CHAPTER SYNOPSIS:
Thumb carpometacarpal (CMC) arthritis is a common problem that can cause significant pain and dysfunction, often in a relatively young and active population. Conservative treatments including nonsteroidal anti-inflammatory drugs (NSAIDs), splints, and cortisone injections can usually provide significant relief. Because the natural history of thumb CMC arthritis is one of progressive joint degeneration, surgical treatment is a common and often successful eventuality. A number of surgical treatment options exist including arthroscopic debridement, osteotomy, arthrodesis, implant arthroplasty, and a variety of interpositional arthroplasties with or without ligament reconstruction.
IMPORTANT POINTS:
Conservative treatment can often provide prolonged satisfactory pain relief.
Complete trapeziectomy is the sine qua non of definitive surgical treatment.
Despite a plethora of ligament reconstructions, none has proved itself distinctly superior.
The role of arthroscopy and new implant materials in the surgical treatment of thumb CMC arthritis continues to evolve.
CLINICAL/SURGICAL PEARLS AND PITFALLS:
The trapezium must be excised entirely; often the trapezium is fragmented and the trapezial space must be carefully cleared.
The radial sensory branch and radial artery must be carefully avoided during a dorsal exposure.
Following interpositional graft, the joint capsule must be carefully closed to avoid graft extrusion.
Look for loose bodies at the metacarpal bases.
Avoid injuring the flexor carpi radialis (FCR) tendon during trapezium removal.
HISTORY/INTRODUCTION
The thumb CMC joint is a common site for symptomatic osteoarthritis (OA), particularly among middle-aged women. Since Kellgren and Lawrence’s landmark 1958 study, the first large-scale documentation of the prevalence of OA in a sample population, including significant rates of degenerative thumb CMC arthritis, many population studies have reinforced the importance of this disease. The radiologic prevalence among postmenopausal women has been found to be 33% in one study of British women, with 33% of them having symptoms. Similarly, the Framingham Study found that after reviewing right hand radiographs of 453 of its study participants, 26% of men and 42% of women had evidence of thumb CMC OA. The average age was 54 years for men and 53 years for women; no data on symptoms were collected. Factors contributing to arthrosis include the size and congruency of the trapeziometacarpal joint, hormonal factors, environmental conditions, and heredity. Although radiographic changes are well described, their severity does not necessarily correlate with clinical symptoms.
Several anatomic characteristics of the thumb CMC joint are important to consider in the etiology and progression of arthritis. The thumb CMC joint is a saddle joint, with the axes of the metacarpal and trapezial “saddles” 90 degrees opposed to one another. Although these axes lie in the radial/ulnar and volar/dorsal planes, the position of thumb opposition necessitates pronation of the metacarpal. This leads to incongruity of the joint surfaces with increased contact stresses on the dorsal–radial and volar–ulnar aspects of the joint. Wear pattern studies corroborate these biomechanical findings with increased wear in the dorsal–radial and volar regions as the disease progresses. Studies have further demonstrated differences between men and women in topographic characteristics of the joint. In studies of arthritic joints, the thumb CMC joint in females is less congruent, is flatter, and has thinner cartilage surfaces. Joint incongruence leads to smaller cartilage contact areas and therefore higher contact pressures.
Ligamentous anatomy also plays a role in thumb CMC arthritis. Multiple ligaments stabilize the CMC joint, although five are commonly referred to in the literature as primary stabilizers—the anterior oblique or “beak” ligament, the ulnar collateral ligament, the first intermetacarpal ligament, the posterior oblique ligament, and the dorsoradial ligament. Although cadaveric studies differ in their assessment of the contribution of each structure, the emphasis on the anterior oblique ligament in CMC stability is consistent. An intracapsular ligament, the anterior oblique ligament resists metacarpal extension, abduction, and pronation, and its degeneration has been linked to progressive CMC osteoarthritis. Several studies have also noted the particular contribution of the dorsoradial ligament, in some cases finding a greater contribution to dorsal subluxation during loaded opposition in cadaveric models. The primary culprit not withstanding, ligamentous attenuation eventually contributes to the dorsal metacarpal subluxation characteristic of advanced disease.
The increased prevalence of thumb CMC arthritis among women leads naturally to the theory that hormonal factors play a role in pathogenesis. The effect of estrogen or relaxin on ligamentous tissue has been most widely studied with respect to anterior cruciate ligament (ACL) injury, and although female hormones seem to have effects on the tensile properties of the ACL, a direct correlation with injury remains difficult to establish. Despite the progress made with respect to the ACL, the fact that the ligamentous attenuation accompanying thumb CMC arthritis is the result of chronic degeneration rather than an acute event suggests that the role of hormones may not be exactly the same in both cases. Further study is required to elucidate the long-term clinical effect of hormones on ligamentous instability and thumb CMC arthritis. Hypermobility is similarly a theoretic cause of thumb CMC arthritis, possibly related to hormonal factors directly. Case-control studies have reached contradictory conclusions, finding both protective and detrimental effects of hypermobility on the development of thumb CMC arthritis.
