A.H. Alexander and D.M. Lichtman
From : Orthopedic Clinics of North America, Vol 17, July 1986
A volume on controversial subjects in hand surgery would be incomplete without a section on Kienbock's disease. To date, its exact etiology and treatment of choice remain a matter of heated debate. Its various other names (lunatomalacia, aseptic necrosis, osteochondritis, traumatic osteoporosis, osteitis, and avascular necrosis of the lunate) reflect this controversy in etiology. The large assortment of surgical procedures for this condition (over 20) indicates the ongoing controversy in treatment selection. The purpose of this article is to sort through the various theories of etiology, describe a simple clas sification system, and, on the basis of this classification system, present a rational approach for the selection of appropriate treatment.
The diagnosis of Kienbock's disease should be considered in any patient presenting with wrist pain of uncertain origin, because in its earlxyn stage, the disease may be clinically indistinguishable from other causes of wrist pain. Roentgenograms during this stage are of little help as they may be negative. Classically, the Patient is 20 to 40 years of age and complains of wrist pain and stiffness of insidious onset usually following trauma. The male to female ratio is two to one. The disorder, though not rare, is uncommon, and the average orthopedist can expect to see a case every I to 2 years. The incidence of bilateral Kienbock's disease is extremely rare, there being few reports of this occurrence.
Usually the patient will note tenderness dorsally about the lunate, sometimes associated with synovial swelling consistent with localized sYnovitis. Early on, however, the patient may appear to have simply a wrist sprain. With progression, synovitis predominates and finally, in the late stage, arthritis.
Invariably, the grip strength is significantly decreased cornpared with that of the normal hand, and the range of motion of the wrist usually lessens. The diagnosis is established through radiographs, particularly in the later stages of the disease when the sclerotic appearance of the lunate is so characteristic. As noted, early in the course of Kienbock's disease, the radiographs may actually be normal. Because of the varying appearance of the radiographs of a patient with Kienbock's disease, at least two classification systems have been devised. (discussed in detail under "Staging") represents a simple classification of Kienbock's disease that is useful for determining the degree of involvement and for choosing appropriate treatment based on the extent of involvement.
ETIOLOGY
Peste, who first described collapse of the carpal lunate in 1843, did so before the advent of the x-ray. He believed the lesion to be a fracture with a traumatic etiology In 1910, Kienbock also ascribed this lesion to trauma. However- Kienbock believed that repeated sprains, contusions, and subluxations led to ligamentous and vascular injury resulting in loss of blood supply to the lunate. Since then, numerous authors have described the pathologic changes as avascular necrosis.
In 1928, Hulten noted that 78 per cent of his patients with Kienbock's disease had a short ulna, whereas only 23 per cent of normal patients had a short ulna. He called this condition ulna minus variant. Since this discovery, many others have confirmed negative ulnar variance in their patients with Kienbock's disease. Theoreticallv, a short ulna relative to the distal articular surface of the radius causes increased shear forces on the ulnar side of the wrist and particularly on the lunate. This is thought to be a Contributing factor in the development of avascular necrosis.
Gelberman and associates described a method for establishing the degree of ulnar variance. This is done by extending the line from the distal radial articular surface toward the ulna and measuring the distance between this line and the carpal surface of the ulna. Palmer and colleagues found that the position of the distal ulna in relation to the distal radial surface changes with varying degrees of forearm rotation and that the change in variance was least with the elbow flexed 90 degrees.- They then utilized a template of concentric circles (similar to the one used in establishing sphericity in Legg-Perthes disease). The concentric circle that best approximates the distal radial surface is selected as a reference and compared in millimeters with the carpal surface of the ulna. The importance of accurate measurement of ulnar variance is enhanced by the recent gain in popularity of ulnar lengthening and radial shortening techniques to treat Kienbock's disease.
Acute fracture or trauma as an etiology has been implicated in many series as the majority of patients report a history of injury predating the exacerbation of symptoms. Beckenbaugh and associates found lines suggestive of fracture on radiographs of 82 per cent of their patients. More and more investigators are documenting the presence of these fractures in Kienbock's disease, particularly with tomography; however, it remains unclear whether the fractures incited the avascular event or the avascular necrosis predisposed the pathologic bone to these fractures.
