Hand Surgery
1st Edition

Four-Corner Fusion
Mark S. Cohen
Stability of the wrist is provided by the tight-fitting anatomic design of the individual carpal bones and the ligamentous interconnections that control movement of one bone relative to another. Wrist instability results from soft tissue or bony disruption, leading to pathologic carpal orientation. As the bones lose their alignment, they assume a collapsed position, with dissipation of the normal stored potential energy within the wrist (1,2). Motion and load-bearing capacity are lost. Pain occurs secondary to abnormal shear forces, synovitis, and ultimately cartilage degeneration.
The most common form of carpal instability occurs between the scaphoid and lunate. When load-bearing potential of the scapholunate interosseous ligament is lost, the scaphoid collapses into a flexed posture and the triquetrum and lunate extend. Scapholunate dissociation significantly alters articular contact areas and stress patterns within the carpus. Experimentally, only 5 degrees of pathologic flexion of the scaphoid results in a 45% reduction in the radio-scaphoid contact area (3). In this setting, arthritic changes begin at the radial styloid articulation with the scaphoid and progress to the proximal radioscaphoid joint. Degeneration then moves to the midcarpal capitolunate joint. This is termed scapholunate advanced collapse (SLAC) arthritis (4). This is the final common pathway for a variety of degenerative disorders of the wrist, the most common being scapholunate dissociation. It is also seen in scaphoid nonunion, avascular necrosis of the scaphoid, rheumatoid arthritis, and calcium pyrophosphate deposition disease (pseudogout) (5).
As a result of its concentric design, the radiolunate joint is unloaded and preserved regardless of the etiology or stage of SLAC degeneration. Four-corner fusion of the midcarpal joint (lunate, triquetrum, capitate, and hamate) with scaphoid excision is based on maintenance of this relationship. Successful arthrodesis results in isolated radiocarpal motion with force transmission through the radiolunate articulation and the triangular fibrocartilage complex.
Patients with advanced wrist arthritis typically present with measurable limitation of motion and function. Pain is variable and is usually aggravated by loading activities. Interestingly, although wrist degeneration has certainly been present for some time (months to years), patients often report a relatively recent onset of symptoms. On examination, soft tissue swelling on the dorsoradial aspect of the wrist is not uncommon. Discomfort can be localized to the radioscaphoid joint both dorsally and radially in the majority of cases. A scaphoid shift test, consisting of palmar pressure applied to the distal scaphoid as the wrist is brought into radial deviation, may reproduce symptoms and elicit crepitation in advanced cases (6). Provocative maneuvers on the ulnar aspect of the wrist are usually negative.
Plain radiographs are most commonly all that is necessary to confirm the diagnosis of SLAC arthritis. A static dissociation between the scaphoid and lunate or a displaced fracture of the scaphoid is typically obvious. Whereas sclerosis and joint space narrowing are easily identified in advanced cases, articular cartilage loss is typically greater than that appreciated on standard radiographic projections (Fig. 1) (7). Conservative treatment measures include activity modification, short-term splintage of the wrist, antiinflammatory medication, and intraarticular cortisone injections. Once advanced carpal degeneration is present, the wrist can be treated only by salvage procedures and there is no urgency for surgical intervention.
The two most common motion-preserving surgical procedures for SLAC wrist arthritis are scaphoid excision and four-corner fusion and proximal row carpectomy (PRC). Proponents of PRC cite technical ease, early mobilization, and the lack of nonunion risk as advantages (8,9,10,11 and 12). Potential disadvantages of PRC include shortening of the carpus, with associated weakness and incongruity between the capitate and the lunate fossa of the distal radius. Ulnar midcarpal fusion was introduced to relieve pain and restore wrist stability and height while providing more physiologic motion through the preserved radiolunate and ulnocarpal joints (4,13). This is a more technically demanding procedure that typically requires a longer period of postoperative immobilization.

