Abstract
Background:
Implant arthroplasty of the wrist offers pain relief with preservation of motion to patients with rheumatoid arthritis, although few studies have investigated the long-term results of this procedure. The purpose of the present study is to report the prospective results of total wrist arthroplasty with use of the Universal wrist prosthesis in a consecutive series of patients with rheumatoid arthritis who were managed by a single surgeon.
Methods:
Twenty-four wrist arthroplasties in twenty patients with rheumatoid arthritis were followed prospectively. Nineteen wrists in fifteen patients were followed clinically and radiographically for a mean of 7.3 years (range, 5.0 to 10.8 years) after the index procedure. Outcome measures included the Disabilities of the Arm, Shoulder and Hand (DASH) score, wrist range of motion, and standard radiographic findings.
Results:
The average DASH score improved from 62 points preoperatively to 40 points at the time of the latest follow-up. The mean wrist flexion and extension at the time of the latest follow-up were 42° and 20°, respectively, for a mean improvement in the total flexion-extension arc of 14°. A total of nine wrists (45%) in eight patients underwent revision surgery because of a loose carpal component at the time of the latest follow-up. One patient underwent wrist arthrodesis because of recurrent wrist instability. Two additional wrists in two patients had radiographic evidence of carpal component subsidence at the time of the latest follow-up. The implant survival rates at five and seven years for the original prosthetic components were 75% and 60%, respectively.
Conclusions:
The results for the Universal wrist prosthesis at a minimum of five years of follow-up include a high rate of failure, most often because of carpal component loosening, resulting in revision of ten (50%) of twenty wrists at the time of the latest follow-up (with the inclusion of one revision in a patient who died before five years). Patients with a stable prosthesis maintained a functional range of motion and had improvement in patient-reported outcome measures.
Level of Evidence:
Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.
Rheumatoid arthritis of the wrist can be debilitating, with resultant interference with activities of daily living. When nonoperative treatments such as splinting and activity modification fail to relieve symptoms, surgical treatment is considered. Wrist arthrodesis provides a durable solution for patients with severe wrist arthritis but sacrifices wrist motion1. Decreased wrist motion is especially problematic for patients with bilateral wrist arthritis, multiple upper extremity joint involvement, or those who require wrist mobility for special needs. Wrist arthroplasty provides useful motion but historically has been associated with a high complication rate2,3. Prosthetic instability, soft-tissue imbalance, and carpal component loosening have often led to poor outcomes. The early results associated with the first Universal wrist prosthesis (KMI, Carlsbad, California) showed good survival rates and improvement in terms of pain and disability4,5. However, to our knowledge, the long-term results associated with this implant in patients with rheumatoid arthritis have not been reported. The purpose of the present study was to prospectively evaluate the results of wrist arthroplasty with use of the Universal total wrist prosthesis in a consecutive series of patients with rheumatoid arthritis after a minimum duration of follow-up of five years.
The present study was approved by our institutional review board, and informed consent was obtained from all patients.
Between 1997 and 2001, twenty-one patients (twenty-five wrists) with rheumatoid arthritis were managed with total wrist arthroplasty by a single surgeon. All patients were prospectively enrolled in the present study and were asked to return for follow-up at yearly intervals. One patient (one wrist) was excluded because she had undergone a previous total wrist arthroplasty with a Volz implant (Howmedica, Rutherford, New Jersey) at another institution. In the remaining twenty patients, arthroplasty was performed on fifteen right and nine left wrists. Fifteen procedures involved the dominant extremity. All patients were female. In order for a patient to be considered as a candidate for wrist arthroplasty, there could be no history of infection in the wrist, no active synovitis of the wrist, no loss of wrist extensor tendon function, no skin problems overlying the wrist, and no need to routinely bear weight on the upper extremities with a handheld device such as a cane, crutch, or walker. Despite the destructive effects of rheumatoid arthritis, the distal carpal row had to have sufficient preservation of bone stock to allow implantation of the carpal component used for wrist arthroplasty. Distal carpal row bone loss in the form of cysts or severe osteopenia may preclude secure fixation of this prosthetic carpal component.