Stahl believed that traumatic compression fracture led to avascular necrosis in a Innate with an already tenuous blood SUpply. Lee found three vascular patterns in cadaver lunates: a single vessel, either volar or dorsal, supplying the entire bone; several vessels at both volar and dorsal surfaces of the lunate without central an-astomosis; and several vessels at both volar and dorsal surfaces of the Innate with central anastomosis. According to Lee, the first two, then, would be at greater risk of developing Kienbock's disease. Injection studies by Panagis and colleagues support this contention,- as they found a single palmar nutrient vessel in 20 per cent of fresh Innate specimens.
Gelberman and associates also studied the extraosseous and the intraosseous blood supply of the Innate in fresh specimens. They found the extraosseous supply to be extensive, with branches of the radial and anterior interosseous arteries forming a dorsal Innate plexus. Branches of the radial, ulnar and anterior interosseous arteries and recurrent deep palmar arch arteries form a volar plexus. In most specimens, vascularity reached the Innate through one or two foramina from the volar plexus and one or two foramina from the dorsal plexus. In 7 per cent of specimens there was only a volar contribution. These lunates theoretically would be at greatest risk for a single traumatic event (for example, fracture or dislocation of the carpus) resulting in avascular necrosis. However, seldom is this history present in a patient with Kienbock's disease. Gelberman and colleagues found that the intraosseous blood supply consisted of three patterns: Y in 59 per cent, I in 31 per cent, and X in 10 per cent with the dorsal and volar anastomosis just distal to the center of the lunate .
Evaluation of the terminal vessels in the Iunate allowed Gelberman and colleagues to conclude that the proximal subchondral bone, adjacent to the radial articular surface, was least vascular. Because of the rich extraosseous blood supply, they discounted the theory held by some that interruption of vessels entering a single pole of the lunate caused avascularity. On the basis of this work, Gelberman and coworkers suggest that intraosseous disruption of vascularity due to repeated trauma with compression fracture causes Kienbock's disease.
In summary, current thinking on the etiology of Kienbock's disease is that acute trauma or repeated minor trauma due to excessive shear force (enhanced by an ulnar minus variance) leads to interruption of the blood supply to the susceptible or "at risk" Innate. Avascular necrosis results. The susceptible Innate is one that has a single nutrient vessel supplying the entire bone or a limited intraosseous blood supply.
DIAGNOSIS
Kienbock's disease is an isolated disorder of the lunate diagnosed from characteristic roentgenographic density changes, often accompanied by fracture lines, fragmentation, and progressive collapse. It should be distinguished from other causes of wrist pain and swelling particularly in the early stages when the roentgenograms may be negative. Disorders to be ruled out include rheumatoid arthritis, post-traumatic arthritis, synovial-based inflammatory disease, acute fracture, carpal instability, and ulnar abutment syndromes. The radiographic hallmark of increased density typically seen in Kienbock's disease should be distinguished from transient vascular compromise. White and Omer recently described this radiographic condition seen following fracture-dislocation or dislocation of the carpus. In 3 of 24 patients sustaining this injury, there was a postinjury transient increase in lunate radiodensity that could have been confused with Kienbock , s disease. This radiodensity lasted from 5 to 32 months and should be treated expectantly.
In more severe Kienbock's disease, as the lunate collapses, there is proximal migration of the capitate, widening of the proximal carpal row, and, frequently, rotation of the scaphoid, causing it to appear foreshortened on anteroposterior radiographs. This foreshortening has been referred to as the "ring" sign Tomograms may be helpful in identifying the linear fractures or localized areas of sclerosis not readily apparent on plain radiographs. Scintigraphic imaging may be of benefit in patients who have otherwise negative radiographs.
Rarely has Kienbock's disease been reported in association with other conditions. There are case reports of Kienbock's disease in sickle cell disease carpal coalition, and gout. One article identified streptococcal infection in several cases and attempted to cite this organism as etiologic. Rooker and Goodfellow found five cases of Kienbock's disease in a group of 53 adults with cerebral palsy. An abnormally flexed wrist posture was the common feature in all five cases, suggesting that this extreme posture compromised the blood supply to the lunate.