Typically, patients will present with the chronic onset of pain at the base of the thumb metacarpal often with radiation into the thenar eminence. Pain is exacerbated by pinching or grasping activities such as turning a car key or opening a jar, and functional disability in addition to pain may manifest itself as a sense of weakness with activities using the hand. Often patients will have tried anti-inflammatory medications or off-the-shelf splints with some degree of temporary relief.
On examination the thumb may appear normal or dorsal subluxation of the proximal metacarpal may be evident. Tenderness may be elicited along the first metacarpal or at the CMC joint. Dorsal tenderness may seem similar to that of DeQuervain’s tenosynovitis, although the distinction can often be made with careful palpation of the first dorsal compartment and a positive Finkelstein’s test result. The two conditions may coexist, at times with a primary CMC condition irritating the overlying tendon sheath. In an equivocal case, diagnostic injection may clarify the diagnosis. The thumb CMC grind test will elicit pain with axial loading of the thumb CMC joint and rotation and translation of the first metacarpal ( Fig. 15-1 ). Forced adduction of the thumb metacarpal also often elicits a painful response. Classically, with more advanced disease an adduction/flexion contracture will limit metacarpal range of motion. Compensatory hyperextension of the metacarpophalangeal (MCP) joint may be evidence of an attempt to facilitate grasp from this adducted position.
Other causes of pain at or radiating to the base of the thumb include flexor carpi radialis tendonitis; carpal tunnel syndrome; metacarpal or scaphoid fracture or arthritis of the MCP, scaphotrapeziotrapezoidal (STT), or radiocarpal joints. Usually history and examination findings will rule out causes other than arthritic joints, of which the thumb CMC is most common. If history and physical examination suggest thumb CMC arthritis, we obtain posteroanterior (PA), lateral, and Bett’s views using fluoroscan. The Bett’s view, alternatively termed Gedda or Clement’s view, allows visualization of the STT joint and enhances the reliability of thumb CMC arthritis classification. Plain x-rays are sufficient to diagnose, stage, and treat thumb CMC arthritis ( Fig. 15-2 ). We do not obtain further studies—computed tomography, magnetic resonance imaging, or bone scan—unless other diagnoses are suspected (i.e., occult fracture, avascular necrosis, tumor, vascular malformation).
CLASSIFICATION SYSTEM
Eaton established the most commonly used classification system for thumb CMC degenerative disease. This radiographic system consists of four stages:
- 1
Normal x-rays or thumb CMC joint widening
- 2
Narrowed joint space; osteophytes or loose bodies <2 mm
- 3
Further narrowing; subchondral sclerosis; osteophytes or loose bodies >2 mm
- 4
Stage III with the addition of scaphotrapezial joint changes
The Eaton classification system has been tested for reliability with PA and lateral views and found to have moderate intrarater and interrater consistency among hand surgeons. Addition of the Bett’s view to standard views enhances the reliability of classification.
INDICATIONS
With few exceptions, all patients with an initial diagnosis should be treated conservatively. Splints, NSAIDs, and activity modification for 1 to 2 months will often be successful. If results are not adequate, a cortisone injection in the thumb CMC joint is performed and usually provides at least some duration of relief. We inject 1 mL of 6 mg/mL Celestone (betamethasone) mixed with 0.5 mL of 2% lidocaine using a 25-gauge needle. Mixing lidocaine in with the injection can help confirm the diagnosis if immediate pain reduction is noted. We ensure intraarticular injection by palpating the base of the metacarpal dorsally with the examiner’s thumb and feeling the sulcus of the CMC joint between the first and third dorsal compartments as the metacarpal is flexed and extended. Metacarpal subluxation or osteophytes can obscure easy access to the joint, but in our experience an intraarticular injection can usually be obtained by angling the needle slightly distally and if necessary gently manipulating the syringe to adjust the starting point and angle and ensure smooth entry into the joint. Signs of a successful injection include ease of flow of the syringe and brief but significant discomfort as the joint is distended. Often patients present initially having had a prior injection, sometimes unsuccessfully. If indicated, we typically repeat the injection and ask the patient how the pain compares. In our experience, if the patient notes significantly more pain with injection, they will respond better than after the previous attempt, which we presume may have been extraarticular. The number of injections provided varies depending on the patient’s desire to avoid surgery, but we advise patients that the marginal benefit and duration of relief of each injection is likely to diminish with each subsequent injection. We allow at least 6 months between injections and seldom provide more than two or three.
