40 or older with knee osteoarthritis in 2020¹. Incidence was 203 per 10,000 person-years in people aged 20 and over, and prevalence in women was 1.69 times that in men¹.

The Global Burden of Disease 2021 analysis estimated that 595 million people had osteoarthritis (at any site) in 2020, 7.6% of the world population, and projected that knee cases will increase by 74.9% by 2050 compared with 2020². A high body mass index contributed to 20.4% of the osteoarthritis burden, and the authors stress that there is no effective cure².

Symptoms and how knee osteoarthritis is diagnosed

The central symptom is pain that comes with activity and eases with rest, often with brief stiffness, crepitus and reduced range of motion. The NICE guideline (NG226) says to diagnose osteoarthritis clinically, without imaging, in people aged 45 or over with activity-related joint pain (among other criteria), and not to use imaging routinely unless there are atypical features or features that suggest another diagnosis⁴.

The 1986 American College of Rheumatology classification criteria combine history, examination, laboratory and radiographic data in algorithms, intended to separate osteoarthritis from other causes of knee pain such as rheumatoid arthritis³. NICE keeps imaging for cases with atypical features⁴.

Standard treatment for knee osteoarthritis today

The three main guidelines agree on the basics. The 2019 ACR and Arthritis Foundation guideline made strong recommendations for exercise, weight loss in people with overweight or obesity, self-management programs, tai chi, cane use, tibiofemoral bracing, topical nonsteroidal anti-inflammatory drugs (NSAIDs) for the knee, oral NSAIDs and intra-articular glucocorticoid injections⁵. The 2019 OARSI guidelines define arthritis education and structured land-based exercise programs, with or without dietary weight management, as core treatments for the knee, and rank topical NSAIDs as level 1A⁶.

Exercise has the best evidence base. A Cochrane review of 54 studies found, with high-quality evidence from 44 trials (3,537 participants), that exercise reduced pain (standardized mean difference -0.49), equivalent to about 12 points on a 0 to 100 scale, and improved physical function by about 10 points⁷.

In people with overweight, the IDEA trial assigned 454 adults aged 55 and over to diet, exercise or both for 18 months: weight loss was 11.4% with diet plus exercise, 9.5% with diet alone and 2.0% with exercise alone, and the diet-plus-exercise group had less pain and better function than the diet-only group (pain 3.6 versus 4.8 on a 0 to 20 scale)⁸.

Among drugs, a network meta-analysis found that all treatments except acetaminophen produced a clinically significant improvement in pain from baseline, and that intra-articular treatments outperformed NSAIDs, possibly because of the placebo effect of the injection itself⁹. NICE advises considering an oral NSAID when topical ones fail, with gastroprotection, and intra-articular corticosteroids for short-term relief; it says not to offer glucosamine, strong opioids, acupuncture, hyaluronan injections or arthroscopic lavage or debridement⁴. In a 2-year trial of 140 patients, intra-articular triamcinolone every 3 months was associated with more cartilage loss than saline (-0.21 mm versus -0.10 mm) and no less pain¹⁰. When disease is advanced, knee replacement remains, which the Global Burden of Disease analysis describes as highly effective for late-stage disease².

What peptide research has shown in knee osteoarthritis

Semaglutide (GLP-1 receptor agonist). The STEP 9 trial, funded by Novo Nordisk, was a 68-week, double-blind, placebo-controlled study at 61 sites in 11 countries. It enrolled 407 adults with obesity (mean BMI 40.3), clinical and radiographic moderate knee osteoarthritis and at least moderate pain, who received semaglutide 2.4 mg weekly or placebo, plus counseling on physical activity and a reduced-calorie diet¹¹. Body weight changed by -13.7% with semaglutide and -3.2% with placebo, and the WOMAC pain score (0 to 100 scale) fell by 41.7 points versus 27.5 (baseline mean 70.9), a difference of about 14 points¹¹. SF-36 physical function improved by 12.0 points versus 6.5¹¹.

This needs careful reading. The drug's contribution was observed alongside a much larger weight loss, so the trial does not separate the effect on pain from the effect of losing weight, and it did not measure cartilage structure¹¹. The FDA-approved label for semaglutide (Wegovy) includes reducing cardiovascular risk, reducing excess body weight and metabolic dysfunction-associated steatohepatitis (MASH) with fibrosis, but not osteoarthritis or knee pain¹⁵.

Liraglutide. Another GLP-1 receptor agonist did not show the same effect on pain. In a 52-week trial of 156 people with overweight or obesity and knee osteoarthritis who had lost more than 5% of body weight on a prior diet, liraglutide 3 mg produced a weight difference of 3.9 kg versus placebo (-2.8 kg versus +1.2 kg), but there was no difference in KOOS pain (0.4 versus -0.6 points; difference 0.9; 95% confidence interval, -3.9 to 5.7)¹². Gastrointestinal events occurred in 50.2% versus 39.2%¹². The article has a published erratum from 2025¹³.