FIGURE 1. A: Anteroposterior view of the wrist in a patient presenting with wrist pain revealing scapholunate dissociation. The patient had palpable radioscaphoid synovitis with crepitation during a scaphoid shift maneuver. Note the apparent lack of radiographic degenerative changes. B: Intraoperative view of the proximal pole of the scaphoid (left above elevator) revealing advanced articular loss. C: Retractor allowing visualization of midcarpal joint revealing arthrosis of the capitate and hamate as well. Articular changes are often well in advance of those appreciated radiographically. D: Note maintenance of the radiolunate joint (tip of forceps) in scapholunate advanced collapse arthritis.

Number of cases Follow-up (mo) Average extension (degrees) Average flexion (degrees) Averageb flexion–extension arc (%) Grip strengthb (%)
100 48 32 42 53 81
24 23 25 23 –– 70
23 41 27 27 –– 79
20 24 39 43 65 78
18 36 26 34 54 67
17 27 36 31 47 74
Avg. of all reported cases   31 36 54 77
aIncludes articles with 15 or more patients.
bPercentage of opposite, unaffected wrist.
Reprinted with permission from Cohen MS, Kozin SH. Degenerative arthritis of the wrist: proximal row carpectomy versus scaphoid excision and four-corner fusion. J Hand Surg [Am] 2001;26:94–104.
Multiple authors have reported successful pain relief with preserved motion and strength after both procedures. A review of the larger series in the literature reveals comparable results in terms of motion and strength (4,14,15,16,17,18,19,20,21,22,23,24,25 and 26). Combining these reports, PRC results in an average flexion–extension arc of approximately 60% of the opposite wrist versus 54% after four-corner fusion (Tables 1 and 2). Average reported grip strength versus the opposite side is 79% after PRC and 77% after four-corner fusion.
There are few reports specifically comparing these two surgical options for the treatment of SLAC arthritis. Wyrick et al. retrospectively compared 17 four-corner fusions with 11 PRC procedures (15). Motion and grip strength were greater in the PRC group. However, due to study limitations, the authors were unable to prove the superiority of one procedure over the other. Tomaino et al. compared nine four-corner patients with 15 patients treated with PRC (14). Although grip strength was similar between groups, range of motion was superior after PRC. Unfortunately, 20% of the PRC patients were not satisfied with their surgical result (versus none of the four-corner group). These authors recommend PRC in the absence of capitate arthrosis. Krakauer et al. compared 23 patients treated with four-corner fusion with 12 PRC patients (24). Wrist motion was greater after PRC, whereas grip strength was superior after four-bone fusion. Pain relief, however, appeared less reliable after PRC, with three of 12 (25%) patients reporting persistent severe pain versus three of 23 (13%) four-corner fusion patients.
Number of cases Follow-up (mo) Average extension (degrees) Average flexion (degrees) Averageb flexion–extension arc (%) Grip strengthb (%)
27 48 84 (extension + flexion) –– 80
24 36–120 46 38 59 ––
23 72 37 37 61 79
22 120 48 48 63 81
18 157 40 36 63 83
17 42 35 28 52 67
15 72 40 37 64 77
15 30 39 43 –– 84
Avg. of all reported cases   41 38 60 79
aIncludes articles with 15 or more patients.
bPercentage of opposite, unaffected wrist.
Reprinted with permission from Cohen MS, Kozin SH. Degenerative arthritis of the wrist: proximal row carpectomy versus scaphoid excision and four-corner fusion. J Hand Surg [Am] 2001;26:94–104.
Cohen and Kozin compared the largest series of two cohort populations from separate institutions performing exclusively four-corner fusion or PRC for SLAC arthritis (27). Patient groups were matched with respect to age, sex, dominance, stage of arthritis, and preoperative measures of pain and function. Evaluations completed by independent examiners revealed remarkably similar outcomes at a 2-year follow-up. Both procedures resulted in approximately an 80-degree flexion–extension arc measuring nearly 60% of the contralateral wrist. Grip strength was slightly greater after four-corner fusion, averaging 79% of the opposite side

in these patients versus 71% after PRC (p >.05). Four-corner fusion did result in a greater radial-ulnar deviation arc, predominantly due to limited radial deviation after PRC (averaging only 7 degrees). This loss of radial deviation has been reported previously after PRC with or without radial styloidectomy (12,19,20,21 and 22,25,26). Pain and function significantly improved in both groups from preoperative levels using a variety of measures, with little difference between the two procedures.