At the time of the latest follow-up, four patients (four wrists) had died. Of those four patients, one patient (one wrist) had undergone conversion of the wrist arthroplasty to a wrist fusion 2.2 years after the index procedure because of recurrent implant instability. Because the final outcome for this implant (revision) was known prior to the patient's death, this patient was included in the survivorship analysis. One patient (one wrist) declined to return for the five-year follow-up. Minimum five-year radiographic and clinical data were available for the remaining nineteen wrists (fifteen patients), and the mean duration of follow-up for this group was 7.3 years (range, 5.0 to 10.8 years).
Surgical Technique
A single surgeon (B.D.A.) performed all procedures with use of the technique described by Menon5, with the exception of modification of the capsular flap as described below. A longitudinal skin incision is made in line with the third metacarpal. Starting in the sixth extensor compartment, the extensor retinaculum is elevated as a radially based single flap. A broad, distally based, rectangular capsular flap is raised to expose the entire joint, including the distal radioulnar joint.
The distal part of the radius and the distal part of the carpus are prepared with use of the alignment and cutting guides, with minimal resection of the distal part of the radius to better preserve the volar capsule. Trial radial and carpal components are inserted, and their position is confirmed with fluoroscopy. The wrist is moved through a range of motion to assess soft-tissue balance. A high-speed burr is used to remove any remaining joint surface between the trapezoid, capitate, and hamate to facilitate intercarpal fusion. Autogenous bone obtained from the partial carpal resection is used to augment the intercarpal arthrodesis. The distal part of the ulna is resected. Sutures are passed through drill holes in the ulna and are secured to the volar capsule to stabilize the distal part of the ulna. The trial components are exchanged for the final components. The radial component and the central stem of the carpal component are fixed with cement. The carpal component is further secured with two screws into the carpus. All implants were the first-generation Universal wrist prosthesis (KMI). During capsular closure, excessive tightness is avoided to prevent loss of flexion. In some cases, either a portion of the extensor retinaculum or an allograft (tensor fasciae latae) was used to augment a deficient capsule due to inadequate length or thickness secondary to the rheumatoid arthritis. The radial wrist extensor and extensor pollicis longus tendons remain superficial to the retinaculum, whereas the other dorsal tendons are placed beneath the extensor retinaculum.
A subcutaneous suction drain is used for twenty-four hours. A bulky gauze dressing incorporating a volar plaster wrist splint is applied. Strict elevation of the hand and early digital motion are encouraged to reduce swelling and the risk of tendon adhesions. If the wrist is stable on intraoperative examination at the completion of implantation, the bandage is changed on the second postoperative day and a removable custom plastic wrist splint is applied. The splint is removed two to three times daily for gentle active wrist motion exercises. If wrist instability was evident intraoperatively, a plaster splint is reapplied and is left in place for four weeks. After four weeks, the patient advances to active wrist motion and begins using the hand for light activities of daily living without the splint. Full use of the wrist without a splint is allowed after eight weeks, except that lifting >10 lb (>4.5 kg) with this hand alone is discouraged indefinitely. Patients may continue to use a splint as needed for more strenuous activities.
Clinical Evaluation
Patients completed the Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire preoperatively, between one and two years postoperatively, and at the most recent evaluation. The DASH questionnaire consists of thirty items that address the degree of difficulty in performing certain tasks as well as symptoms such as pain and stiffness. The DASH questionnaire is scored from 0 to 100, with 0 representing a perfect score. Preoperative clinical examination included measurements of wrist flexion and extension, radial and ulnar deviation, and pronation and supination with use of a hand-held goniometer. Range-of-motion measurements were repeated at each follow-up visit.
Radiographic Evaluation
Preoperative posteroanterior and lateral wrist radiographs were evaluated by the two authors (C.M.W. and T.K.) who did not perform the surgical procedures. Each wrist was classified with use of the Simmen system6. Simmen Type I (“ankylosis”) is characterized by spontaneous ankylosis of the intercarpal joints with or without radiocarpal fusion. Simmen Type II (“osteoarthritic”) is characterized by moderate articular surface destruction with associated subchondral sclerosis and occasional osteophytes, but no evidence of destabilization. Simmen Type III (“destabilization”) is characterized by carpal collapse, ulnar translocation of the carpus, and osseous destruction.
