6.3 Kienböck’s disease

Kienböck described avascular necrosis of the lunate in 1910. Over twice as many men than women are affected, with onset usually in the twenties or thirties. Kienbock’s disease is usually unilateral. Although relatively uncommon, the pain, stiffness, and loss of strength may be devastating for otherwise healthy individuals, often in manual employment.

Kienböck’s disease results from vascular compromise to the lunate. Like the scaphoid, most of the surface of the lunate is covered by hyaline cartilage. However, the dorsal and volar poles have small areas with ligamentous and capsular attachments including blood supply. Lee originally described the arterial anatomy in 1963. Gelberman elaborated on this using latex injection; most lunates have vessels entering from both the dorsal and volar sides with intraosseous anastomosis.

The lunate has been described as the ‘keystone’ of the carpus. It transmits over 50% of the compressive load across the wrist joint. Approximately 80% of the load is transmitted through the radius and 20% through the ulna via the triangular fibrocartilage complex. The different compliance of the two interfaces may cause shear within the lunate.

The shape of the lunate may also affect force transmission; a triangular lunate may be more susceptible to microfracture than a square or rectangular lunate.

Vascular compromise is the common endpoint in Kienböck’s disease; however, the relative contributions of intrinsic anatomical variants versus extrinsic trauma are contentious.

Ulnar variance as a risk factor for Kienböck’s disease was first described in 1928 by Hultén, who compared the radiographs of 400 normal wrists with those of 23 patients with Kienböck’s disease; 23% of normal wrists were ulnar negative, in comparison to 74% of the Kienböck’s patients. Sixteen per cent of normal wrists were ulnar positive as opposed to none of the patients with Kienböck’s disease.

Some other clinical studies have confirmed this association, others have not. Some studies have questionable validity, poor radiographic standardization, or controls. The consensus remains that Kienböck’s disease is more common with an ulnar negative variance.

Biomechanical studies demonstrate the significant effect of changes in ulnar length on lunate strain; ulnar lengthening or radial shortening of 2mm more than halves lunate strain.

Kienböck speculated that injury may disrupt the lunate’s blood supply. The rarity of Kienböck’s disease following lunate and perilunate dislocations shows that the lunate can usually survive the loss of a major proportion of its vascular attachments.

External injury as an aetiological factor is suggested by increased frequency in manual labourers and in the dominant hand and by its infrequent bilateral occurrence. Clinical studies and cadaver models show susceptibility to fracture in hyperextension.

Vascular and traumatic susceptibility may be cofactors. The arterial supply has few or sometimes no anastomoses. Even a non-displaced stress fracture could disrupt this, causing devascularization. Conversely, a primary vascular injury would weaken the lunate, making it vulnerable to micro-shear fracture.

Patients usually present with a gradual onset of wrist pain, but may describe a precipitating injury.

The physical findings are non-specific, but consistent. Inspection may reveal dorsal wrist swelling. Pain and tenderness is localized in the mid-dorsal area. Kienböck described exacerbation of pain on percussion of the third metacarpal head longitudinally. Wrist flexion and extension are decreased, as are radial and ulnar deviation to a lesser extent. Forearm rotation is usually normal. Grip strength is typically approximately halved.

Confirmation of the diagnosis requires imaging; including plain radiography, radionuclide bone scan, computed tomography (CT) and/or magnetic resonance imaging (MRI).

Plain radiographs including posteroanterior and lateral views are essential. The posteroanterior view must be taken with the forearm in neutral rotation, for true assessment of ulnar variance.

Changes on plain radiographs follow a consistent pattern, as described in the staging section. Initially, the lunate may appear normal or may have subtle horizontal lines suggesting compression fracture. Next the lunate appears more dense, consistent with avascularity. Lucency may be seen below the subchondral bone, similar to that seen in avascular necrosis of the femoral head. As the lunate collapses and fragments, the capitate begins to migrate proximally (Figure 6.3.3). The scaphoid falls into palmar flexion, similar to the pattern seen with scapholunate dissociation. In advanced cases, degenerative changes of the radiocarpal and intercarpal joints are seen.