The indication for four-corner fusion most commonly involves radioscaphoid degeneration with or without midcarpal involvement. In advanced cases of scapholunate dissociation or scaphoid nonunion, there is typically little question regarding the degree of arthrosis, and preoperative planning is straightforward. In some cases, direct inspection of the radioscaphoid joint articular cartilage is required to determine the need for salvage surgery. Once advanced scaphoid fossa degeneration exists, four-corner fusion is a viable option in symptomatic patients who fail conservative measures. Other less common indications for midcarpal fusion include midcarpal instability without radioscaphoid arthrosis (28), avascular necrosis of the scaphoid with collapse, and failed scaphocapitate or scaphotrapezial–trapezoid fusion. Contraindications include radiolunate arthrosis and ulnar translocation of the carpus.
The surgical procedure is typically performed under regional anesthesia. A dorsal longitudinal incision is centered over the proximal carpal row. The wrist joint is entered though the third dorsal compartment, with resection of the posterior interosseous nerve, which lies on the floor of the fourth compartment. The joint capsule is opened and subperiosteally dissected radially and ulnarly. Deep retractors and longitudinal traction help expose the individual carpal articulations. Most commonly, the scaphoid is identified and excised. In nonunion, one can choose to retain the proximal scaphoid pole, which is typically spared from arthrosis with the lunate (29). This increases the surface area of the resultant radiocarpal articulation. It does, however, theoretically increase the chance of late degeneration between the proximal scaphoid pole and the radius, as this joint is not concentric and can be placed under shear if not perfectly reduced.
In scapholunate dissociation, scaphoid resection is facilitated by transecting the bone in the mid-waist perpendicular to its axis. Threaded Steinmann pins are placed into the distal and proximal poles and used as joysticks. Care is taken not to violate the palmar extrinsic radiocarpal ligaments, especially the more ulnar long radiolunate ligament. This is required to stabilize the carpus after fusion. A limited radial styloidectomy can be performed in individuals with large osteophytes to improve visualization.
Reduction of the collapse deformity to realign the midcarpal joint is critical to the success of the procedure. The lunotriquetral relationship can be first reduced by aligning the distal concavity of the two bones from the midcarpal joint. This is secured with 0.045 Kirschner wires. An additional wire can be placed into the lunate and used as a joystick for the proximal carpal row during removal of the cartilage (Fig. 2). The fusion surfaces between the lunate, capitate, hamate, and triquetrum are then denuded down to cancellous bone with a low-speed bur, rongeurs, and curettes. Of note, the lunate often has thick subchondral bone, and it is frequently not possible to denude this surface down to a soft cancellous level. Fine Kirschner wire holes can be placed in the lunate concavity to facilitate vascular ingrowth and union.
Pure cancellous bone graft can be obtained from the excised scaphoid, although this provides only marginal material. Additional autograft is available from the distal radius just proximal to Lister’s tubercle (30). The graft can be compressed in a syringe and meticulously packed into the fusion sites. Wrist distraction is then released, and the proximal carpal row is reduced. Care is taken to flex the lunate and palmarly and ulnarly translate the capitate. A slight overreduction (volarflexed intercalated instability) is preferred. With compression maintained across the midcarpus, the reduction can be stabilized with either multiple Kirschner wires or cannulated screws. Wires are placed retrograde from the capitate and hamate into the lunate. It is often helpful to introduce these pins from the second and fourth web spaces, respectively, to begin palmar enough to obtain adequate purchase in the lunate. One additional pin is placed from the triquetrum into the capitate (Fig. 2). Cannulated screws are typically placed antegrade from the lunate and triquetrum into the distal carpal row (Fig. 3). Wrist flexion and extension are tested on the table and greater extension than flexion should be confirmed. The remaining bone graft is applied dorsally across the fusion site. The longitudinal capsulotomy is loosely approximated––but not repaired––to the radius, and the extensor pollicis longus is left outside the repaired retinaculum.