Once the diagnosis of Kienbock's disease is established, the degree of involvement should be determined in order to assist in guiding one through the maze of treatment options. Casting for 2 to 3 weeks in uncertain cases will usually reveal the diagnosis by relative disuse osteoporosis of the adjacent carpal bones.
Stahl 's original classification of Kienbock's disease has been modified by Lichtman and colleagues 50 and consists of four stages.
Stage 1. Roentgenograms are normal except for the possibility of either a linear or a compression fracture. Unless this compression fracture is visible, this stage is clinically indistinguishable from a wrist sprain. Scintigraphic imaging may be helpful.
StageII. There are definite density changes apparent in the lunate relative to the other carpal bones; however, the size, shape, and anatomic relationship of the bones are not significantly altered. Significant fracture lines may be noted. Later in this stage, anteroposterior roentgenograms show loss of height on the radial side of the lunate. The patient exhibits symptoms of recurrent pain, swelling, and tenderness in the wrist.
Stage III. The entire lunate has collapsed in the frontal plane and is elongated in the sagittal plane. The capitate migrates proximally. Scapholunate dissociation, rotation of the scaphoid (ring sign), and ulnar deviation of the triquetrum may be seen on the anteroposterior roentgenograms. To better assess the degree of collapse in stage III, one should establish the carpal height ratio. Carpal height is the distance between the base of the third metacarpal and the distal radial articular surface as determined on a posteroanterior roentgenogram of the wrist. The carpal height ratio is defined as the carpal height divided by the length of the third metacarpal. In normal individuals, this ratio is 0.54 ± 0.03. Carpal height ratio is becoming more important, as the factors determining results of treatment in stage III appear tied to the degree of collapse. Clinically, patients in this stage have the same symptoms as those in stage II, but with increased level of wrist stiffhess.
Stage IV. All findings characteristic of stage III are present as well as generalized degenerative changes in the carpus.
TREATMENT
Kienbock's disease may be treated by immobilization, revascularization, ulnar lengthening or radial shortening, simple excision, silicone replacement arthroplasty (SRA), softtissue replacement arthroplasty, limited intercarpal fusion, or salvage procedures.
Immobilization. Prolonged immobilization of the wrist has been tried in all stages of Kienbock's disease. Stahl advocated immobilization, yet in some series, it has been shown to lead to continued collapse of the lunate or to otherwise unsatisfactory results due to the need for prolonged treatment. Lichtman and colleagues reported on 22 patients treated with immobilization, 17 of whom had progressive collapse while immobilized and 19 of whom had unsatisfactory results. For stage 1, however, immobilization may be indicated in hopes that the vascular insult is kept to a minimum and that the lunate is given a chance to heal. Because diagnosis in this stage often is difficult, a trial period of immobilization may result in the characteristic radiographic changes that establish the diagnosis. Since there are effective methods of treating later stages, delay in providing more aggressive treatment may also be warranted in stage II, depending on the patient's symptoms, desires, and occupational requirements.
Revascularization. In stage 11 Kienbock's disease, before the lunate has collapsed, it is possible for the lunate to regain blood supply. Braun has described a method by which a small piece of volar radial bone, still attached to the pronator quadratus muscle, is grafted to the avascular lunate. This procedure is done through a volar approach with division of' the palmar and transverse carpal ligaments in order to mobilize the median nerve. After the volar wrist is entered tile lunate is burred with a highspeed drill in preparation to receive a 1-1.5-cm piece of radial bone still attached to the pronator
quadratus muscle. This is secured with pull-out wires, Lind the fixation is augmented with transarticular Kirschner wires. Braun has found this to he successful in 7 of 8 patients, with the longest follow-tip being 7 years. He recommends the procedure be done only in those cases without significant collapse.") Chacha has also reported on this technique in three patients. A similar revascularization procedure uses the pisiform. Pisiform transfer on its vascular pedicle was reported by Erbs and Bohm in 32 patients . They found uniformly good results at 5-year followtip. Presumably, most of these were in stage II of the disease. Of 14 patients with "advanced" Kienbock's disease, however, 50 per cent became symptom-free or had pain only under stressful conditions. Eckardt reported the same technique with good results in two patients with 4-year follow-up for stage 1.
