Thumb Metacarpophalangeal Joint Arthrodesis

A salvage operation that trades a little motion for a stable, painless column. Thumb MCP arthrodesis eliminates pain at the MCP and converts an unstable link into a rigid strut, at the cost of only the small MCP flexion arc (the CMC and IP provide the dominant thumb motion). It is reserved for situations where motion-preserving options have failed or are not viable. Absolute indications - Chronic ulnar collateral ligament (UCL) insufficiency not amenable to reconstruction — gross instability with volar subluxation during pinch, a failed prior UCL repair, or a failed graft reconstruction.

• Post-traumatic MCP joint destruction with combined arthritis and instability where motion-preserving procedures (arthroplasty, osteotomy) are not viable.
• Rheumatoid Z-collapse deformity with painful MCP volar subluxation as part of the zigzag collapse pattern — fusion stabilises the middle link of the collapse. Relative indications - Primary MCP osteoarthritis (uncommon in isolation — more often part of pan-trapezial arthritis) causing pain and functional loss refractory to non-operative treatment for greater than 6 months.
• MCP instability after infection or tumour resection where joint reconstruction is not possible.
• Failed MCP arthroplasty (silicone or pyrocarbon) with painful instability or loosening — revision to fusion.
• Salvage after failed external fixation or failed dynamic external fixation for complex MCP peri-articular fractures. Contraindications. Absolute: active infection at or adjacent to the MCP joint; an uncorrectable soft-tissue deficit precluding wound closure over the fixation hardware. Relative: symptomatic CMC arthritis that should be addressed first (staged before MCP fusion or simultaneously); pre-existing IP joint arthritis (fusion at MCP may overload the IP — consider whether IP fusion alone would suffice); and a patient whose occupation requires a full MCP flexion arc (some musicians, specialised manual workers) who should be counselled thoroughly. Try non-operative treatment first. A resting thumb spica splint reduces MCP loading and gives symptomatic relief in early disease; static splinting at 10-15 degrees of flexion can reduce pain during activities of daily living. The evidence is largely case series and expert opinion — no RCTs specifically compare splinting with surgery for MCP arthritis. An intra-articular corticosteroid injection gives temporary relief (weeks to months), mainly in inflammatory arthritis; in primary OA the small capsule makes injection inconsistent, though it can serve as a diagnostic test confirming the MCP as the pain source before committing to surgery. Arthroplasty or arthrodesis? MCP arthroplasty (silicone, pyrocarbon, or interposition) is an option for patients who wish to retain MCP motion, but silicone implants carry synovitis, implant fracture and subsidence over time, and pyrocarbon and resurfacing implants have shorter follow-up and higher revision rates than fusion. Arthrodesis gives a predictable, permanent and painless solution at the cost of MCP motion, and the bulk of evidence supports fusion as the gold standard for young, active, high-demand patients. Key decision points - Young, active patient with good bone stock and UCL insufficiency — headless compression screw: the best balance of compression, low profile and no routine hardware removal.
• Rheumatoid patient with osteoporotic bone — dorsal mini-plate: the most rigid fixation in poor-quality bone; accept the higher rate of hardware removal.
• Low-resource or cost-sensitive setting — tension band wiring with K-wires: reliable and inexpensive; plan for hardware removal at 4-6 weeks.
• Concomitant CMC disease — address the CMC first (trapeziectomy and LRTI), then reassess the MCP; some patients no longer need MCP fusion after CMC stabilisation.
• IP joint arthritis present — if mild, proceed with MCP fusion and monitor the IP; if severe, consider IP fusion alone (which eliminates tip pinch) or staged procedures. Consent specifically for nonunion (5-15%), malposition requiring revision, hardware prominence and possible removal (10-20% with a plate), infection (less than 1%), IP joint overload, CMC overload, and the certainty of permanent loss of MCP motion.

The goal is to expose the joint through the dorsal approach, protect the radial digital nerve, prepare two flat bleeding cancellous surfaces, fix the joint at 10-15 degrees of flexion with slight pronation, and splint the thumb in the functional position. The dorsal exposure between EPB and EPL — with the radial digital nerve identified and protected — is the heart of the operation and is laid out step by step below.

