Fracture in an arthrodesed Charcot’s knee joint

Discussion

Charcot’s joint is a neuro-arthropathy resulting in a functionally impaired and unstable joint without pain. Before the era of penicillin, neuro-syphilis was most commonly attributed to Charcot’s joint. Diabetes mellitus is one of the leading causes of Charcot’s joints. The predilection for joint involvement is variable and poorly understood. The knee joint is commonly affected by neuro-syphilis. Foot and ankle are affected more in diabetic when compared with non diabetics. In patients with syringomyelia, shoulder and elbow involvement is higher. Overall, the ankle and foot are most commonly affected in the musculoskeletal system, followed by knee, hip, shoulder and elbow joints.3 The prevalence of Charcot’s joint in patients with spina bifida is extremely rare (one in every 100 cases).4

The pathogenesis is understood by a neurotraumatic and neurovascular theory.5 In Charcot’s joint, there is primary degeneration of the dorsal ganglion. It results in loss of proprioception and progressive damage to the weight-bearing axis, resulting in repeated microtrauma leading to arthropathy.5 Charcot’s joint is thought to be a neural mediated continuous process of inflammation and repair resulting in severe osteolysis that results in fractures and dislocations at the affected joint. The neurovascular theory hypothesis suggests hypervascularity in the subchondral bone resulting in vascular shunting and subsequent osteopenia.5 Eichenholtz classified Charcot’s joint into three stages: (1) development, (2) coalescence and (3) reconstruction and reconstitution. Stage 1 shows localised osteopenia, periarticular debris and fragmentation. In Stage 2, there is early sclerosis, absorption of periarticular debris and bony consolidation noted. In the final stage (Stage 3), there is smoothening of the bony fragments resulting in fibrous or bony ankylosis.6

Treatment for a neuro-arthropathic knee joint is challenging. Eichenholtz proposed protected weight-bearing along with the use of cylindrical casts, long knee splints and pneumatic braces in Stages 1 and 2 of the disease. However, surgical management is more definite and relieves the symptoms by providing a stable and functional joint. Stage 3 cases are managed surgically that include arthrodesis, arthroplasty (excision and replacement).6 Arthrodesis of the involved joint is the most preferred treatment option, and arthroplasty is rarely done in a neuro-arthropathic joint. Though arthroplasty is considered a contraindication for Charcot’s joint, some considerations in patient selection have shown promising results. Possible complications of early prosthesis loosening, patellar dislocations, recurrent instability and peri-prosthetic fracture outweigh the benefits of an arthroplasty.7 Long-term results of a case series reported by Babazadeh et al and Bae et al have shown satisfactory results.8 9 Therefore, the choice of Total knee arthroplasty (TKA) in Charcot’s joints requires a thorough assessment and selective patient criteria; though clear guidelines are not available for it currently. Yoshino et al have suggested implantation in the reconstructive stage (Stage 3) with constrained long component implants, augmenting bony defects and good ligamentous balancing has given them better results.10

Achieving a sound fusion/arthrodesis in a Charcot’s joint is often challenging in an underlying neuropathic condition. Often fusion results in pseudoarthrosis due to inadequate bone stock, no contact of the solid bone, existing infection or presence of synovium even after fusion.3 We encountered similar problems when our patient was operated on for the first time for arthrodesis. Hence, we recommended a complete and thorough excision of the synovium with complete contact of the bony surfaces and additional bone grafts at the site of arthrodesis. Vaishya et al have reported successful elbow arthrodesis in a case following syringomyelia in the cervical cord. They achieved good bony fusion with the use of a long-moulded LCDCP, spanning the elbow joint.11

A fracture through a previously arthrodesed joint has not been reported in any joint, making this case a novel and rare case report. However, fractures through an ankylosed spine in ankylosing spondylitis patients have been reported even after minor trauma.12 A fracture through the fused knee in our case was akin to the fracture through the fused spine, being transverse like a ‘broken chalk’. Fracture in an ankylosing spondylitis is associated with osteoporosis in a fused ‘bamboo’ spine.13 The fracture line that results passes through all the three columns of the spine transversely entirely different from those sustained on a normal spine. We believe that in our case, the ‘fused knee’ was associated with osteoporosis, and a long-lever arm of the lower limb resulted in a fracture through it. There is a possibility of incomplete fusion somewhere in this arthrodesed knee, which was not obvious and became the weakest link of the lever arm of the lower limb, leading to a fracture.

Arthrodesis is also done in chronic intractable fulminant infection after TKA. A study conducted by Robinson et al showed 87% bony fusion of arthrodesis with a mean duration of 11.3 months following infective TKA.14 The overall reported rate of union varies from 73.7% to 100% for intramedullary nailing and 68.4% to 100% with an external fixator. The time to a union for knee arthrodesis in these cases ranges from 3.6 to 9.3 months.14 Their findings indicate that the union at the arthrodesis is variable and takes a long time. Therefore it requires continuous and longer follow-up.

Knee arthrodesis for a Charcot’s joint can be done using a long intramedullary nail (spanning the knee joint), ring (Ilizarov) fixator, plating, mono-axial external fixator or with Charnley’s clamps. The main objective of arthrodesis is to achieve stability by fusing the damaged joint in a functional position. Intra-medullary nail is a load-sharing implant that provides a rigid fixation. The nail is inserted with the opposing bony edges angulating at 10–20 degrees. These nails are custom made and are being increasingly used recently. But, these are not universally available, expensive and technically demanding. Open reduction and arthrodesis with Charnley’s clamp or Illizarov are also performed. Good expertise in the technique and instrumentation is required for Illizarov (ring) fixation. Hence, we preferred an open reduction and internal fixation using a broad LCDCP plating. This implant is commonly used for fracture fixation, is easily available, and not expensive. It provides a stable fixation across the joint. A few cases of peri-implant fractures have been reported, hence we suggest using a long-spanning plate.15 Wood and Conway16) have discussed the problem of nonunion after knee arthrodesis. They have advised to retain the intramedullary nail (if not broken and not infected), and do supplemental plate fixation, with bone grafting, in these cases. Supplemental plate provides a mechanical advantage as it increases the bending and torsional stiffness. The plate needs contouring to accommodate the flares of the distal femur and proximal tibia. A broad locking plate is preferred with a minimum of four bicortical screw fixation in both the femur and tibia. The plate allows good bony contact and opposition of the bone ends and provides stability and good fixation.16

We believe that the surgical fixation of the fracture through previously arthrodesed joint is the treatment of choice, as a conservative treatment with a cast is not likely to succeed. However, the surgery is challenging as careful and detailed planning of the surgical procedure requires consideration of various parameters like the skin condition, prevention of infection, choice and placement of implant, stability of fixation, and the degree of osteoporosis.