Hod and colleagues described direct transplantation of a vascular bundle into the vascular lunate They had successful results in eight of nine patients.
It must be remembered, however, that none of the revascularization procedures are likely to work in the face of severe collapse (stage III), for even if they are successful in re-establishing blood supply, lunate height and normal carpal kinematics will riot be restored.
Ulnar Lengthening and Radial Shortening. On the basis of the theory that ulnar minus variance is a significant etiologic factor in Kienbock's disease, some have advocated equalization of the distal articular surfaces by either ulnar lengthening or radial shortening. Both of these procedures have bad good results reported. However, it seems unlikely that leveling of the distal artictilar surfaces of the radius and lunate can restore an already collapsed lunate; therefore, these procedures remain questionable in advanced stage III.
Both radial shortening and ulnar lengthening require osteotomy. A segment of bone is removed when radial shortening is done and a segment of bone graft inserted when ulnar lengthening is done. After either, fixation is usually accomplished with a compression plate. It is generally recommended that the ulnar variance be changed to I- to 2-mm positive variance by placing an appropriate-sized interpositional graft during ulnar lengthening. Sundberg and Linscheid found this successful in all but I of 19 patients followed for an average of 8.2 months. They reported no nonunions. Ulnar lengthening per se "burns no bridges," and therefore further treatment in the event of failure is not precluded.
Radial shortening may be preferable to some because it does not require a second surgical incision to harvest bone graft. Like ulnar lengthening, it burns no bridges. Almquist and Burns obtained good results in 11 out of 12 patients (minimum follow-up 5 years) and noted that the literature reported good results in an additional 58 of 67 patients (87 per cent) treated with radial shortening.
Excision of the Lunate. Lunate.excision was one of the first surgical procedures for Kienbock's disease. The rationale of this procedure is to remove sequestered bone that is provoking painful synovitis. Some have reported good results from simple excision others criticize the operation predicting late proximal migration of the capitate. There is at least one report of the same results from either excision or immobilization . Nahigian- and colleagues ccmbined simple excision with dorsal capsular flap arthroplasty to prevent migration of the capitate and reported good results in four patients. Schmitt and associates reported a similar technique of capsuloplasty using "epitendinous" tissue from the flexor tendons to fill the gap.-, I They reported 80 per cent satisfactory results in 42 cases. Another similar technique is excisional arthroplasty and replacement with a rolled tendon graft (palmaris longus). This is performed much like the procedure of metacarpotrapezial joint arthroplasty described by Froinison . Ishiguro also reported on the use of autogenous tendon graft (generally the palmaris longus, plantaris, or portion of flexor carpi radialis) placed in the bed of the excised lunate. Twenty-four of 26 patients with average followup of 2.5 years were satisfied with their result. Using the criteria of Lichtman and coworkers, Ishiguro, further noted that 6 of 10 stage III patients and 11 of 16 stage IV patients had satisfactory results. More recently, Kato and colleagues treated patients with either silicone replacement arthroplasty or a "coiled palmaris longus tendon" and concluded that the latter is preferable once carpal collapse has occurred . Because in stage I and stage II Kienbock's disease there still exists a chance for lunate revascularization, soft-tissue (palmaris longus) replacement arthroplasty is best reserved for stages III and IV.
Silicone Replacement Arthroplasty. SRA is simply another way to prevent migration following Innate excision. There are many advocates of this procedure. After excision, the lunate may be replaced by a handcarved Silastic wafer or, more commonly, by a Swanson design carpal Innate manufactured from high-performance silicone elastoine (Dow-Corning, Midland, Michigan). SRA is no indicated in stage IV disease, once pan carpal arthrosis is present. In stages II and III, how ever, SRA has proved successful in many series, though prior to the development of the latest designed implant, dislocation of the prosthesis contraindicated BRA in stage III. Subse quently the deeper concavity for capitate articulation and the more anatomic design of the newer implant have reduced the risk of dislocation.