A combination of conservative measures can provide a surprisingly long period over which symptoms are managed to the patient’s satisfaction. Recent meta-analyses support the efficacy of both intraarticular corticosteroids and viscosupplementation in the knee. Studies on the effectiveness of corticosteroids and viscosupplementation in the thumb CMC joint are less numerous and less robust. The only randomized, blinded trial of corticosteroid injection versus saline was performed in patients with moderate to severe arthritis and found no difference between the two groups in terms of subjective pain or objective measures at 24 weeks. Several European studies have evaluated the effectiveness of hyaluronic acid injection both alone and in comparison with corticosteroid injection, finding generally favorable results with hyaluronic acid. Randomized, blinded studies of corticosteroid injection, particularly for early-stage arthritis, and randomized studies comparing hyaluronic acid to placebo are necessary to further elucidate the efficacy of these two treatments. If pain persists and a patient is unable to function at a satisfactory level, surgery is the next option. Although indications for various procedures overlap, in the context of the Eaton stages some general principles prevail. Basic arthroscopic procedures are appropriate for Stage I or early Stage II disease in which the articular surface is preserved, although it may also be helpful as a diagnostic tool in later stages. Partial trapeziectomy should be undertaken with caution in Stage III disease when the likelihood of pain derived from the scaphotrapezial or trapeziotrapezoid joints becomes greater. Stage IV disease necessitates excision of the trapezium and several interpositional arthroplasty options.
SURGICAL TECHNIQUE
Trapeziectomy
Trapeziectomy was first described by Gervis in 1949 and remains the essential step of surgical treatment for thumb CMC joint disease. Surgery can be performed under general anesthesia or using a forearm bier block or axillary block. The arm should be exsanguinated and a tourniquet should be used. A dorsal approach may be used consisting of a longitudinal incision through the skin and identification of the interval between the extensor pollicis brevis (EPB) and abductor pollicis longus (APL) tendons ( Fig. 15-3 ). The joint capsule is then incised directly over the trapeziometacarpal joint. The superficial sensory branch of the radial nerve should be avoided. Full-thickness flaps are raised, the trapezium is exposed, and soft tissue capsular attachments to the trapezium are incised. A rongeur is used to excise the trapezium while taking care to avoid the radial artery as it courses dorsally toward the interval between the heads of the first dorsal interosseous muscle. With an adequate grip on the trapezium, continuous rotation of the instrument may allow en bloc excision of the entire trapezium. If not, the trapezium can be split into two or three pieces with an osteotome and more easily removed. The space should be examined for remaining pieces of bone that should be sharply excised. The joint capsule should be closed with interrupted sutures prior to skin closure.
Interposition Graft
Interposition grafting as described by Froimson may be performed through either a dorsal or volar approach using the proximally detached flexor carpi radialis (FCR) as a soft-tissue spacer to prevent proximal migration of the thumb metacarpal. After excising the trapezium, the FCR tendon is released proximally at the musculotendinous junction and pulled distally out through the skin incision. A figure of eight suture is secured to capsular tissue deep on the ulnar-most aspect of the space vacated by the trapeziectomy. The same suture is then woven into the tendon creating an “anchovy” that is compressed into a tight ball by pulling the suture tight. Finally, the suture is tied to itself deep in the wrist, tethering the anchovy within the trapezial space. The superficial capsular tissue should be carefully closed with absorbable suture to help maintain stability of the joint and prevent extrusion of the interposed tendon. K-wire fixation of the first to the second metacarpal will maintain joint position in the early healing period. We recommend K-wire immobilization for 5 weeks following trapeziectomy, interposition arthroplasty, ligament reconstruction with tendon interposition (LRTI), or any of the variant suspension procedures. Temporary fixation allows the soft tissues to heal in a stable position and theoretically helps to prevent subsidence of the metacarpal.
Ligament Reconstruction Tendon Interposition
Ligament reconstruction with tendon interposition, originally described by Burton and Pellegrini, seeks to reconstruct the stabilizing function of the many ligaments surrounding the CMC joint ( Fig. 15-4 ). As with tendon interposition, the approach may be dorsal or palmar. The palmar approach avoids the sensory branch of the radial nerve and improves visualization of the FCR tendon, whereas the dorsal approach improves visualization of the radial artery and may avoid painful volar scarring. A trapeziectomy is performed and the FCR, in whole or split in half, is harvested proximally in the forearm and freed up distally to its insertion at the base of the index metacarpal before being brought through the skin incision. We always use the entire FCR tendon. Having performed the procedure both ways, we have noticed no difference in complaints of wrist flexion weakness or pain whether one half or the entire tendon is harvested. Furthermore, constructing the interposition graft using the whole tendon allows for significantly more bulk and avoids the technical difficulties inherent in splitting the tendon. A drill hole is made through the dorsal cortex of the proximal thumb metacarpal leaving an approximately 5-mm cortical bridge at the base. A small curette is used to extend the drill hole proximally in the medullary canal of the metacarpal until it communicates through the base of the metacarpal into the trapezial space. The FCR tendon is then passed into the base of the metacarpal and out through the dorsal cortex. The tendon is tensioned and sutured to itself, thus tethering the first metacarpal to the second. In a similar fashion to tendon interposition, a suture is placed deep in the trapezial space and the remaining FCR tendon is condensed and interposed ( Fig. 15-5 ). Likewise, careful capsular closure and K-wire fixation of the first to the second metacarpal should be performed.
