Abstract
Avascular necrosis of the femoral head (AVN) is associated with collapse of the femoral head and arthritic degeneration of the joint. The combination of an implant inserted into the femoral head that provides mechanical support and bone grafting to promote bone formation may offer a possible joint-preserving solution1. Seventeen such procedures were performed between November 2012 and March 2014 during an IRB approved clinical trial. Thirteen out of 18 patients remained unrevised at a minimum of 12 months; the results of radiographic and histological analysis of four revisions are presented.
The investigational device (Figure 1) was developed as a joint preserving treatment for AVN with a clinical grade of IIC or less according to the ARCO grading system2.
The device consisted of a braided spherical Nitinol cage with a Titanium / Nitinol orientation feature. It was implanted using fluoroscopic navigation into a spherical cavity cut into the femoral head via an 11mm diameter access tunnel. Once deployed, the implant was filled with a lightly impacted mixture of autologous bone graft and bone marrow soaked Conduit TCP (DePuy CMW, Blackpool, UK). The implant's purpose was to provide mechanical support to the weakened subchondral surface while the bone graft mixture re-integrated with the host bone.
The retrieved femoral heads were trimmed to leave approximately 3mm of bone around the implant, dehydrated, embedded in methacrylate resin, sectioned and thinned into 50–70µm coronal slices for histological analysis. The following observations were made (Figure 2):
Case 1 (Female, age 70, ARCO IIB, revised after 2 days): The patient was revised for spontaneous sub-trochanteric fracture secondary to osteoporosis. Contact between the native bone and bone graft was observed. Marrow elements and repair tissue were visible within the pores in the graft (Figure 2a).
Case 2 (Male, age 67, ARCO IIIC, revised after 82 days): Two wires were broken but retained within the braided structure. A radiolucent gap caused by the presence of fibrous tissue between the graft mixture and native bone was evident suggesting that the implant was unable to prevent progression in this case.
Case 3 (Female, age 70, ARCO IIC, revised after 482 days): The cavity penetrated the subchondral surface; at revision the implant was found to have breached the articular cartilage. There was partial separation of the proximal osteonecrotic fragment and no evidence of graft revascularisation or remodelling within the implant.
Case 4 (Male, age 42, ARCO IIC, revised after 469 days): There was no indication of bone graft re-integration. Collapse of the necrotic bone and deformation of the implant was diagnosed from 1 year follow-up x-rays.
Conclusion
This treatment has preserved the joints of fourteen patients. Of the four revised, two patients had clinical grades or bone quality contra-indicated for the device and three had lesions occupying more than 30% of the femoral head: Improved criteria for patient selection may be required. The device is only partially load-bearing and incapable of stabilising fractures: The radiolucent band associated with fibrous tissue formation may be an early indication of failure.