An alternative method of fixation involves the use of a conical plate designed in 1999 to specifically stabilize the midcarpal joint after arthrodesis (Kinetikos Medical, San Diego, CA). This plate is recessed beneath the dorsal articular margin of the carpus after provisional wire stabilization of the reduced midcarpal relationship (Fig. 4). A conical rasp is used to create a recess for the plate, which has eight holes. This allows two screws to be placed into each carpal bone in optimal circumstances. Care must be taken to follow similar principles for arthrodesis when using plate fixation. The fusion surfaces must still be denuded of articular cartilage (at least half of the anteroposterior depth of the joints) and the autograft meticulously packed to optimize healing. The screws are tightened sequentially to obtain stability and circumferential compression. Additional graft can be added



to the center of the four-bone articulation after fixation through the central plate opening.
FIGURE 2. A: Preoperative posteroanterior and (B) lateral radiographs depicting scapholunate advanced collapse arthritis. Note the scapholunate gap, radioscaphoid narrowing, and the extended posture of the lunate on the lateral film. C: Intraoperative photograph taken with traction after preparation of the midcarpal articular surfaces for fusion. Note the Kirschner wire in the lunate that is used as a joystick. D: Bone graft obtained from the distal radius (note defect) has been meticulously packed into the fusion surfaces. E: The midcarpal joint has been reduced and pinned with additional autograft packed dorsally. F: Postoperative posteroanterior and (G) lateral views revealing scaphoid excision and four-corner pin fixation. Note the position of the retrograde midcarpal pins into the lunate. Mid-axis placement is facilitated by starting these in the second and fourth web spaces, respectively. H: Final posteroanterior and (I) lateral radiographs depicting consolidation of four-corner arthrodesis.
FIGURE 3. A: Posteroanterior and (B) lateral radiographs of four-corner arthrodesis accomplished with cannulated screws. These are placed antegrade with the wrist maximally flexed. Screw fixation allows for more rapid rehabilitation.
After surgery, patients are placed in a compressive dressing with an internal short arm splint. Digital motion is encouraged immediately after the regional block wears off. At 10 to 14 days postoperatively, a short arm cast is applied in those individuals treated with pins alone. These are typically removed at approximately 8 weeks postoperatively, at

which time therapy is begun. In those treated with screws or a plate, early gentle active motion can be started with interval splinting for comfort and support if stable fixation was achieved at the time of surgery. Gentle passive motion is typically not begun until 3 to 4 weeks postoperatively in these individuals. Strengthening is initiated when the union matures, commonly at approximately 10 to 12 weeks postoperatively.
FIGURE 4. A: Posteroanterior and (B) lateral radiographs of a four-corner fusion performed with a conical plate. This provides stable fixation without violation of the articulating surfaces of the lunate and triquetrum.
FIGURE 5. A: Schematic diagram and (B) clinical example depicting undercorrection of the midcarpal joint during four-corner arthrodesis. The extended posture of the lunate results in dorsal radiocarpal impingement with limited wrist extension. C: Diagram depicting slight overcorrection of the capitolunate relationship. This improves wrist extension by relaxing the palmar radiocarpal ligaments and removing the dorsal impingement. (Figs. 5A and 5C reprinted with permission from Cohen MS, Kozin SH. Degenerative arthritis of the wrist: proximal row carpectomy versus scaphoid excision and four-corner fusion. J Hand Surg [Am] 2001;26:94–104.)