Postoperatively, patients are immobilized for 5 to 6 weeks in a well-padded short arm thumb spica splint with the interphalangeal (IP) joint left free. The K-wire is removed at 6 weeks and CMC/MP range of motion exercises are begun. Strengthening exercises may be started at 12 weeks postoperatively, at which point patients may begin to resume full activities as tolerated. Patients are advised to expect improvement in strength and comfort for up to one year postoperatively.
Other Suspension Procedures
A variety of suspension procedures using the abductor pollicis longus and FCR tendons have been described. Thompson described a variation of the LRTI that uses the APL to perform a similar function. After performing a trapeziectomy, a slip of the APL, usually dorsal, is detached at its musculotendinous junction. Drill holes are made in the first and second metacarpals. The hole in the first metacarpal is similar to that of the LRTI. The second metacarpal drill hole is made from trapezial articular surface of the metacarpal through the dorsal cortex at the insertion of the extensor carpi radialis longus (ECRL). The distally based APL tendon slip is then woven through the first metacarpal from dorsal cortex to base, then through the second metacarpal from base to dorsal cortex. The APL tendon is then tensioned and sutured to the ECRL tendon at its insertion.
Sirotakova has published a variation of the Thompson APL suspension arthroplasty. In this variation a slip of the APL is brought into the trapezial space and wrapped around the FCR from dorsal to volar, tensioned and sutured to the FCR. The remaining APL slip is threaded around the ECRL from volar to dorsal at its insertion and similarly tensioned and sutured, creating a sling to suspend the first metacarpal.
Viegas described a split FCR tendon suspension in which half of the FCR tendon based distally is passed through a drill hole in the first metacarpal from volar to dorsal, passed through a split APL, tensioned distally, and tenodesed to the APL. The remaining FCR tendon is threaded around the intact half of the FCR at its insertion, rolled into a spacer, and sutured into the trapezial space.
Following the various suspension procedures, closure and K-wiring are performed in similar fashion to the LRTI.
Arthroscopy
Arthroscopy of the thumb CMC joint may be used in diagnosis of the degree of degeneration of the articular surfaces and evaluation of the integrity of the stabilizing ligaments, primarily the anterior oblique, or beak, ligament. In addition, synovectomy or thermal shrinkage of capsular tissue may be performed, although the benefit of these procedures has not been well established. One case series has reported the results of a number of arthroscopic procedures, which included capsular shrinkage and distal trapezial hemi or complete resection. With a varied combination of procedures 90% of 22 patients had good or excellent outcomes. Partial trapieziectomy has also been described with placement of autograft, allograft, or xenograft tissue as an interpositional graft.
As with open procedures, thumb CMC arthroscopy can be performed with general or regional anesthesia. The upper extremity is prepped and draped and a tourniquet is applied. Either vertical or horizontal traction devices can be applied with 5 to 10 pounds of traction, allowing easy access to the dorsal wrist. After identifying and marking the tendinous structures traversing the area of incision—FCR, APL, EPB, EPL (extensor pollicis longus)—a narrow-gauge needle can be used to provisionally access the thumb CMC joint and verify correct portal placement. The radial portal is placed just volar to the APL tendon, and the ulnar portal is placed just ulnar to the EPB tendon ( Fig. 15-6 ). Two to 3 mL of normal saline are injected into the joint. A 1.9-mm 30-degree arthroscope is used for joint visualization. Multiple procedures may then be carried out. Soft tissue debridement can be performed with a shaver by shifting the arthroscope and working portals back and forth to access the entire joint. In addition, shrinkage capsulorrhaphy may be performed in an effort to create an inflammatory reaction with subsequent healing to denser tissue, although the benefit of this technique has not been proven. For more progressive disease, distal trapezial excision is performed with a burr and an interposition graft can be placed using a hemostat inserted into one portal and out through the opposite to facilitate passage of the graft. Postoperatively, patients are placed in a thumb spica splint in an abducted position for up to 4 weeks.
