Complications of Intercarpal Fusions

INTRODUCTION

∗ Dr. Gaston, Dr. Lourie, and I would like to dedicate our chapter to Louis G. Bayne, MD, and Waldo E. Floyd Jr, MD, true gentlemen, innovative surgeons, and outstanding mentors.

Intercarpal fusions are a well-recognized, time-tested treatment option in the management of a variety of pathologic conditions affecting the wrist. The indications are broad and typically include arthritis, nonunion, Kienböck’s disease, and carpal instability. Virtually every combination of limited wrist fusions has been described, with the rationale of each being to transfer load from a painful or structurally incompetent area of the carpus to a nonpainful structurally sound region. Although the literature regarding limited wrist fusions is generally favorable, many reported complications exist. Complications can be considered in the context of general (inherent to all limited wrist fusion procedures), patient- or disease-specific, or procedure-/technique-specific. The surgeon has little or no control over the first two types of complications, but a thorough understanding of carpal mechanics and meticulous attention to surgical technique can minimize procedure- and technique-related complications. In this chapter, we review complications of limited wrist fusions with particular attention to specific commonly performed intercarpal fusions.

The first reported limited wrist fusion was in 1924, when Thornton reported a case of midcarpal fusion for a neglected capitate dislocation. Scattered case reports then existed until the 1980s and 1990s, when a proliferation of biomechanical studies and clinical series were published. Commonly performed midcarpal fusions include scaphotrapeziotrapezoid (STT), scaphocapitate (SC), capitate-hamate-lunate-triquetrum (four-corner), and capitolunate (CL). Within the proximal row, lunotriquetral (LT) arthrodesis can address specific lunotriquetral pathology, and scapholunate (SL) arthrodesis has been attempted as well.

The most common form of degenerative wrist arthritis is secondary to scapholunate advanced collapse (SLAC wrist). Watson and Ballet have detailed this predictable pattern of degenerative change that occurs in the presence of an incompetent scapholunate ligament beginning with the distal pole of the scaphoid and the radial styloid (stage I). Next, the proximal scaphoid and scaphoid fossa are involved (stage II) followed by the midcarpal joint (stage III), and finally pan-carpal arthritis (stage IV). It is the preservation of the lunate fossa that serves as the basis for motion-preserving procedures such as proximal row carpectomy and midcarpal fusions with scaphoid excision (four-corner and capitolunate). The lunate fossa is spared because with disruption of the scapholunate ligament, the lunate (with its shorter dorsal pole relative to a slightly taller volar pole) rotates into extension. This position of the lunate gives the classic dorsal intercalated segmental collapse (DISI) seen with scapholunate dissociation. When rotated into extension the lunate is no longer able to contribute to the dissipation of load across the carpus; therefore, the scaphoid bears an increased pathologic load, and progressive degenerative change is seen. When performing intercarpal fusions for a SLAC wrist, it is critical to correct this lunate extension to allow the lunate to again participate in load sharing. A direct correlation between restoration of lunate position and return of more normal load distribution and contact pressure across the lunate facet has been shown biomechanically as well. Furthermore, the position of the lunate in the fusion mass has been shown biomechanically to be predictive of the attainable arc of motion with intercarpal fusions. Predicted loss of motion with all combinations of limited wrist fusions has been shown biomechanically and is described with each specific arthrodesis, but in general the midcarpal joint has been shown to contribute 30% to 36% of wrist flexion-extension. Let us now consider individual intercarpal fusion techniques and complications.

FOUR-CORNER FUSION

Traditionally, four-corner fusion has been the recommended limited wrist fusion for SLAC wrist based on the large surface area available for fusion and its proven track record in published series. Scaphoid excision is typically performed in conjunction with the procedure. Other articles and chapters well describe the proper surgical technique, but the keys to the surgery include proper decortication of all bone surfaces, restoration of proper bone carpal alignment (specifically correction of the lunate extension and restoration of the collinear lunate-capitate relationship), and adequate postoperative immobilization. Failure to properly decorticate all bony surfaces and failure to properly immobilize patients postoperatively can contribute to increased rates of nonunion. The published rate of nonunion varies considerably for four-corner fusions, but meta-analyses have shown a published mean of 4.3% to 8.4% overall. Attention should be paid as well to the presence of ulnar carpal translation on preoperative radiographs. Significant ulnar carpal translation suggests incompetence of the critical volar ligamentous support and is a contraindication to limited wrist fusion.

Implant selection, specifically with regard to circular plates, has been implicated by many authors as a significant independent contributor to nonunion. Recent studies have reported nonunion rates of 25% to 62% using these plates. Furthermore, higher DASH scores ( d isabilities of the a rm, s houlder, and h and), decreased range of motion, and decreased mean grip strength have been demonstrated in patients undergoing four-corner fusion with circular plates compared with traditional techniques in one head-to-head comparison study. Our experience with circular plates has also been disappointing, and we have experienced several cases of nonunion with this technique ( Fig. 35-1 ). Other authors, however, have reported excellent results with circular plates. Recently, Merrell and colleagues presented a series of 28 consecutive four-corner fusions with the use of circular plates with no cases of nonunion.