The advantages of SRA include rapid rehabilitation and return to work. It is also technically an easy operation to perform, and it does not require a second operation. However, recent reports of silicone synovitis, periarticular cyst formation, and foreign-body giant-cell reaction are disturbing.
The preceding reports have led the manufacturer to issue a warning that "wear particles from silicone elastomer implants... may participate in or exacerbate synovitis or bone cyst complications in contiguous bone. "
It is our impression that most cases of silicone synovitis are seen in young active patients, patients with preoperative cysts and/or degenerative changes, patients who, had temporary Kirschner wire or suture fixation of the implant, and/or those who had postoperative implant instability. Therefore, SRA should not be done in the young patient with extreme functional demands or in patients with dorsal carpal intercalary segment instability (DISI), unless it is combined with a procedure that reduces the bearing and shear stress and concomitant microfragmentation of the implant surface (for example, radial shortening, ulnar lengthening, or limited carpal arthrodesis). Furthermore, the surgeon must avoid Kirschner wire or suture fixation and must achieve intraoperative stability of the Innate implant.
Limited Intercarpal Fusion. Graner and colleagues described arthrodesis of the lunate to adjacent carpal bones for advanced Kienbock's disease-presumably stage III or stage IV. They considered it as treatment only if conventional methods failed to relieve symptoms. Its most important advantage is that radiocarpal motion is maintained, unlike with complete wrist arthrodesis. In this series, 18 patients with a 22-month average follow-up underwent limited intercarpal arthrodesis; all had satisfactory results. Patients with severe fragmentation of the lunate undergo resection of the necrotic bone, osteotomy of the capitate in its midportion, and proximal displacement of the proximal capitate fragment , which is secured to the scaphoid and triquetrum with bone pegs. Essentially, the space vacated by excision of the Innate is filled by the proximal half of the capitate, and the space left by osteotomy of the capitate is filled by autogenous bone graft. The procedure is then completed by performing arthrodesis of contiguous surfaces of the hamate, capitate, scaphoid, and triequetrum by first denuding articular surfaces and securing the bones with small cortical bone pegs. When the lunate remains suitably intact, osteotomy of the capitate is omitted, as is lunate excision, and the contiguous surfaces of the lunate, scaphoid, triquetrum, hamate, and capitate are arthrodesed .
Capitohamate Fusion. Chuinard and Zeman,-, advocated capitohamate fusion as a method of preventing proximal capitate migration in the face of a collapsed Innate. Depending on the degree of collapse, SRA was considered in conjunction with capitohamate fusion. Capitobainate fusion alone is not a method currently recommended by the authors, as the harnate is already bound to the capitate by strong ligaments; furthermore, collapse of ihe capitate is accorm)Anied bv proximal migration of the en tire distal carpal row (including the hamate) and widening of the proximal carpal row. Thus, it seems unlikely that capitohamate fusion alone can prevent proximal migration of the distal carpal row.
Triscaphe Fusion. Watson and colleagues described triscaphe arthrodesis with and without SRA Patients underwent fusion of the scaphoid, trapezium, and trapezoid, and half underwent concomitant SRA. In their article, Watson and colleagues graphically demonstrate clenched-fist roentgenograms of a patient with SRA showing a 22 per cent reduction (compression of the implant) in capitoradial space when compared to the relaxed position. They suggested that triscaphe arthrodesis is capable of supporting the remaining carpus either in the case of a collapsed Innate or when SRA is performed. Of 16 patients (average follow-up 20 months), 10 had complete relief of symptoms and 6 reported only postactivity ache in the affected wrist. We believe that' this procedure should be reserved for patients with a significant scapholunate dissociation and DISI deformity preoperatively.
Salvage Procedures. For severe Kienbock's disease (stage IV), good results have been reported from proximal-row carpectomy and wrist arthrodesis. As noted- limited intercarpal arthrodesis by Graner's method is another option. Denervation of the wrist joint has been described by several authors to be successful in relieving pain without impairing function or mobility.
PITFALLS AND COMPLICATIONS
Because of the varied treatments advocated for Kienbock's disease, it is apparent that no single treatment stands out as the best. Choice of treatment must be predicated on the experience of the surgeon; the desires, activity level, and goals of the patient; and the stage of the disease. Finally, if considered, the risks involved with each treatment, including pitfalls and complications, may help one to select the optimal treatment for an individual patient.
