Abstract
There is no consensus regarding the aetiology or treatment of carpal boss. Recurrences or carpometacarpal instability is reported after simple resection. Carpometacarpal arthrodesis has been proposed as a primary treatment and after failure of simple resection. This paper reports the results at a mean follow-up of 17 (range 13–28) months of treatment by wedge-shaped joint resection, corticocancellous radial bone graft and arthrodesis with a shape memory staple of seven of 32 patients who presented with a painful carpal boss. The study group included five women and two men of mean age at operation 29 (range 18–40) years. Fusion was achieved in all cases and all patients were pain free at follow-up. No complications occurred and all the patients were pleased with the aesthetic aspect of the treatment.
Rarely symptomatic, carpal boss was initially described as a curiosity without functional impairment (Fiolle, 1931). However, some patients complain of significant pain or catching of the index finger extensors. After the failure of conservative treatment, wedge resection of the carpal boss has become the surgical treatment most commonly performed (Cuono and Watson, 1979; Lenoble and Foucher, 1992). Some authors have recorded good results with this technique. Other publications, however, report recurrences and/or carpometacarpal instability (Citteur et al., 1998; Clarcke et al., 1999; Hazlett, 1992; Hultgren and Lugnegard, 1986).
The alternative surgical approach has been arthrodesis of the second carpometacarpal joint, offered either as a primary treatment or after failure of simple resection (Clarcke et al., 1999). However, the technique of carrying out this arthrodesis is not standardised and a high rate of pseudarthrosis has been reported (Clarcke et al., 1999).
The aim of this prospective study was to describe and to evaluate a technique of second carpometacarpal joint arthrodesis, with the help of a radial bone graft and a shape memory staple osteosynthesis, with the intention of avoiding carpometacarpal joint instability or painful recurrence of the boss.
PATIENTS AND METHODS
During the period October 2002 to October 2005, 32 patients presented to one surgeon with a painful carpal boss. Conservative treatment, including splinting at the night and during heavy manual activities and intralesional corticosteroids injection, was instituted in all cases. Sixteen patients did not respond to conservative treatment, but symptoms were minimal and surgery was not felt to be indicated.
Surgery was, indeed, only proposed after failure of conservative treatment when pain was severe enough to interfere with the patient’s occupation. Neither occasional discomfort nor cosmetic deformity was considered as an indication for surgery. Of the 32 patients, seven patients underwent surgical treatment to seven hands. They included five women and two men of mean age at operation 29 (range 18–40) years. All the patients underwent surgery to the dominant side.
These patients complained of intermittent pain localised to the dorsum of the wrist and metacarpal bases during repetitive, or heavy, manual activities. Although carpal boss is usually described as being located at the base of the second or third metacarpal base, we gained the impression from palpation in our cases that the boss was always situated at the third metacarpal base. Snapping of the extensor tendons of the index finger over the boss was noted in three hands. In all cases, the metacarpal stress test was positive (Fusi et al., 1995). As initially described, a metacarpal stress test is considered positive when pain is exacerbated by distracting the second or third metacarpal while simultaneously attempting to supinate and pronate it with the MCP joint held in flexion (Fusi et al., 1995). In our patients, the stress test was never positive when applied to the second ray, but always positive on application to the third ray. X-ray examination, using the ‘‘carpal boss view’’, which is a lateral view of the hand and wrist with the forearm held in 30° pronation and the wrist maintained in the neutral extension with 20° of ulnar deviation, confirmed the presence of a carpal boss in all seven patients. In these seven patients, we did not see an os styloideum, which is believed to originate from a separate ossification centre and is situated dorsally between the trapezoid, capitate and third metacarpal (Fig 1). Some cases were seen in the patients treated conservatively, but we are unable to report an exact incidence because X-rays were not done systematically when no operation was planned.
Surgical technique
The operation is performed as a day case using axillary block anaesthesia. A short transverse incision is made at the apex of the carpal boss, which, in our experience, was always situated at the third, or middle, carpometacarpal joint. The tendons of the fourth compartment are retracted ulnarwards and the extensor carpi radialis longus and brevis tendons retracted radially (Fig 2). A rongeur is used to resect the carpal boss until the joint space between the third metacarpal and capitate is seen (Fig 3a). A wedge-shaped joint resection is then performed with bone scissors (Fig 3b). A corticocancellous bone graft is harvested from the distal radius through a second transverse incision at Lister’s tubercule and intercalated under slight compression between the capitate and third metacarpal to maintain the normal height of the third ray (Fig 3c). Osteosynthesis was then performed using a shape memory staple (Autograf ATM (Memometal technologies), Bruz, France) placed longitudinally between the capitate and third metacarpal, bridging the bone graft (Figs 3d and 4). This metal alloy staple is conserved at 4 °C until used. The alloys change their mechanical properties on warming. The angle between the body and legs, which is 90° at 4 °C, closes at body temperature, leading to compression. A palmar splint, including the wrist and metacarpals, is worn for 4 weeks. The metacarpophalangeal joints are left free to allow light daily life activities. Strenuous activities and contact sports are avoided for 2 months.