Radionuclide imaging is useful in early disease, when history and examination are suggestive but radiographs are normal (Figure 6.3.4). Increased uptake (‘hot spots’) may be seen in all three phases: flow (angiogram), immediate (blood pool), and delayed (bone), suggesting hyperaemia of the soft tissues and bony reaction or repair.

MRI scanning in Kienböck’s disease is now well established for early detection, assessment of bone viability, and evaluation of the response to treatment, rendering CT and bone scanning much less important (Figure 6.3.5).

Cancellous bone has adipose and haematopoietic tissue, and so high signal intensity on T₁-weighted images. Acute and chronic processes, including fracture, ischaemia, necrosis, and repair, alter marrow composition, decreasing signal intensity. MRI can identify lunate injury when plain films are still normal; the T₁ image abnormalities in Kienböck’s disease are consistent. MRI cannot reliably distinguish between necrosis and reactive tissue; the zone of avascularity may be exaggerated.

MRI is also used for postoperative evaluation; return to normal signal intensity is observed in patients who undergo radial osteotomies for diseased lunates with normal shape.

A diagnostic algorithm is shown in Figure 6.3.6. If history and examination suggest Kienböck’s disease, plain radiographs are taken. If they show changes consistent with avascular necrosis (i.e. sclerosis, fragmentation, collapse), then no further tests are necessary.

With normal radiographs, a bone scan or MRI is indicated; either can show Kienböck’s disease before plain radiographs. MRI is usually preferred to radionuclide imaging after normal plain films, as it can also show differential diagnoses and anatomical detail. Bone scanning remains useful for patients in whom MRI is contraindicated.

Staging is based on imaging and is divided into five distinct groups (Figure 6.3.7):

Treatment options depend on staging, and include immobilization, revascularization, lunate unloading procedures, intercarpal fusions, and salvage procedures. Analysis of results is complicated by the poor correlation of clinical with radiographic findings, as with many wrist disorders.

Radiographic deterioration is common. Long-term studies show that no lunate ever returns to normal and that degenerative changes developed in half even when they are immobilized. However, radiographic changes did not correlate with clinical symptoms; despite these findings, most had either no pain or pain only with heavy work.

In a series of patients treated with dorsal metacarpal vascular bundle implantation (the majority of patients also underwent lunate core debridement and iliac crest grafting combined with either external fixation or scaphotrapeziotrapezoid fusion) after a mean of nearly 6 years, 50 of 51 patients had decreased pain and average improvement in grip strength of 13kg.

These techniques seem promising, especially combined with other techniques to protect or unload the lunate, during revascularization.

Other procedures designed to reduce stress through the lunate include ulnar lengthening, radial shortening, radial wedge osteotomy, capitate shortening, and intercarpal fusion.

‘Joint-levelling’ procedures were developed in response to the perceived relationship between ulnar minus variance and Kienböck’s disease. Despite the controversy regarding the significance of variance, multiple biomechanical studies and computer models have convincingly demonstrated significant decreases in lunate strain and radiolunate contact stress with increases in relative ulnar length. Ulnar lengthening by 3mm reduces lunate strain by more than 70%, with 90% of that in the first 2mm.

Radial shortening has similar clinical success in patients with ulnar minus variance, and has fewer complications than ulnar shortening (Figure 6.3.8). The radius should be shortened to the level of the ulna, with compression plate fixation.

Long-term follow-up of patients with Kienböck’s disease who had radial shortening osteotomies show that grip strength was, on average, 90% of the unaffected side.

For individuals with positive ulnar variance, and for some with neutral variance, relative ulnar lengthening may risk creating ulnar abutment. Alteration in the radial inclination modifies load tran smission through the carpus without significantly changing ulnar variance.