Individuals with SLAC wrist arthritis who fail conservative treatment measures are candidates for four-bone fusion. Once cartilage loss occurs at the radioscaphoid articulation, either a PRC or midcarpal fusion is a motion-preserving option. A PRC is contraindicated in the presence of advanced arthrosis of the capitate head. The four-corner fusion can be used in either situation. As previously stated, contraindications include radio-lunate degeneration and ulnar translocation of the carpus.
The most common complication after four-corner fusion is dorsal radiocarpal impingement in wrist extension (14,16). This occurs secondary to inadequate reduction of the capitolunate relationship (Fig. 5). Care must be taken to not leave the lunate in an extended posture (dorsiflexed intercalated segment instability) at the time of fixation. This markedly limits postoperative wrist extension and can lead to impingement. It is recommended to actually overreduce the midcarpal joint into a slight volarflexed intercalated instability pattern to avoid this complication (27). This results in increased wrist extension [which is more important than flexion functionally (31)] without an equal loss of wrist flexion. Flexion of the lunate relaxes the intact palmar radiolunate ligaments, augmenting wrist extension (Fig. 5). In turn, flexion appears less limited by the dorsal capsule and remaining dorsal extrinsic ligaments, neither of which is repaired to the distal radius after exposure and arthrodesis.
Nonunion of four-corner arthrodesis is relatively rare. Originally, the midcarpal fusion technique was limited to the capitolunate articulation. However, the triquetrum and hamate were subsequently added to increase the fusion surface area and improve healing rates. When nonunion occurs, it can be related to technical errors, patient variables (e.g., tobacco use), or a combination of factors (Fig. 6). If pins were originally used, repeat fusion is probably best accomplished with fresh autograft and screws or plate fixation.

FIGURE 6. A: Posteroanterior radiograph depicting established nonunion of midcarpal fusion. B: Final posteroanterior and (C) lateral views after revision arthrodesis with auto-graft and internal screw fixation. Note the loss of correction with incomplete midcarpal reduction due to the nonunion. (Reprinted with permission from Cohen MS, Kozin SH. Degenerative arthritis of the wrist: proximal row carpectomy versus scaphoid excision and four-corner fusion. J Hand Surg [Am] 2001;26:94–104.)
Scaphoid excision with four-corner fusion is a reliable treatment method for individuals with SLAC arthritis who have failed conservative care. After successful arthrodesis, one can expect maintenance of approximately 60% to 80% of wrist motion and 75% to 80% of grip strength with predictable pain relief. Results are contingent, however, on adherence to the technical principles of the procedure. These include adequate bony preparation, meticulous placement of autograft, and proper reduction of the midcarpal relationship.
1. Cohen MS. Ligament injuries and instability patterns of the wrist. In: TR Light, ed. Hand surgery update, 2nd ed. Rosemont, Ill: American Society for Surgery of the Hand and American Academy of Orthopaedic Surgeons, 1999:97–106.
2. Zdravkovic V, Jacob HA, Sennwald GR. Physical equilibrium of the normal wrist and its relation to clinically defined “instability.” J Hand Surg [Br] 1995;20:159–164.
3. Burgess RC. The effect of rotatory subluxation of the scaphoid on radioscaphoid contact. J Hand Surg [Am] 1987;12:771–774.
4. Watson HK, Ballett FL. The SLAC wrist: scapholunate advanced collapse pattern of degenerative arthritis. J Hand Surg [Am] 1984;9:358–365.
5. Chen C, Chandnani VP, Kang HS, et al. Scapholunate advanced collapse: a common wrist abnormality in calcium pyrophosphate dihydrate crystal deposition disease. Radiology 1990;177:459–461.
6. Watson HK, Ashmead D, Makhouf MV. Examination of the scaphoid bone. J Hand Surg [Am] 1988;13:657–660.
7. Peh WC, Patterson RM, Viegas SF, et al. Radiographic-anatomic correlation at different wrist articulations. J Hand Surg [Am] 1999;24:777–780.

8. Ferlic DC, Clayton ML, Mills MF. Proximal row carpectomy: review of rheumatoid and nonrheumatoid wrists. J Hand Surg [Am] 1991;16:420–424.