Postoperative posteroanterior and lateral radiographs were analyzed for evidence of carpal or radial component subsidence, osteolysis, intercarpal fusion, and implant breakage. Osteolysis was indicated by progressive radiolucency around the component. To our knowledge, there is not a standardized method for defining carpal component subsidence. In order to quantify carpal component subsidence for the purposes of the present study, we measured the distance from the intersection of the central post with the baseplate of the carpal implant to the base of the third metacarpal (Fig. 1-A). In some cases, the base of the third metacarpal could not be consistently identified on sequential radiographs. In those cases, the distance was measured to the base of the fourth metacarpal. A decrease in this distance on sequential radiographs indicated carpal component subsidence. The reliability of this measurement technique has not been established.Fig. 1-AFigs. 1-A, 1-B, and 1-C Case 4 (see Appendix). Fig. 1-A Posteroanterior radiograph of the left wrist and hand, made three months after wrist arthroplasty. The black arrows denote the start and end points of the measurement used to quantify carpal component subsidence. Note the lack of fusion between the capitate and the hamate. Fig. 1-B Posteroanterior radiograph, made three years after the index procedure, demonstrating that the carpal component has subsided so the distance between the baseplate and the base of the third metacarpal has decreased substantially.
Statistical Methods
A paired t test was used to compare preoperative and postoperative ranges of motion and DASH scores. When comparing wrists with or without revision surgery with regard to patient age, hand dominance, and preoperative DASH score, a t test was used. For all analyses, the level of significance was set at p < 0.05.
A Kaplan-Meier survivorship curve was created with use of data from the nineteen wrists that had minimum five-year follow-up as well as the one patient who had revision of the wrist before the time of death. The three patients who died before a minimum of five years of follow-up and the one patient who did not return for evaluation were excluded from the survivorship analysis.
Source of Funding
We received no external funding for this study.
Clinical Evaluation
At the time of the latest follow-up, ten wrist prostheses (50%) in eight patients remained functional. In these patients, the average DASH score improved from 62 points (range, 42 to 80 points) preoperatively to 40 points (range, 18 to 72 points) at a mean of 14.5 months (range, eight to twenty-four months) postoperatively and remained at 40 points (range, 18 to 80 points) at the time of the latest follow-up at a mean of 7.8 years after surgery.
There was no significant change in wrist flexion or extension, radial deviation, ulnar deviation, or forearm supination between the preoperative and postoperative measurements. The mean wrist flexion-extension arc increased from 48° (range, 15° to 95°) preoperatively to 62° (range, 30° to 90°) postoperatively. The mean flexion increased from 32° (range, 15° to 55°) preoperatively to 42° (range, 30° to 55°) postoperatively (p = 0.06). The mean extension increased from 16° (range, 0° to 40°) preoperatively to 20° (range, –20° to 45°) postoperatively (p = 0.46). The mean radial deviation changed minimally, from 6° (range, –10° to 18°) preoperatively to 8° (range, 0° to 35°) postoperatively (p = 0.80). The mean ulnar deviation was 15° (range, 0° to 40°) preoperatively and 17° (range, –20° to 32°) postoperatively (p = 0.45). The mean supination increased from 50° (range, 0° to 90°) preoperatively to 71° (range, 40° to 90°) postoperatively (p = 0.12). The mean pronation increased significantly from 54° (range, 0° to 90°) preoperatively to 83° (range, 60° to 90°) postoperatively (p = 0.04). The increase in pronation and supination was likely a result of distal ulnar resection.
Radiographic Evaluation
On preoperative radiographs, nine of the twenty-four wrists were classified as Simmen Type I; five, as Type II; and nine, as Type III. Preoperative radiographs could not be located for one patient.
At the time of the latest follow-up, intercarpal fusion was seen in all ten wrists with retained functional prostheses. Five of the ten wrists that were treated with revision had radiographic evidence of intercarpal fusion prior to the revision. In the wrists that were not treated with revision, radiographic changes at the time of the latest follow-up included carpal component subsidence and osteolysis (one wrist), osteolysis without subsidence (one wrist), and carpal component subsidence without osteolysis (one wrist). There was no radiographic evidence of implant breakage or radial component subsidence.
Complications
Short-term complications included wound-healing problems (17%; four of twenty-four wrists) and persistent prosthetic instability (4%; one of twenty-four wrists). Superficial skin necrosis or delayed superficial wound-healing occurred in four patients (four wrists; 17%), but all lesions healed with dressing changes. There were no deep or periprosthetic infections. Recurrent prosthetic dislocation occurred in one wrist. Dislocation was first identified two months postoperatively and recurred despite closed reduction and casting. The patient underwent carpal component revision two months after the index procedure, but redislocation recurred and was treated with wrist arthrodesis 2.2 years after the index procedure.
