Rationale and Basic Science

Scapholunate dissociation is a vexing problem facing hand surgeons. It is the most common form of carpal instability.^1–4 Untreated scapholunate interosseous ligament (SLIL) disruption leads to significant degenerative changes of the carpus.^4–8 A multitude of treatment options are described, which suggests that an optimal strategy has yet to be discovered.^5,9–41 Partial wrist fusion, although a durable solution, trades off improvements in pain with decreased range of motion and strength.³⁸

The great strides made in anatomical anterior cruciate ligament reconstruction over the past 30 years have led some hand surgeons to consider a similar model for SLIL reconstruction. Biomechanical studies show that the dorsal portion of the ligament is strongest and most important functionally.^8,42–47 The yield strength of the dorsal region was 260N, more than twice that of the palmar region.⁴² Reconstruction of the dorsal ligament also is appealing given the ease of exposure through a standard dorsal approach. Various composite grafts have been investigated, including bone-retinaculum-bone from the distal radius,³⁰ navicular-cuneiform autograft,^27,41 SLIL allograft,^5,12 metacarpal-carpal composite,⁴⁸ and capitohamate ligament,⁴² in an attempt to match the strength, stiffness, and functionality of the native SLIL (Fig. 50-1).

FIGURE 50-1 Schematic diagram of the various sources of autograft used in scapholunate ligament reconstruction. All of the techniques have the same method of inset into the scapholunate interval as described by Weiss.^30,31

Indications and Contraindications

The ideal candidate is a patient with an acute or chronic rupture of the SLIL and only dynamic instability. Patients with a history consistent with an SLIL tear, pain over the scapholunate interval, or stress views showing widening at the scapholunate interval were considered for operative intervention. The definitive diagnosis of dynamic instability was made by direct arthroscopic examination in all patients. The reconstruction is rarely necessary in Geissler grade 1, but is commonly performed for grades 2 through 4. Patients with static scapholunate widening generally have not done well with bone-tendon-bone reconstruction, which is a relative contraindication. Sclerosis of the lunate (sometimes seen in cases of perilunate or lunate dislocation) also is a relative contraindication because of the risk of graft pullout. Degenerative changes of the carpus would be an absolute contraindication. An inability to reduce anatomically the scapholunate gap and rotation would be another absolute contraindication. Low-demand patients and elderly patients may be better served with a definitive procedure, such as a proximal row carpectomy.

Navicular-Cuneiform Autografts

Davis and colleagues⁴¹ reported on the use of the dorsomedial portion of the navicular–first cuneiform ligament. Although weaker than the SLIL, its minimal donor morbidity made it an acceptable clinical choice. The graft is harvested through a medial incision with posterior retraction of the tibialis anterior. Each bone plug measures 10 × 5 × 5 mm and is secured with small screws. The same group of investigators reported several cases of tissue elongation without loss of angulation.⁴⁹ Commentary on donor site morbidity has not been published at this point. Svoboda and associates²⁷ reported on the biomechanics of several other bone-tendon-bone complexes in the foot (e.g., tarsometatarsal ligament); however, none of these have been reported in clinical use.

Metacarpal-Carpal Autografts

Metacarpal-carpal grafts have several appealing features. The articulation at the base of the second or third metacarpal is relatively immobile and perhaps expendable. It can be harvested through the same dorsal incision as the reconstruction. These grafts more closely approximate the strength and stiffness of the SLIL than distal radius retinaculum grafts.⁴⁸ Early results from a small case series are encouraging.⁵⁰ The two main complications—graft pullout and graft stretching—are similar to reports with other types of bone-tendon-bone grafts. Graft stretching typically occurs after several months, perhaps during the revascularization phase of healing. There is loss of a tight SLIL interval, but the scapholunate angle is usually maintained. Graft pullout is seen more often in cases with poor lunate bone quality. A vascularized third metacarpal-carpal bone graft may have some benefit in these cases.⁴⁹

Capitohamate Autografts

Berger’s group⁵¹ pioneered the use of the capitohamate graft for SLIL reconstruction. There are three capitohamate interosseous ligaments—dorsal, deep, and palmar. The dorsal ligament has a load to failure of 133N. The dorsal ligament is clinically feasible to harvest, and preservation of the deep and palmar ligaments is likely to maintain the capitohamate orientation. Berger supplements the repair with a capsulodesis and pins. Protected motion is started at 4 weeks, and pins are removed at 8 weeks. There are no clinical results published yet on this technique. Berger has commented that he does not perform this reconstruction if the lunate is translated ulnarly more than 50% out of the fossa.⁴⁹

Authors’ Preferred Technique (Distal Radius Retinaculum Graft)

A standard dorsal midline approach is made. The extensor pollicis longus is released on the ulnar side and transposed radially. Using a small osteotome, a single 20 × 8 × 8 mm bone block with the overlying periosteum and retinaculum is harvested from the distal radius near Lister’s tubercle between the second and third compartments (Fig. 50-2). A 2- to 3-mm segment of cancellous and cortical bone is removed from the midportion of the bone block with a fine rongeur. Care is taken not to disrupt the overlying periosteum and retinaculum.