SRA has consistently yielded good results in both stage II and stage III, but growing concerns over silicone synovitis and progressive cystic degeneration in contiguous bones make it inadvisable as a sole procedure in the young patient with carpal instability or in the patient who will have great functional demands. The other significant risk with SRA is the complication of implant dislocation. This is best avoided by proper selection of implant size, preservation of the lunate volar shell of bone along with its soft-tissue attachments, and careful suture of the dorsal capsule. The 20- and 30-year follow-up of SRA is yet to be determined, as is the fate of the silicone implants over the long run in relatively young patients. In some of our patients, we have noted a slight amount of scapholunate dissociation accompanied by mild scaphoid rotation detectable on postoperative radiographs. To date, this seems not to have affected results of SRA, but these patients, too, will have to be followed to determine if the rotation progresses or if implant fragmentation and synovitis occur.
The pronator muscle pedicle flap is a tedious procedure requiring fastidious attention to detail. It is important to release the pronator fascia sufficiently in order to ensure adequate length of the muscle and to prevent its vascular compromise. Dislodgement of the graft has also been reported, and if one is using pull-out wires, it is possible to entrap the extensor tendons. This procedure, even if successful in re-establishing blood supply to the lunate, cannot be expected to restore lunate height and normal kinematics. Therefore, it is not recommended for advanced stage III.
Radial shortening and ulnar lengthening are relatively simple procedures and have yielded consistently good results in stage II and early stage III. There are a few drawbacks, though. With radial shortening, nonunion of the radius is possible. Furthermore, a second operation is usually required for plate removal. This operation as well as ulnar lengthening should not be done in the presence of significant collapse of the lunate, as they are unlikely to restore height and normal carpal kinematics. Their application in patients with neutral and positive ulnar variance is questionable, though Grassi and colleagues noted good results even in these patients.
Like radial shortening, ulnar lengthening may result in nonunion of the ulna or delayed incorporation of the bone graft, requiring prolonged immobilization. In this procedure, too, a second operation is required for plate removal. Because the ulna is lengthened, the anatomy of the distal radioulnar joint is affected and may be a source of postoperative discomfort. Care must be taken not to overlengthen the ulna. For the same reason, ulnar deviation may also be restricted by this procedure.
Following limited intercarpal arthrodesis, there may be moderate to marked limitation of the range of motion of the wrist. Loss of intercarpal motion may also lead to arthrosis of the radiocarpal joint.
In triscaphe arthrodesis, overcorrection of the scaphoid rotation can result in decreased range of motion and incongruity at the scaphoradial joint. Furthermore, if performed alone, SRA may be required as a second procedure. Overall, Watson and colleagues reported two of eight patients requiring delayed SRA after triscaphe arthrodesis. Also, the long-term results of triscaphe arthrodesis must be evaluated.
SUMMARY
Kienbock's disease is an isolated disorder of the lunate resulting from vascular compromise to the bone. The symptoms include wrist pain, limited range of motion, and decreased grip strength. The diagnosis is made from characteristic changes seen in the lunate on radiograms of the wrist. The severity of the disease can be categorized by staging the degree of involvement. This is helpful in guiding the practitioner through the maze of treatment options. Initial treatment of Kienbock's disease is conservative and includes immobilization, analgesics, and/or anti-inflammatory medication. If symptoms are not relieved, then based on the degree of involvement, several surgical options exist that will provide a successful result. These include autogenous tendon replacement arthroplasty, revascularization, radial shortening, ulnar lengthening, limited intercarpal arthrodesis, and silicone replacement arthroplasty. Salvage procedures for Kienbock's disease include wrist denervation, wrist arthrodesis, and proximalrow carpectomy.
Currently, we prefer immobilization for treatment of stage I Kienbock's disease. For stage II, a revascularization procedure may be attempted or ulnar lengthening/radial shortening done, particularly if there is significant negative ulnar variance. In stage III, replacement arthroplasty and/or limited interearpal arthrodesis is our treatment of choice, and for stage IV, one of the salvage procedures is indicated.