Revision Surgery
Nine patients (ten wrists) underwent a total of twelve surgical revision procedures (see Appendix). One prosthesis (Case 1) was revised to a wrist arthrodesis because of recurrent instability. That patient died of an unrelated cause 2.2 years after the index procedure. In the group of nineteen wrists with a minimum of five years of follow-up, nine prostheses (in eight patients) had revision surgery because of carpal component loosening (Figs. 1-A, 1-B, and 1-C). The mean time to revision surgery for carpal component loosening was 5.5 years (range, 2.7 to 9.3 years). Prior to revision, four of the eight patients had developed painful swelling of the wrist. Revision surgery revealed polyethylene wear, metallosis, and carpal component loosening in all of the wrists. There was no evidence of gross radial component loosening in any wrist at the time of revision, although resorption under the tray of the radial component often prompted revision of the distal radial component. With inclusion of the patient who died, ten (50%) of twenty prostheses had been revised by the time of the latest follow-up. The survival rates of the original implant at five and seven years were 75% (fifteen of twenty) and 60% (twelve of twenty), respectively (Fig. 2). Two of the wrists with carpal component loosening were revised to wrist arthrodeses 3.5 and 5.3 years following wrist arthroplasty with use of Steinmann pins and demineralized bone matrix for grafting. Both wrists had clinical and radiographic evidence of fusion one year after the arthrodesis procedure. A third wrist underwent arthrodesis with a Synthes wrist fusion plate 9.3 years after the index procedure. Less than one year of follow-up was available for this patient. Two wrists underwent revision of the carpal component and wrist capsule supplementation with allograft fascia lata at 2.9 and 3.1 years after the index procedure. One of these implants was stable at the time of the latest follow-up (Figs. 1-A, 1-B, and 1-C). The other wrist underwent a third procedure 5.3 years after the revision, with removal of the carpal component and fascia lata allograft interposition.Fig. 1-CPosteroanterior radiograph, made two years after carpal component revision with bone-grafting, showing that the prosthesis has remained stable.Fig. 2Kaplan-Meier survivorship curve for the Universal wrist prosthesis.
Five wrists had insufficient carpal bone stock to support a revision carpal component. These wrists underwent removal of the carpal component, bone-grafting of the carpal defect with use of allograft corticocancellous chips mixed with demineralized bone matrix, and insertion of rolled allograft fascia lata as an interposition between the carpus and the radial component at an average of 6.7 years (range, 4.2 to 8.3 years) after the index procedure. In two of these wrists, the radial component was revised to a Universal 2 component (Integra LifeSciences, Plainsboro, New Jersey) because of bone resorption around the radial component. Despite resorption beneath the distal articular surface of the radial implant, all implants were well fixed in the radial shaft at the time of surgery. The stable radial component was retained in the remaining three wrists. The wrist capsule was carefully closed and the wrist was immobilized in a cast for two to three months after the interposition procedure. The salvage procedure alleviated pain but led to reduced wrist extension, decreased strength, and a resting posture in some ulnar deviation.
There was no significant difference between the ten revised and ten nonrevised wrists in terms of age at the time of surgery (p = 0.92), Simmen classification (p = 0.62), dominance (p = 0.86), or preoperative DASH score (p = 0.56). At the time of the latest follow-up, ten of ten wrists with a functional prosthesis had radiographic evidence of intercarpal fusion. Of the nine wrists that were revised because of carpal component loosening, four did not have radiographic evidence of intercarpal fusion at the time of revision. There was no significant difference between these two groups (p = 1.0).
Wrist arthroplasty offers pain relief with preservation of motion in patients with severe rheumatoid arthritis of the wrist. Compared with patients who are managed with wrist arthrodesis, those who are managed with wrist arthroplasty have less difficulty with personal hygiene and fine dexterity that requires wrist flexion7. Just a few degrees of wrist motion can result in substantially increased digit positioning in space3. Most patients with a wrist arthroplasty on one side and a wrist arthrodesis on the contralateral side prefer the wrist on the side of the arthroplasty8. Unfortunately, wrist arthroplasty has resulted in a high number of complications and implant failures. A recent meta-analysis of wrist arthroplasty and wrist arthrodesis in patients with rheumatoid arthritis demonstrated an overall complication rate of 30% for wrist arthroplasty and 17% for wrist arthrodesis1. In our series of patients with rheumatoid arthritis, 50% of implants had failed by the time of the latest follow-up.