Patients receiving osteotomies to decrease inclination are either pain free or have pain only with strenuous activity.

A variety of intercarpal fusions have been described that either directly (e.g. capitate shortening with capitate–hamate fusion) or indirectly (e.g. scaphotrapeziotrapezoid fusion) unload the lunate. They are particularly useful alternatives in ulnar positive variance or possibly when some lunate collapse has already occurred.

A 2- to 4-mm thick section of the capitate waist is removed with internal fixation of the remaining portions, generally combined with capitate–hamate fusion. Two-dimensional modelling of the procedure has predicted a 66% decrease in compressive load across the radiolunate articulation, with increases in scaphotrapezial load and triquetrohamate load.

Scaphotrapeziotrapezoid fusion has been used in Kienböck’s disease to unload the lunate indirectly (Figure 6.3.9); it also treats the rotation of the scaphoid seen in advanced cases. Biomechanical studies demonstrate that scaphotrapeziotrapezoid fusion with the scaphoid in neutral flexion reduces lunate strain by approximately 70%.

Patients treated by scaphotrapeziotrapezoid fusion had either no pain or mild pain with strenuous activity, but reduced range of motion (flexion 18%, extension 23%, radial deviation 69%); 13 of 15 returned to their original occupation. Some patients will require secondary excision of the lunate, and others will require late salvage procedures.

Scaphocapitate fusion can both unload the lunate and stabilize rotatory subluxation of the scaphoid. Cadaver and two-dimensional modelling studies predict significant unloading of both the capitolunate and radiolunate articulations. Clinically reports show wrist flexion and radial deviation are reduced by more than 60%, but there is good relief of pain and most patients returned to their original occupation. Although carpal height is maintained, some wrists show early radiocarpal degenerative change.

Isolated capitatohamate fusion has been used as a lunate unloading procedure. Although clinical success has been reported, cadaver studies and two-dimensional modelling do not demonstrate significant decreases in lunocapitate loading, radiolunate loading, or lunate strain.

Silicone lunate replacement, which was introduced by Swanson (1970), gained popularity for treatment of stages II to IV of Kienböck’s disease. It has also been used in combination with intercarpal fusions to limit the loading of the implant. The original implant, made of standard-grade elastomer, was later replaced by a high-grade elastomer with a deeper concavity to resist dislocation, and a titanium implant has also been designed, but not widely used. Despite early clinical successes, reports have continued linking silicone elastomer implants to particle debris synovitis. The long-term follow-up of patients with implants has shown radiographic and clinical deterioration.

Other primary treatments reported in the literature include sensory denervation of the wrist, forage of the distal radius, simple lunate excision, excision of the lunate with soft tissue interposition, and bone grafting of the lunate combined with external fixation.

Total wrist arthrodesis is the procedure of choice following failed treatment, particularly when secondary osteoarthritis has developed or if return to manual occupation is required. If the articular surfaces are relatively preserved and loss of motion is unacceptable, proximal row carpectomy is an alternative as both a primary or salvage procedure. Most patients will have either mild or no pain. Wrist motion is generally improved slightly while grip strength more than doubles.

As designs improve, third-generation total wrist arthroplasty is becoming an accepted alternative to arthrodesis in non-manual workers.

The diagnosis of Kienböck’s disease is based on history, examination, and imaging. Staging is based on plain radiographs, with additional information primarily from MRI.

For stages I, II, and III, the aim of treatment is prevention of further collapse during healing and revascularization. Protection or unloading of the lunate is achieved by various techniques, depending on ulnar variance. In negative or neutral ulnar variance, radial shortening has established effectiveness. Capitate shortening, radial wedge osteotomy, proximal row carpectomy and intercarpal fusions are alternatives for patients with positive variance.

In stage IIIB, a scaphocapitate or scaphotrapeziotrapezoid fusion also corrects rotatory scaphoid subluxation.

For stage IV a wrist arthrodesis is generally indicated.