When steroid injections stop working but surgery feels too far

The pattern tends to be consistent: a sharp pain at the base of the thumb when gripping a steering wheel, unscrewing a jar lid, or pinching a pen. A steroid injection brings four or five months of relief — then the pain returns, often a little sooner than the time before. A second injection helps less. At that point, most patients are told to consider surgery.

The problem is that neither corticosteroids nor hyaluronic acid address what is actually happening inside the joint — a focal area of cartilage thinning at the trapeziometacarpal (TMC) joint, the small saddle joint at the thumb's base. A 2024 meta-analysis of ten randomised controlled trials found no meaningful difference between those two injection types for pain, grip strength, or function, and a 2025 double-blind trial of 90 patients confirmed that a single PRP injection performed no better than saline. These are palliative treatments; once the cartilage loss reaches a certain point, their effect fades.

Surgery — trapeziectomy, the removal of the trapezium bone — remains the UK gold standard for advanced disease, but it is irreversible and carries a substantial rehabilitation period. Many patients with moderate-grade damage have not reached that threshold and are not yet ready for it.

This article examines the evidence for ChondroFiller injection — a CE-marked, ultrasound-guided, outpatient injectable collagen scaffold — as a biologically active option in that intermediate space, covering its mechanism, the clinical data, and which patients are and are not suitable candidates.

What happens inside the TMC joint — and why cartilage loss matters here

The TMC joint sits where the first metacarpal of the thumb meets the trapezium bone at the wrist. Its saddle shape allows the thumb to rotate, oppose the fingers, and move through the wide arc that makes gripping, pinching, and turning possible. That same mobility comes at a cost: the joint surfaces bear forces several times the weight of the hand during ordinary activities, leaving the cartilage here unusually vulnerable to progressive wear.

Eaton-Littler grading describes four stages of that deterioration. Grades I and II represent early wear — mild joint-space narrowing with small osteophytes and cartilage still largely intact. Grade III shows more substantial narrowing, focal areas of exposed subchondral bone, and larger osteophytes. Grade IV is end-stage bone-on-bone contact. ChondroFiller injection is indicated at grades I through III, where a focal cartilage deficit exists but sufficient structural integrity remains for a biological scaffold to function; grade IV disease, where no residual cartilage surface remains to anchor and surround the scaffold, is a contraindication.

The joint's small size and depth also shape the options available. Rigid surgical implants cannot easily reach the TMC space without significant dissection, whereas a fine-needle injectable approach under ultrasound guidance can access it directly as an outpatient procedure.

How ChondroFiller injection works — gelation, scaffold, and cell recruitment

Injectable as a liquid, ChondroFiller begins to transform the moment it contacts the joint environment. Within approximately three to five minutes, it undergoes in-situ gelation — shifting from a flowable solution into a self-adherent gel that forms fibrin bonds with the surrounding cartilage surface and subchondral bone beneath it. No sutures or mechanical fixation are needed; the scaffold anchors itself through those chemical bonds and, in the confined space of the TMC joint, the anatomy itself helps retain it in position.

The material contains no donor cells — it is entirely acellular. Its biological role is to act as a recruitment framework: a structured collagen matrix that signals the body's own progenitor cells to migrate into the damaged area and begin producing repair tissue. This process is sometimes called matrix-induced chondrogenesis. A 2025 ex vivo osteochondral study provided direct evidence of this recruitment, demonstrating a 2.4-fold increase in DNA content within ChondroFiller-treated tissue explants by day 14 — confirming that cells do enter the scaffold. That finding establishes the biological rationale; it does not imply guaranteed cartilage regrowth, and outcomes in individual patients depend on the grade and extent of the defect as well as the surrounding tissue quality.

From the patient's perspective, the procedure is straightforward. It is performed as an outpatient appointment under local anaesthetic, with image guidance confirming accurate needle placement into the joint. There is no general anaesthesia, no theatre booking, and no overnight stay. After the injection, a short period of rest and reduced loading of the thumb is recommended — the biomechanical evidence suggests that early cyclic stress on the joint before the gel has fully set may limit the scaffold's initial stability, so avoiding forceful grip and pinch in the first days after treatment is a practical and clinically relevant precaution.

What the clinical evidence shows for the thumb base specifically

Two prospective Italian studies by Corain and colleagues form the primary evidence base at this joint.

The more recent, published in 2025 (NCT06881186), enrolled 64 patients across all four Eaton-Littler grades and delivered a single percutaneous injection under ultrasound guidance. All four outcome measures — NRS pain, DASH function, Jamar grip, and Pinch strength — improved significantly (P<0.001), sustained over two-plus years with no adverse events. The most substantial gains occurred between two and six months. Importantly, the pattern of benefit divided by disease stage: patients with earlier damage (grades I–II) recovered more grip and pinch strength, while those with more advanced disease (grades III–IV) showed greater functional improvement on the DASH score. That distinction matters when setting expectations before treatment.

The 2023 Corain study (43 patients, fluoroscopic guidance) added a structural dimension. Post-treatment MRI showed reduced bone-marrow oedema, diminished periarticular effusion, and visible joint-space widening — confirming that the response extends beyond symptom relief to measurable change within the joint itself.

The case for a scaffold-based approach becomes clearer against the comparator data. A 2024 meta-analysis of 10 RCTs (673 patients) found no meaningful difference in pain, grip, or pinch between corticosteroid and hyaluronic acid or PRP for TMC OA. A 2025 double-blind RCT (90 patients) confirmed that PRP performed no better than saline at six months (p=0.22). These findings do not discredit palliative injections for earlier presentations, but they mark the therapeutic gap that an injectable biological scaffold targets.

