Why Can Pain Scores Improve Without Cartilage Regrowth on MRI?
Published November 16, 2026
Protocol: 8 steps in this guide
Because pain and cartilage structure are different things, measured in different ways. Knee pain reflects more than cartilage thickness: inflammation, the bone and soft tissues around the joint, muscle strength, activity and expectations all play a part. A pain score is what a patient reports; an MRI score is what a reader measures in an image. The two can move separately. A 2026 meta-analysis of 28 randomized trials of stem cell-based knee injections found improvements in some pain and function scores, while “MRI-based WORMS scores were non-significant, indicating no consistent structural benefit.” Feeling better is real, and it does not by itself show that cartilage grew back.
Pain comes from more than cartilage
Osteoarthritis involves the whole joint. MedlinePlus explains on its osteoarthritis page that as cartilage wears away, “the fragments of the cartilage or the bones rub together,” and that “bony spurs or extra bone may form around the joint.” It adds: “The ligaments and muscles around the joint may become weaker and stiffer.” It also notes that some people “might not have symptoms” even when x-rays or a physical exam show the changes of osteoarthritis. If structure and symptoms can diverge in untreated osteoarthritis, they can also diverge after treatment.
Inflammation, muscles and function
The American Academy of Orthopaedic Surgeons’ page on arthritis of the knee describes an arthritic knee that “may be painful and inflamed,” notes that pain and swelling “may be worse in the morning, or after sitting or resting,” and describes exercises that “help strengthen the muscles in your leg.” A change in inflammation, swelling, strength or how a person moves can change pain without any change in cartilage. This may be one reason exercise and physical therapy can help symptoms in a joint whose cartilage looks the same on imaging.
Patient-reported pain versus imaging measures
Pain scales such as VAS, WOMAC and KOOS ask patients how they feel. MRI scores are assigned by a reader looking at images. They answer different questions, and a study should report them in separate tables rather than combine them into a single claim that a treatment “worked.” How to judge the size of a symptom change is covered in what clinically meaningful improvement means in a joint pain study.
What WORMS measures
The Whole-Organ Magnetic Resonance Imaging Score (WORMS) grades several joint structures on MRI, such as cartilage, bone marrow changes and other features, giving a picture of joint structure rather than of symptoms. MRI can detect relevant tissue changes; the point is that a structural score and a pain score measure different things. Even a better MRI score would not on its own show that the new tissue is healthy cartilage. Cartilage repair is slow to show because, as a 2023 review by Guo and colleagues in the Journal of Tissue Engineering, Regeneration of articular cartilage defects, explains, articular cartilage “has limited capacity for regeneration and self-repair after being damaged because of its low cell density, alymphatic and avascular nature.” That limit is biology, not proof that every treatment fails.
Placebo and natural history
Part of the improvement after any injection comes from contextual factors: expectation, attention and the experience of receiving treatment. A 2025 meta-analysis by Yin and colleagues in Frontiers in Medicine, Contextual effects of mesenchymal stem cell injections for knee osteoarthritis, pooled 8 placebo-controlled trials with 467 patients. Based on low-certainty evidence, contextual factors accounted for about 63% of pain reduction at 6 months and about 50% at 12 months. The authors note that this share bundles natural fluctuation in symptoms, regression to the mean and possible mechanical effects of the injection together with expectation, so it is a group-level estimate, not the part of one patient’s improvement that came from expectation. A placebo response is a real change in how a person feels, not an imaginary one. Natural ups and downs in pain also matter: people who start treatment during a flare can improve as the flare settles. Control groups are explained in the difference between a placebo-controlled trial and an observational study.
An example from knee cell-therapy research
The 2026 meta-analysis cited at the top, by Awad and colleagues in Clinical Rheumatology, shows the split directly:
| Outcome type | Measure | Result across 28 randomized trials |
|---|---|---|
| Symptoms | VAS pain and some KOOS scores | Significantly improved |
| Symptoms | WOMAC, KOOS quality of life, Lequesne Index | No significant difference |
| Structure | WORMS scores on MRI | Not significantly improved |
The authors concluded that the therapies “serve a primarily symptom-modifying rather than structure-modifying role.” The full trial evidence is in what clinical trials show about stem cell injections for knee osteoarthritis, and the regrowth question in whether stem cell products can regrow damaged joint cartilage in humans.
A schematic of discordant outcomes
This is a schematic, not a real patient:
| Panel | Pain score | Knee function | Cartilage on MRI |
|---|---|---|---|
| Before treatment | High | Limited | Thinned |
| 6 months after | Lower | Better | Unchanged |
| Possible explanations | Less inflammation, stronger muscles, contextual effects, natural fluctuation | Activity and strength | No measurable regrowth |
How to read discordant results
When a study or clinic reports better pain but no structural change, read it as a symptom benefit, and ask how it compared with a control group. When it reports a structural change, ask how it was measured and whether symptoms changed too. A claim that an injection “rebuilds cartilage” needs structural evidence from controlled studies, as covered in how to recognize unsupported claims about cartilage regeneration. In an individual patient, it is not possible to say which explanation accounts for an improvement.
Pain scores can improve without cartilage regrowth on MRI because pain reflects inflammation, surrounding tissues, muscles, activity, expectations and natural fluctuation as well as cartilage, so a symptom improvement is a real benefit but not evidence of structural repair.
Cells linked to this one
This guide leads on to
- I9 What Does Clinically Meaningful Improvement Mean in a Joint Pain Study?
- I8 What Is the Difference Between a Placebo-Controlled Trial and an Observational Study?
- F1 What Do Clinical Trials Show About Stem Cell Injections for Knee Osteoarthritis?
- F3 Can Stem Cell Products Regrow Damaged Joint Cartilage in Humans?
- I7 How Do You Recognize Unsupported Claims About Cartilage Regeneration?
Guides that lead here
- H6 For Ankle Cartilage Lesions, Does Adding BMAC Improve Microfracture Outcomes?
- A2 What Is the Difference Between Tissue Repair and Tissue Regeneration?
- F5 What Is the Evidence for Cell-Based Treatment of Avascular Necrosis of the Femoral Head?
- F7 Can Stem Cell Treatments Repair a Torn Meniscus?
- D2 How Does A2M Block Enzymes That Damage Cartilage?
- D5 Can A2M Injections Rebuild Damaged Cartilage?
- F6 What Is Known About Cell-Based Therapy for Degenerative Disc Disease?
Add a remark