In a review of eight published reports totaling 431 cases of four-corner fusions, dorsal carpal impingement was found to be the most common complication, occurring in 4.4% of cases. In a review of over 1000 cases, Watson and collegues found this to be the most common complication in their series, with an incidence of 13%. Again, this is one complication that can be minimized with proper attention to surgical technique. As alluded to previously, restoring a collinear relationship of the lunate and capitate is critical to the success of this procedure. Failure to do so predisposes patients to dorsal carpal impingement because the capitate rests in a dorsally subluxated position and abuts the dorsal rim of the radius with wrist extension. Dorsal carpal impingement can result in loss of wrist extension in addition to wrist pain. Figure 35-2 illustrates the failure to properly reduce the lunate extension. One technical tip that can be helpful intraoperatively in maintaining lunate reduction is to place a Kirschner (K) wire through the dorsal aspect of the radius into the lunate. Flexing the wrist and using K wires to joystick the lunate can also aid in lunate reduction initially. When dorsal carpal impaction does arise, resection of the dorsal ridge of the radius has been shown to be effective in relieving pain and gaining an average 10 degrees of extension. When dorsal hardware is present, such as a circular plate or staples, the risk of dorsal impingement is even higher, and a concerted effort to recess the hardware must be made. With circular plates, dorsal impingement has been reported in up to 22% to 25% of cases in some reports.

Other complications of four-corner fusions that the surgeon has less control over include loss of wrist range of motion, loss of grip strength, superficial and deep infection (3% and 0.5%, respectively), complex regional pain syndrome (3%), and progressive degenerative change necessitating revision surgery (2%). When compared among all studies, range of motion tends to be between 55% and 75% of that the contralateral side, and grip strength roughly 80% of that of the contralateral side.

A recent complication of four-corner fusion that we have encountered and reported is that of pisotriquetral arthritis. We reported this complication in nine patients after limited and total wrist arthrodeses. In a separate, recent retrospective series, 33% (3 of 18 patients) developed pisotriquetral arthritis after undergoing four-corner arthrodesis. In a series of capitolunate fusions with triquetral retention, 29% of patients (2 of 7) required subsequent pisiform or triquetral resection. Scattered reports of pisiform excision after limited wrist fusions can be found in the literature as well. Our biomechanical study found substantial alterations in the kinematics of the pisotriquetral joint after four-corner fusion. A 33% loss of coronal plane motion and 31% loss of sagittal plane motion at the pisotriquetral joint were noted after simulating a four-corner fusion. We also found a progressive increase in pressure across the pisotriquetral joint with progressive wrist extension. In nearly two thirds of our cases in which pisotriquetral arthritis developed, we retrospectively noted that adequate correction of lunate extension had not been achieved. We now routinely examine the pisotriquetral joint clinically and radiographically to counsel patients concerning this complication preoperatively.

CAPITOLUNATE ARTHRODESIS

The rationale of the capitolunate arthrodesis is similar to that of the four-corner fusion in that preservation of the lunate fossa allows a reduced lunate to bear the load of the carpus. It is interesting that in Watson’s and Ballet’s landmark article on the management of SLAC wrist, 3 of the reported 16 patients who underwent limited wrist fusions had a capitolunate fusion with results similar to those of patients who had undergone four-corner fusion. Isolated capitolunate fusion was first reported in 1966 for Kienböck’s disease, and early applications of this technique to the SLAC wrist were met with poor outcomes secondary to a high nonunion rate. Indeed, a review of early studies on the use of this technique reveals that nonunion was a major complication as reported in 33% to 50% of cases. Refinements of the capitolunate fusion technique (specifically changing fixation from K wires to compression screws) have greatly decreased the incidence of nonunion, and the technique offers the advantages of easier lunate reduction following triquetrum excision, a reduced or even eliminated need for bone graft, and elimination of pisotriquetral arthritis as a complication.

One unique complication of isolated capitolunate fusion that we have found is screw backout ( Fig. 35-3 A and B). In a recent presentation, this complication was noted in 31% of cases and necessitated at minimum a return trip to the operating room for screw removal, and in 40% of these cases it required conversion to total wrist arthrodesis. Sixty percent of the cases of screw backout involved triquetrum retention. It is possible, though not proven, that triquetral retention limits the amount of compression that can be achieved between the lunate and capitate and predisposes to screw backout. Dorsally placed compression staples could offer an alternative means of fixation to prevent this complication, but this introduces an increased risk of dorsal impingement. Only one study has directly compared four-corner with capitolunate fusion. This study found no significant difference between the two techniques at an average 3-year follow-up with respect to range of motion, grip strength, DASH scores, and visual analogue (pain) score (VAS).