FIGURE 50-2 Overall diagram of the basic method for implanting a bone-retinaculum-bone autograft.

The capsule between the second and fourth compartment is incised longitudinally or via a ligament-sparing incision to expose the scapholunate interval. The wrist is flexed to improve visibility of the joint, and Kirschner wires (K-wires) are placed into the scaphoid and lunate for use as joysticks (Fig. 50-3). After reduction is confirmed on fluoroscopy, two 0.045-inch K-wires are drilled from the scaphoid into the lunate to capture the reduction. The joystick K-wires are removed. With a small osteotome, troughs are cut in the scaphoid and lunate adjacent to the joint, matching the size of the harvested graft (Fig. 50-4). The graft is digitally impacted into the scaphoid and lunate. A 1.3-mm screw is placed into each graft, securing them to the scaphoid and lunate (Fig. 50-5). Intraoperative fluoroscopy is used to check the alignment and placing of the screws and K-wires (Fig. 50-6). The capsule is repaired with imbrication as needed. Short arm spica immobilization in wrist extension is used for 6 weeks.

FIGURE 50-3 The use of Kirschner wires as joysticks to reduce the scapholunate interval before transfixing with Kirschner wires is essential.

FIGURE 50-4 From the bone block removed with its retinaculum and periosteum intact, a 2- to 3-mm segment of cancellous and cortical bone is removed from the midportion of the bone block with a fine rongeur resulting in the bone-retinaculum-bone autograft. Care is taken not to disrupt the overlying periosteum and retinaculum. Recipient troughs are fashioned in the scaphoid and lunate of a similar size to accept the bone-retinaculum-bone graft.

FIGURE 50-5 A press-fit application of the bone-retinaculum-bone graft into the troughs is performed followed by fixation with mini-screws.

FIGURE 50-6 Intraoperative fluoroscopy is used to check for appropriate placement of all hardware.

Harvey and colleagues⁴⁹ reported on 20 patients with a distal radius bone-retinaculum-bone reconstruction augmented with a Herbert screw, placed percutaneously from the radial side. The screw is removed at 9 to 12 months. One patient had screw pullout with gapping of the scapholunate interval, and another had a fixed dorsal intercalated segment instability deformity. Eighty percent of their first 20 patients returned to their previous occupations.

Results

In 1998, Weiss³¹ first reported on 19 patients treated with distal radius bone-retinaculum-bone fixation of the SLIL. Three of the five patients in this group with a static deformity went on to further surgery (proximal row carpectomy, radioscapholunate fusion, wrist arthrodesis). In a longer term examination of the 14 patients (15 wrists) with dynamic instability, we report the following results. Five could not be located. Of the remaining nine, three had subsequent surgery (one had proximal row carpectomy, and two had wrist arthrodesis). The reason for failure of these three is unknown. It could be that the ligament reconstruction tears through the soft tissue substance. We have not seen evidence that the bone plugs displace. For the other six, follow-up averaged 8.5 years. Average grip strength was 119 lb. Wrist flexion averaged 49 degrees, and wrist extension averaged 58 degrees (Fig. 50-7). Three patients had no pain, and three had mild pain with activity. All returned to work at their previous occupation. Mayo wrist scores were three excellent, two good, and one fair. The graft does not seem to stretch out over time. The natural history postoperatively is that if the graft remains intact for 6 to 12 months, it will stay so. We have noticed minor degenerative changes (usually around the junction of the scaphoid and lunate with slight spurring), but no loss of height of the radiolunate or radioscaphoid joints.

FIGURE 50-7 A 3-year follow-up examination revealed excellent maintenance of the scapholunate gap and little degenerative changes on posteroanterior and lateral radiographs. The patient has excellent function without pain even with strenuous exercise.

Over the follow-up period there have been few modifications to the procedure. If the quality of the soft tissue is robust, we use only the bone-retinaculum-bone graft with a single mini-screw fixation in each plug. In one case, the soft tissue quality was suspect, so we augmented the repair with a dorsal capsulodesis. In one case, not included in this series of pure bone-retinaculum-bone reconstructions, we found the tissue of such poor quality that we placed a headless screw across the scapholunate joint for additional temporary fixation.