View larger

Rehabilitation | Phase | Timing | Immobilisation | Therapy | |-------|--------|----------------|---------| | 1 | 0-4 weeks | Thumb spica cast/splint, MCP in the fusion position, IP free | Active IP flexion/extension from day 1; gentle CMC opposition within the cast; elevation for 48 hours | | 2 | 4-6 weeks | Removable thumb spica splint (after 4-week X-ray shows early bridging) | Gentle IP and CMC motion; light lateral and pulp pinch; scar massage from 6 weeks | | 3 | 6-8 weeks | Splint for heavy tasks only | Progressive pinch and grip strengthening with therapy putty; splint weaning | | 4 | 8-12 weeks | Night splint only | Full activity; non-contact sport; contact sport and heavy manual work at 3-4 months | Sutures are removed at 10-14 days. Radiographs are taken at 4 and 8 weeks to track bridging trabeculae across the fusion site. Most patients return to desk work by 8-12 weeks and heavy manual work by 3-4 months, and driving once they can comfortably grip the wheel and perform an emergency stop (usually 6-8 weeks). If a plate was used, discuss planned removal after confirmed union at 4-6 months. Key rehabilitation principle — IP motion from day 1. The IP joint compensates for the lost MCP arc and must stay supple; a stiff IP after MCP fusion is a poor functional result. Encourage gentle CMC opposition even during immobilisation, and retrain pinch with a hand therapist, since the thumb now works as a single unit from CMC to IP with the MCP fixed. Expected outcomes. Union is achieved in 85-95% across all fixation methods (highest with a plate, lowest with K-wires alone), typically over 8-14 weeks (longer in smokers, diabetics and rheumatoid bone). Pain relief is reported by 85-95%, patient satisfaction by 80-95%, and grip and pinch strength are maintained or improved — particularly where instability was the pre-operative problem. Complications

Functional anatomy. The thumb MCP is a condyloid joint with a biconcave proximal phalanx base articulating with a biconvex metacarpal head. The metacarpal head is wider volarly than dorsally, which contributes to the natural slight flexion of the resting thumb. The collateral ligaments are eccentrically attached — taut in flexion and lax in extension — which is why the UCL is examined with the MCP in flexion. The capsule is thin dorsally (allowing surgical access) and reinforced volarly by the volar plate. Why fusion at 10-15 degrees sacrifices little. The total MCP flexion-extension arc is only 5-10 degrees of extension to 50-60 degrees of flexion, and most thumb function (pinch, grip, opposition) occurs with the MCP between 5 and 20 degrees of flexion. The CMC joint provides the majority of opposition and circumduction, and the IP joint provides tip flexion for fine pinch — so fusing the MCP at 10-15 degrees removes relatively little useful motion. Collateral ligament complex and the unstable MCP. The proper ulnar collateral ligament runs from the ulnar side of the metacarpal head (dorsal tubercle to volar edge) to the base of the proximal phalanx and is the primary restraint to valgus (radial deviation) stress during pinch, preventing volar subluxation of the proximal phalanx. It is tested by applying valgus stress with the MCP in 30 degrees of flexion, which tightens the proper UCL and isolates it from the accessory ligament. A Stener lesion — the UCL torn and displaced superficial to the adductor aponeurosis — cannot heal and requires acute repair; chronic insufficiency after a failed repair is a classic route to MCP fusion. The accessory UCL (from the volar metacarpal neck to the volar plate and ulnar sesamoid) stabilises the volar plate and is taut in extension. The radial collateral ligament restrains varus stress and rarely needs reconstruction in isolation. Extensor tendons and the dorsal interval. The EPL passes over Lister's tubercle at the wrist and crosses dorsally to insert on the distal phalanx base; at the MCP level it is the most ulnar dorsal extensor and forms the ulnar boundary of the approach interval. The EPB inserts on the proximal phalanx base (dorsal and slightly radial) and forms the radial boundary; it may be absent or fused with APL in up to 30% of individuals. APL lies radial to EPB, inserting on the first metacarpal base, and is not routinely encountered unless the incision is extended proximally. The standard dorsal approach proceeds between EPB (radial) and EPL (ulnar), identified proximal to the MCP and developed distally onto the capsule. Neurovascular structures at risk. The radial digital nerve runs volar to the extensor tendons, crossing into the dorsal plane in the proximal wound, where it lies superficially in the subcutaneous fat and is vulnerable during skin-flap elevation — it must be identified and protected before the interval is developed. The ulnar digital nerve lies volar-ulnarly and is not routinely encountered unless dissection extends volarly. Small dorsal sensory branches from the superficial radial nerve supply the MCP dorsum and are transected during the skin incision (warn the patient about a small area of numbness). The volar blood supply is dominant, so avoid excessive dorsal periosteal stripping to preserve cancellous vascularity for fusion healing. Epidemiology and outcomes. Primary MCP osteoarthritis is uncommon in isolation and is more often post-traumatic or part of pan-trapezial disease; rheumatoid disease is a major driver through the Z-collapse pattern. Across published series, thumb MCP arthrodesis reliably eliminates MCP pain and provides a stable platform for pinch and grip, with union in 85-95%, significant pain relief in 85-95%, and patient satisfaction of 80-95%. Fixation-method studies consistently show that all techniques achieve acceptably high union rates and that functional outcomes are determined primarily by fusion position, bone preparation and rehabilitation rather than by the specific device — hardware-related complications (especially plate prominence and removal) are the main differentiator between methods.