RESULTS
The procedure was performed on seven hands in seven patients. The mean tourniquet time was 50 (range 40–65) minutes. No complications occurred, either at surgery or during follow-up.
The mean follow-up was 17 (range 13–28) months, which is too short to see any long-term repercussions of this arthrodesis. Radiological fusion was obtained by the eighth week in all patients (Fig 5). All patients were relieved of pain at final follow-up. To date no patients have required re-operation and no staples have required removal. The staple has never been palpable, but some deep scar tissue was always palpable and a visible boss due to the staple was present in three cases.
All the patients were pleased with the appearance of the short transverse scars (Fig 6). Even the patients who still had a visible boss were satisfied. However, we have now modified our technique slightly to locate the staple at a deeper level and avoid the postoperative visible boss.
DISCUSSION
There is no consensus regarding the aetiology of carpal boss. Rupture of the dorsal ligament due to trauma, or repetitive strain, congenital predisposition, os styloideum or post-traumatic periostitis have been suggested as aetiological factors (Van der Aa et al., 1999). Whatever the pathophysiology, most authors agree that carpal boss is a degenerative osteoarthritis of the second, or third, carpometacarpal joint, with some degree of instability (Joseph et al., 1981).
The second and third metacarpals are rigidly fixed to the carpus, providing the central cantilever around which the other metacarpals move in a conical manner. This rigidity is usually attributed to the anatomy of the articular surfaces and the dorsal and volar ligaments of these carpometacarpal joints (Joseph et al., 1981). Interestingly, clinical experience with middle ray amputation has shown that the second and fourth metacarpals come together towards the central axis of the hand if the third metacarpal base is entirely removed (Lyall and Elliot, 1996; Sood and Elliot, 2000), suggesting that the third metacarpal base acts like the keystone of an arch. To our knowledge, the relative importance of the different factors in the stability of the carpal arch and each carpometacarpal joint, individually, has never been investigated. The major axis of stress in the normal wrist is in the line of the second and third metacarpals, the capitate and the scapholunate joint. Stability here is of prime importance. Citteur et al. (1998) showed, in a cadaver, that wedge excision of the dorsal part of the third carpometacarpal ligaments significantly increased the passive range of motion at this joint. During removal of a carpal boss, this dorsal ligament between the capitate and the third metacarpal bone, which is incorporated into the capsule, is inevitably cut, increasing the pre-operative state of instability. The normal anatomical relations within the third, or middle, carpometacarpal joint are significantly disturbed and instability occurs on applying weights of only 1 and 2 kg in the laboratory (Citteur et al., 1998). Therefore, from a theoretical point of view, arthrodesis of the involved CMC joint should be a safer procedure than simple wedge resection of the carpal boss.
Most of the series reporting results of simple resection of carpal boss present good results (Artz and Posch, 1973; Cuono and Watson, 1979; Fusi et al., 1995; Lenoble and Foucher, 1992). However, removal of the carpal boss does not always result in a complete relief of symptoms or pain and recurrences of the boss have been reported, with some cases requiring either a second excision or a secondary carpometacarpal joint arthrodesis (Artz and Posch, 1973; Cuono and Watson, 1979; Fusi et al., 1995; Hazlett, 1992; Hultgren and Lugnegard, 1986; Lenoble and Foucher, 1992). Citteur et al. (1998) wrote that they had the ‘‘impression that many patients treated by wedge excision remain symptomatic’’. Fusi et al. (1995) reported six revision operations in a series of 60 patients, i.e. a 10% re-operation rate in reviewed cases. Clarcke et al. (1999) reviewed 46 patients with 48 painful carpal bosses. Ten of these patients were referred from another institution after previous, unsuccessful surgery. They operated on 40 patients, including 18 simple excisions and 22 carpometacarpal joint arthrodeses. They achieved only 50% complete resolution of symptoms with patient satisfaction in both groups. In this study, osteosynthesis was performed with various materials, viz. K-wires, Herbert bone screws, plates, staples, tension band and, even, without any implanted foreign material but immobilisation only in three cases. Without giving details about the rate of non-union, the authors stated that they encountered some difficulties in achieving sound union. There is very little information in the literature about carpometacarpal arthrodesis and no consensus about the type of bone graft and the fixation that should be used (Carroll and Carlson, 1989; Clarcke et al., 1999; Joseph et al., 1981).
We think that pain from carpal boss is due to degenerative changes secondary to some instability or conflicting mobility at the third carpometacarpal level. Simple resection is only able to remove osteoarthritic changes but is not addressing the main problem. To reduce the incidence of recurrence of pain after simple resection, we believe it is worthwhile to fuse the third carpometacarpal joint primarily when surgery is indicated. We have never had to fuse the second carpometacarpal joint and have, so far, been unable to demonstrate a carpal boss at the second ray level. Our follow-up and the series are too short to give definite conclusions, but the experience reported in this paper supports this view.