aHand Surgery Service, New York University Hospital for Joint Diseases Orthopaedic Institute, Department of Orthopaedic Surgery, The New York University School of Medicine, 301 East 17th Street,
New York, NY 10003, USA bHand Surgery Service, The Institute of Plastic and Reconstructive Surgery, The New York University School of Medicine, 550 1st Avenue, New York, NY 10016, USA
According to ICD-9-CM diagnostic codes in 1998, an estimated 188,000 to 226,000 carpal fractures were treated in emergency rooms across the United States . Although the scaphoid accounts for 62% to 87% of carpal fractures, clinicians should not underestimate the frequency by which the remainder of the carpus is fractured. These injuries are often difficult to diagnose and the treatments rendered inadequate and delayed.
In recent literature, the incidence of carpal fractures in overall hand injuries has ranged from 8% to 19% since 1990 [2-4]. The frequency of such fractures has not changed significantly since Emmet and Breck  reported a 17% incidence in 1958. Nonscaphoid fractures account for 3.2% to 7.7% of these injuries. A breakdown between the proximal and distal row separately demonstrates that fractures of the distal row account for 0.8% to 1.4% of all hand injuries. The hamate is injured most frequently in the distal row, whereas the triquetrum is the second most commonly injured behind the scaphoid with respect to the proximal row [2,3]. Triquetral fractures vary from 4% to 20% with different studies, whereas the remaining carpal bones (trapezium, hamate, capitate, lunate, pisiform, and trapezoid) are less frequently injured, with a range of 0.2% to 3% [2,3,6-9]. The rarity of such fractures offers few outcomes data.
The strategy behind carpal fracture treatment is to make accurate diagnoses, determine the degree of displacement, severity of symptoms, and to
* Corresponding author.
E-mail address: [email protected] (S.K. Lee).
address concomitant injuries. Treatment should be directed not only at the fracture, but possible surrounding associated injuries. Carpal fractures are relatively uncommon, but care must be given to diagnose and treat these injuries appropriately. Misdiagnosed and untreated carpal fractures may lead to nonunion, malunion, avascular necrosis, carpal instability, articular incongruity with resultant osteoarthrosis, neurovascular compression, or late tendon rupture, among other conditions [7-9].
Fractures of the capitate are rare and account for only 1.3% of all carpal fractures . Most of these fractures occur in association with additional carpal pathology, particularly scaphoid fractures; isolated fractures of the capitate comprise only 0.3% of carpal injuries .
Harrigan  reported on the first case of an isolated capitate fracture in 1908. In 1962, Adler and Shaftan  reported on 48 cases of isolated capitate fractures. Of the 16 cases with known treatment, 14 were treated by immobilization and 2 by excision. Results were reported in only 8 cases; five patients had a ''good'' result and three had a ''poor'' result. One of these three underwent capitolunate fusion. Unfortunately, many of these cases had incomplete data, and no data were available regarding the incidence of nonunion.
Since 1962, only 25 cases of isolated capitate fractures have been reported [10,13-27]. These reports emphasize that early diagnosis is important because delayed treatment may lead to avascular necrosis (AVN), nonunion, and post-traumatic arthritis .
The low incidence of isolated capitate fractures is postulated to be because of its anatomical position; the capitate is protected from injury by its surrounding bones, namely the third and fourth metacarpal, hamate, lunate, scaphoid, and trapezoid bones. In addition, these fractures may be underdiagnosed; little, if any, displacement of fracture fragments occur, because of stabilization by intracarpal ligaments [13,28]. The fracture may also be initially missed because of a paucity of symptoms and a radiographically occult fracture.
Studies by Gelberman and colleagues [29-31] demonstrated that the capitate is vulnerable to post-traumatic avascular necrosis because of its blood supply. The capitate has a dorsal blood supply with two to four vessels entering the distal two thirds on its concave surface. These vessels supply the body and head in 67% of specimens. One to three vessels enter from the palmar side. The blood vessels to the head of the capitate originate entirely from the palmar surface in 33% of specimens. This retrograde blood flow (similar to the scaphoid) is believed to place the capitate with a waist fracture at risk for AVN [29-31].
Of the three mechanisms considered to cause isolated fractures of the capitate, the more frequent is a fall on the palm with the wrist extended [12,32]. Biomechanical cadaver studies have demonstrated that the dorsal lip of the radius may strike the capitate with hyperextension . The other causes are axial load or a direct blow over the dorsum of the wrist.