Adjacent-joint data offer supporting context. A 2025 wrist study showed significantly better cartilage quality at follow-up arthroscopy after ChondroFiller injection (median Outerbridge score 1.5 versus 3), and hip arthroscopy data from 26 patients followed over three to five years returned 17 of 21 evaluable patients with good-to-excellent outcomes — with patients who had severe pre-existing OA faring poorly, a grade-dependent pattern consistent with the TMC findings.

No randomised controlled trial comparing ChondroFiller injection directly with corticosteroid, hyaluronic acid, or trapeziectomy at the TMC joint yet exists, and long-term comparative data remain limited. The available studies are uncontrolled prospective series — notable for their consistency and absence of adverse events, but a gap the current evidence base has not yet filled.

Who is likely to benefit — and who is not a suitable candidate

Candidacy rests on two overlapping questions: how much cartilage remains, and whether the surrounding joint architecture can still support a regenerative scaffold.

Patients likely to benefit

The strongest candidates are those with focal TMC cartilage damage at Eaton-Littler grades I to III — meaning some cartilage remains, the surrounding borders are largely intact, and the joint has not collapsed into end-stage bone-on-bone contact. Symptoms should be persistent despite a trial of conservative management (splinting, physiotherapy, oral analgesia, or earlier palliative injections), yet not so advanced that surgery has become unavoidable. This is the intermediate group for whom standard steroid and hyaluronic acid injections have ceased to offer meaningful relief but who do not meet the threshold for trapeziectomy.

Expectations should be calibrated honestly at consultation. Both prospective Corain studies confirm that meaningful improvement across pain and function is achievable, but the character of that improvement varies with how much damage is already present — a point the evidence makes clearly, and one that a pre-treatment conversation should address individually rather than through broad reassurance.

Patients who are not suitable

End-stage grade IV disease with no cartilage reserve — bone grinding directly on bone — is a clinical contraindication. Without viable surrounding tissue for the scaffold to adhere to and cells to migrate from, the biological mechanism has no foundation to work with. Trapeziectomy remains the appropriate pathway for these patients.

Additional contraindications include active joint infection, known allergy to collagen-based products, and any clinical situation where precise injection volume cannot be controlled. This last point has practical weight: wrist data show that overfilling causes fibrous tissue formation rather than repair tissue, which means accurate volume delivery by an experienced clinician is a technical requirement, not a precaution that can be set aside in the interest of thoroughness.

A biomechanical in-vitro study also underlines that early loading before the gel has fully set carries risk — patients who cannot or will not protect the joint in the days immediately after injection are likely to achieve a worse result.

Clinical imaging — typically ultrasound to guide the injection and MRI to characterise the defect and surrounding cartilage — is the starting point for determining whether any of this applies to an individual patient.

Getting assessed in London — costs, access, and next steps

Taken together, the evidence places ChondroFiller injection in a genuinely useful but still-developing position for TMC osteoarthritis. The functional and structural outcomes across two prospective studies are consistent and encouraging; the absence of adverse events over two-plus years is notable. What is missing is a randomised controlled trial against the existing palliative options, and long-term head-to-head data versus trapeziectomy. Patients considering this treatment deserve both sides of that picture before they take the next step.

On access and cost, the position is straightforward and worth knowing early. ChondroFiller injection is available in the UK on a private basis only — it is not NHS-funded, and standard private medical insurance policies do not typically cover it. At London Cartilage Clinic, a single-session assessment and treatment package starts at approximately £3,000, inclusive of consultation, ultrasound, the product, the injection, and a six-week follow-up.

Assessment at any centre should involve clinical examination, imaging review, and a candidacy conversation whose purpose is to determine whether the defect is focal, whether sufficient surrounding cartilage remains, and whether expectations align with what the evidence actually shows. At London Cartilage Clinic on Harley Street, that evaluation sits within the clinical scope of Professor Paul Y. F. Lee, whose practice focuses on cartilage pathology and image-guided joint procedures. Patients based outside London may find the MSK Doctors group's Lincolnshire and Grantham sites a more practical point of access for the same pathway.

For those who want to explore whether a ChondroFiller injection assessment is appropriate for their situation, further information is available at londoncartilage.com.

1. [1] Efficacy of Intra-Articular Corticosteroid Injection for TMC Osteoarthritis: Systematic Review and Meta-Analysis. (2024). https://doi.org/10.1016/j.jhsa.2024.02.001 https://doi.org/10.1016/j.jhsa.2024.02.001
2. [2] Platelet-Rich Plasma Injection for Painful Trapeziometacarpal Osteoarthritis. (2025). https://doi.org/10.2106/JBJS.OA.25.00027 https://doi.org/10.2106/JBJS.OA.25.00027
3. [3] The Use of Collagen-Based Filler for Trapeziometacarpal Osteoarthritis: Long-Term Follow-Up and Future Applications (PMC). (2025). https://pmc.ncbi.nlm.nih.gov/articles/PMC12213533/ https://pmc.ncbi.nlm.nih.gov/articles/PMC12213533/
4. [4] The Use of Collagen-Based Filler for Trapeziometacarpal Osteoarthritis: Long-Term Follow-Up and Future Applications. (2025). https://doi.org/10.1177/19476035251354926 https://doi.org/10.1177/19476035251354926