9. Inglis AE, Jones EC. Proximal-row carpectomy for diseases of the proximal row. J Bone Joint Surg Am 1977;59:460–463.
10. Inoue G, Miura T. Proximal row carpectomy in perilunate dislocations and lunatomalacia. Acta Orthop Scand 1990;61:449–452.
11. Neviaser RJ. On resection of the proximal carpal row. Clin Orthop Rel Res 1986;202:12–15.
12. Salomon GD, Eaton RG. Proximal row carpectomy with partial capitate resection. J Hand Surg [Am] 1996;21:2–8.
13. Watson HK, Goodman ML, Johnson TR. Limited wrist arthrodesis. Part II: intercarpal and radiocarpal combinations. J Hand Surg [Am] 1981;6:223–233.
14. Tomaino MM, Miller RJ, Cole I, et al. Scapholunate advanced collapse wrist: proximal row carpectomy or limited wrist arthrodesis with scaphoid excision. J Hand Surg [Am] 1994;19:134–142.
15. Wyrick JD, Stern PJ, Kiefhaber TR. Motion-preserving procedures in the treatment of scapholunate advanced collapse wrist: proximal row carpectomy versus four-corner arthrodesis. J Hand Surg [Am] 1995;20:965–970.
16. Ashmead D, Watson HK, Damon C, et al. Scapholunate advanced collapse wrist salvage. J Hand Surg [Am] 1994;19:741–750.
17. Culp RW, McGuigan FX, Turner MA, et al. Proximal row carpectomy: a multicenter study. J Hand Surg [Am] 1993;18:19–25.
18. Gill DRJ, Ireland DCR. Limited wrist arthrodesis for the salvage of SLAC wrist. J Hand Surg [Br] 1997;22:4:461–465.
19. Green DP. Proximal row carpectomy. Hand Clin 1987;3:163–168.
20. Imbriglia JE, Broudy AS, Hagberg WC, et al. Proximal row carpectomy: clinical evaluation. J Hand Surg [Am] 1990;15:426–430.
21. Jebson PJ, Hayes E, Engber WD. Proximal row carpectomy: a minimum 10-year follow-up. Presented at the American Society for Surgery of the Hand, 52nd Annual Meeting, Denver, 1997.
22. Jorgensen EC. Proximal row carpectomy. J Bone Joint Surg Am 1969;51:1104–1111.
23. Kirschenbaum O, Schneider LH, Kirkpatrick WH. Scaphoid excision and capitolunate arthrodesis for radioscaphoid arthritis. J Hand Surg [Am] 1993;18:780–785.
24. Krakauer JD, Bishop AT, Cooney WP. Surgical treatment of scapholunate advanced collapse. J Hand Surg [Am] 1994;19:751–759.
25. Neviaser RJ. Proximal row carpectomy for posttraumatic disorders of the carpus. J Hand Surg [Am] 1983;8:301–305.
26. Tomaino MM, Delsignore J, Burton RI. Long-term results following proximal row carpectomy. J Hand Surg [Am] 1994;19:694–703.
27. Cohen MS, Kozin SH. Degenerative arthritis of the wrist: proximal row carpectomy versus scaphoid excision and four-corner fusion. J Hand Surg [Am] 2001;26:94–104.
28. Lichtman DM, Bruckner JD, Culp RW, et al. Palmar midcarpal instability: results of surgical reconstruction. J Hand Surg [Am] 1993;18:307–315.
29. Viegas SF. Limited arthrodesis for scaphoid nonunion. J Hand Surg [Am] 1994;19:127–133.
30. Bruno RJ, Cohen MS, Berzins A, et al. Bone graft harvesting from the distal radius, olecranon and iliac crest: a quantitative analysis. J Hand Surg [Am] 2001;26:135–141.
31. Palmer AK, Werner FW, Murphy D, et al. Functional wrist motion: a biochemical study. J Hand Surg [Am] 1985;10:39–46.