Early diagnosis cannot be overemphasized. Unfortunately, the frequent paucity of symptoms contributes to the possible delay in diagnosis [28,33,34]. If an isolated nondisplaced capitate fracture is missed and not immobilized, the proximal segment may rotate with wrist movements producing AVN or nonunion caused by interruption of the vascularization of the head (proximal pole) .
When a displaced capitate fracture occurs as an isolated injury, plain radiographs are usually diagnostic (Fig. 1A,B); however, nondisplaced isolated capitate fractures may be radiographi-cally occult [13,36,37]. Multiple radiographic studies may be required for diagnosis. PA radial and ulnar deviation views may help make nondis-placed capitate waist fractures visible on plain radiographs. Hopkins and Ammann  found that early diagnosis of a capitate fracture could be obtained with a 99M-Tc-methylene disphosphonate nuclear medicine bone scan and confirmed with CT or MRI. Other authors [13,37-40] have also reported the usefulness of CT and MRI .
Calandruccio and Duncan  reported a case of isolated capitate fracture in which initial plain radiographs were considered normal. Treatment was delayed until the fracture was diagnosed with the use of MRI. The authors prefer using MRI to confirm an occult capitate fracture.
Nondisplaced isolated capitate fractures should be treated with short-arm thumb spica cast immobilization for 6 to 8 weeks . Displaced fractures require anatomic reduction to restore normal carpal kinematics . In a long-term follow-up study of capitate fractures, Rand and coworkers  recommended anatomic reduction (by open technique if necessary) and immobilization until the fracture united. Volk and colleagues  reported an excellent outcome with open reduction and stabilization using a Herbert screw or Kirshner wires (K-wires). Internal fixation by K-wires or Herbert screws has also been reported by others .
The most substantial and under recognized complication of isolated capitate fractures is that of nonunion. Of the 25 cases of isolated capitate fractures reported in the literature since 1962, 14 (56%) developed nonunion [10,18,19,22,23,33,42]. Yoshihara and coworkers  reported 12 cases of nonunion in the literature, and reported the incidence of nonunion amongst isolated capitate fractures as being 19.6%. This percentage took into account the 48 cases reported by Adler and Shaf-tan in 1962  and assumed that none of them developed nonunion, although this was not specifically stated.
Of the 14 cases of nonunion in the literature, the average patient age was 27 years old, ranging from 13 to 54 years. All were diagnosed late, with the average period from injury to definitive diagnosis being 1 year and 11 months (range 3 months to 7 years). The regions of nonunion were the proximal third in 3 cases, the middle third in 10 cases, and the distal third in one case. Of the 14 cases of nonunion, 8 did not receive initial treatment after injury because of missed diagnoses. The other 5 were initially treated with immobilization because of suspected contusion or "sprain," but the fractures failed to unite. The details ofthe remaining 1 case as to initial treatment are unknown. The treatment in 10 cases was cancellous or corticocancellous bone grafting, with or without screw fixation, and observation in 4 cases . Nine out of 10 operated cases obtained union. The other cases did not achieve union. Rico and colleagues  reported that cancellous or corticocancellous bone grafting after correction of the rotation of the fragments
achieved bone union and restored the length of the capitate, but with some reduction in mobility.
The long-term probability of arthritis after an isolated capitate nonunion is unknown and has not been reported; however, 66% of patients who had scaphocapitate syndrome developed post-traumatic arthritis .
AVN of the capitate after an isolated capitate fracture is rare, with only three cases reported in the literature [15,24,43]. Grend and coworkers  reported that a fracture through the capitate jeopardizes the blood supply to the proximal portion of the bone by interference with the intraosseous circulation, thus potentially resulting in AVN.
The high incidence of nonunion of isolated capitate fractures (56%) has not been previously recognized. All cases were associated with late diagnosis, highlighting the importance of early diagnosis and treatment. A patient suspected of having a capitate fracture based on clinical examination and history should undergo further imaging with MRI when initial radiographs are negative. Nondisplaced fractures warrant 6 to 8 weeks in a short-arm thumb spica cast. Displaced fractures require closed versus open reduction and internal fixation (ORIF), with K-wires or headless compression screws, depending on the individual fracture pattern. Box 1 provides diagnosis and treatment guidelines.
Was this article helpful?