Patients in our study in whom the implant did not fail maintained a functional range of motion of the wrist. Performance of most activities of daily living requires a relatively modest wrist flexion-extension arc of 35° and a radial-ulnar deviation arc of 25°9. The mean postoperative flexion-extension arc in the present series was 62° (range, 30° to 90°), and only one of our patients had a flexion-extension arc of <35°. The mean postoperative radial-ulnar deviation arc in our patients was 25° (range, 15° to 40°). There was a significant increase in pronation postoperatively, but there was no significant improvement in wrist flexion or extension, radial deviation, ulnar deviation, or supination postoperatively. The average flexion-extension arc of 62° and the average radial-ulnar deviation arc of 25° are similar to previously reported results with this prosthesis3.
Despite the theoretical advantages of wrist arthroplasty, enthusiasm for this procedure has been tempered by reports of early complications, with few reports of long-term results. Early silicone wrist replacements were prone to breakage and osteolysis10. Later nonanatomic designs such as the Meuli wrist prosthesis have been affected by wrist imbalance, resulting in substantial ulnar deviation and prosthesis dislocation in 8% of patients11.
Early failures of previous prostheses prompted changes in design to improve wrist stability. The Universal prosthesis has a large articular surface area designed to resist wrist imbalance. Nevertheless, Menon reported that volar dislocation (prevalence, 14%) was the most common complication in his series of patients who were managed with the Universal prosthesis5. One of our patients experienced wrist instability and was ultimately managed with revision to a wrist arthrodesis.
Carpal component loosening has been a common mode of failure of previous wrist prostheses11,12 and was the most commonly encountered complication in our series. Nine wrists (45%) in eight patients underwent revision secondary to carpal component loosening, and an additional two patients had radiographic evidence of carpal component subsidence. After a mean duration of follow-up of 6.7 years, Menon reported that thirty-four of thirty-seven Universal wrist prostheses remained in place, for a 92% survival rate5. None of his patients had evidence of carpal component loosening at the time of the latest follow-up, whereas 45% of the wrists in our study had a revision because of loose carpal components. Menon emphasized the importance of achieving intercarpal fusion “to [form] a solid bony support for the carpal component and [reduce] the chance of loosening by decreasing the toggle.”5 A higher rate of intercarpal fusion may have contributed to the lower carpal loosening rate in that study, as Menon reported radiographic evidence of intercarpal fusion in his patients but did not supply specific numbers5. The patients in our study who had carpal component loosening were less likely to have radiographic evidence of intercarpal fusion than those who had stable prostheses, although this difference did not reach significance.
Carpal component failure, metallosis, and polyethylene wear debris were always seen at the time of revision surgery. A finite-element computer model has shown that the toroidal articulation of the Universal implant is subject to low contact areas and high peak pressures when moved even small degrees from the neutral position of the wrist13. Such motion in vivo could explain the high rate of polyethylene wear. In the absence of a solid intercarpal fusion, wear particles could more easily migrate distally and contribute to osteolysis and carpal component loosening.
To address the problems of polyethylene wear and carpal component loosening, the next generation of the Universal prosthesis (the Universal 2 prosthesis) utilizes an ellipsoidal articulation rather than the toroidal articulation seen in the first Universal prostheses. Although the change in shape may appear subtle, in a finite-element model, the ellipsoidal articulation resulted in lower peak contact pressures and greater implant stability than the toroidal articulation did13. In addition, the central stem of the Universal 2 carpal component has a porous coating to promote osseointegration. In what we believe to be the only published study of this implant14, the failure rate was much lower than that in our series. In that study, van Winterswijk and Bakx reported carpal component loosening in one of seventeen wrists after twenty to seventy-four months of follow-up.