Retatrutide (GIP, GLP-1 and glucagon receptor agonist). It is an investigational agent. In December 2025, Eli Lilly announced topline results from TRIUMPH-4, a 68-week phase 3 trial of 445 adults with obesity or overweight and knee osteoarthritis, randomized 1:1:1 to retatrutide 9 mg, 12 mg or placebo: weight fell by up to 28.7% and WOMAC pain fell by up to 4.5 points (75.8%) versus 2.4 points (40.3%) with placebo¹⁴. This is a company announcement, not a peer-reviewed paper; I found no peer-reviewed publication as of writing¹⁴.

BPC-157. It is a synthetic pentadecapeptide (derived from a gastric protein) that is not approved for human use. Animal studies suggest it may promote the outgrowth and migration of tendon cells and speed healing of a transected Achilles tendon in rats¹⁶¹⁷, but none of those models is knee osteoarthritis. A 2025 systematic review found 36 studies, 35 preclinical and 1 clinical, and concluded the evidence was level IV and V²⁰. The clinical study is a retrospective chart review of 17 patients at one clinic, 16 of whom answered a phone call about knee pain of mixed causes; 14 of 16 reported relief, with no function or quality-of-life measures¹⁹. Reviews summarize that same work differently (11 of 12 improved versus 7 of 12 with relief lasting over 6 months), which illustrates how fragile it is¹⁹²⁰.

A 2026 systematic review identified only three human studies of BPC-157 (30 people in total), judged all of them to be of poor methodological quality, rated overall certainty as very low and concluded it should not be used clinically in humans outside ethically approved, regulated, high-quality trials²¹. More in BPC-157: from Rogan's podcast to an FDA vote and in the BPC-157 and TB-500 guide.

TB-500. It is a synthetic version of an active region of thymosin beta-4, not the whole protein²³; see TB-500 is not thymosin beta-4. In a study of 32 rats with a transected Achilles tendon, TB-500 improved maximum load and several histological scores, and combining it with BPC-157 gave no added benefit¹⁸. I found no human trial in knee osteoarthritis.

Safety: adverse effects and precautions

The FDA label for semaglutide (Wegovy) carries a boxed warning for thyroid C-cell tumors: it causes these tumors in rodents and it is unknown whether it does so in humans; it is contraindicated with a personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia type 2, and with prior hypersensitivity¹⁵. Common reactions (5% or more) in the 2.4 mg weight-reduction trials include nausea (44%), diarrhea, vomiting, constipation, abdominal pain, headache, fatigue, dyspepsia and dizziness¹⁵. In STEP 9, serious adverse events were similar in the two groups and discontinuation for adverse events was 6.7% with semaglutide and 3.0% with placebo, mostly for gastrointestinal reasons¹¹. On muscle and bone during weight loss, see GLP-1s: muscle, rebound, eyes and pancreas.

In the TRIUMPH-4 announcement, discontinuation for adverse events was 12.2% with 9 mg, 18.2% with 12 mg and 4.0% with placebo, and dysesthesia occurred in 8.8% and 20.9% versus 0.7%¹⁴.

For BPC-157 and TB-500, the FDA lists them among bulk drug substances nominated for compounding that were later withdrawn, and says it identified no, or only limited, safety information for the proposed routes of administration of BPC-157, and no human exposure data for the thymosin beta-4 fragment; it warns of immunogenicity risk and peptide-related impurities²². The 2025 review found no clinical safety data and stressed the risks of unregulated manufacturing and contamination²⁰.

Summary by agent

Frequently asked questions

Is any peptide approved for knee osteoarthritis? No. Semaglutide has FDA-approved indications for weight, cardiovascular risk and MASH, but knee osteoarthritis is not on its label¹⁵.

What did semaglutide show in knee osteoarthritis? In STEP 9, in adults with obesity, WOMAC pain fell by 41.7 points with semaglutide and 27.5 with placebo at 68 weeks, and body weight fell by 13.7% versus 3.2%¹¹.

Is the pain relief from the drug or from losing weight? The trial cannot separate them. Weight loss through diet and exercise already improves pain (IDEA), and liraglutide lowered weight without reducing pain in another trial, so the link is not automatic⁸¹¹¹².

Does BPC-157 repair cartilage? It has not been shown in humans. The evidence is animal studies of tendon and a small retrospective series of knee pain of mixed causes; a 2026 review rated the certainty as very low¹⁷¹⁹²¹.

What do guidelines recommend first? Education, structured exercise and weight loss if there is overweight, with topical NSAIDs as the first drug for the knee⁴⁵⁶.

Has retatrutide been shown to reduce knee pain? There is a preliminary company announcement from a phase 3 trial of 445 participants, and I found no peer-reviewed publication as of writing¹⁴.