The five-year survival of the implants in our series was lower than that reported previously for the BIAX Total Wrist System (DePuy Orthopaedics, Warsaw, Indiana)8,15. Carpal component loosening was the most common cause of failure reported by Cobb and Beckenbaugh15, resulting in revision of eight (13%) of fifty-seven biaxial wrist prostheses with an additional six implants showing radiographic signs of loosening at the time of the latest follow-up after a minimum of two years. Takwale et al. found radiographic evidence of loosening in fourteen of sixty-three BIAX wrist prostheses after a minimum duration of follow-up of only twelve months, resulting in revision surgery in five wrists8. Symptomatic soft-tissue imbalance was also noted in eleven wrists in the study by Takwale et al.8, and seven wrists in the study by Cobb and Beckenbaugh had radiocarpal subluxation or dislocation15.
A major limitation of the present study is the small number of patients. Total wrist arthroplasty is not a common procedure, and accumulating the large number of patients necessary for detailed analysis of possible factors leading to failure of a wrist arthroplasty prosthesis would likely require the cooperation of multiple institutions.
In conclusion, wrist arthroplasty with use of the Universal prosthesis for the treatment of rheumatoid arthritis resulted in a high rate of carpal component loosening. Our failure rate was higher than previously reported for this prosthesis. Carpal component loosening remains an obstacle to predictable good long-term survival of this implant in patients with rheumatoid arthritis.
A table showing the revision procedures performed on nine patients (ten wrists) is available with the online version of this article on our web site at jbjs.org.
Cavaliere
CM;
Chung
KC. A systematic review of total wrist arthroplasty compared with total wrist arthrodesis for rheumatoid arthritis. Plast Reconstr Surg.
2008;122:813-25.[CrossRef][PubMed]
Adams
BD. Complications of wrist arthroplasty. Hand Clinics.
2010;26:213-20.[CrossRef]
Radmer
S;
Andresen
R;
Sparmann
M. Total wrist arthroplasty in patients with rheumatoid arthritis. J Hand Surg Am.
2003;28:789-94.[CrossRef]
Divelbiss
BJ;
Sollerman
C;
Adams
BD. Early results of the Universal total wrist arthroplasty in rheumatoid arthritis. J Hand Surg Am.
2002;27:195-204.[CrossRef]
Menon
J. Universal Total Wrist Implant: experience with a carpal component fixed with three screws. J Arthroplasty.
1998;13:515-23.[CrossRef]
Zangger
P;
Kachura
JR;
Bogoch
ER. The Simmen classification of wrist destruction in rheumatoid arthritis. Experience in patients with early disease. J Hand Surg Br.
1999;24:400-4.[CrossRef]
Murphy
DM;
Khoury
JG;
Imbriglia
JE;
Adams
BD. Comparison of arthroplasty and arthrodesis for the rheumatoid wrist. J Hand Surg Am.
2003;28:570-6.[CrossRef]
Takwale
VJ;
Nuttall
D;
Trail
IA;
Stanley
JK. Biaxial total wrist replacement in patients with rheumatoid arthritis. Clinical review, survivorship and radiological analysis. J Bone Joint Surg Br.
2002;84:692-9.[CrossRef]
Palmer
AK;
Werner
FW;
Murphy
D;
Glisson
R. Functional wrist motion: a biomechanical study. J Hand Surg Am.
1985;10:39-46.
Kistler
U;
Weiss
AP;
Simmen
BR;
Herren
DB. Long-term results of silicone wrist arthroplasty in patients with rheumatoid arthritis. J Hand Surg Am.
2005;30:1282-7.[CrossRef]
Cooney
WP
3rd;
Beckenbaugh
RD;
Linscheid
RL. Total wrist arthroplasty. Problems with implant failures. Clin Orthop Relat Res.
1984;187:121-8.
Talwalkar
SC;
Hayton
MJ;
Trail
IA;
Stanley
JK. Management of the failed biaxial wrist replacement. J Hand Surg Br.
2005;30:248-51.[CrossRef]
Grosland
NM;
Rogge
RD;
Adams
BD. Influence of articular geometry on prosthetic wrist stability. Clin Orthop Relat Res.
2004;421:134-42.[CrossRef]
van Winterswijk
PJ;
Bakx
PA. Promising clinical results of the Universal total wrist prosthesis in rheumatoid arthritis. Open Orthop J.
2010;4:67-70.
Cobb
TK;
Beckenbaugh
RD. Biaxial total-wrist arthroplasty. J Hand Surg Am.
1996;21:1011-21.[CrossRef]