Implant survival, adverse events, and bone remodeling of osseointegrated percutaneous implants for transhumeral amputees.

Background: Osseointegrated percutaneous implants provide direct anchorage of the limb prosthesis to the residual limb. These implants have been used for the rehabilitation of transhumeral amputees in Sweden since 1995 using a two-stage surgical approach with a 6-month interval between the stages, but results on implant survival, adverse events, and radiologic signs of osseointegration and adaptive bone remodeling in transhumeral amputees treated with this method are still lacking.

Questions/purposes: This study reports on 2- and 5-year implant survival, adverse events, and radiologic signs of osseointegration and bone remodeling in transhumeral amputees treated with osseointegrated prostheses.

Methods: Between 1995 and 2010, we performed 18 primary osseointegrated percutaneous implants and two implant revisions in 18 transhumeral amputees; of those, 16 patients were available for followup at a minimum of 2 years (median, 8 years; range, 2-19 years). These include all transhumeral amputees who have received osseointegrated prostheses and represented approximately 20% of the all transhumeral amputees we evaluated for potential osseointegration during that time; general indications for this approach included transhumeral amputation resulting from trauma or tumor, inability to wear or severe problems wearing a conventional socket prosthesis, eg, very short residual limb, and compliant patients. Medical charts and plain radiographs were retrospectively evaluated.

Results: The 2- and 5-year implant survival rates were 83% and 80%, respectively. Two primary and one revised implant failed and were removed because of early loosening. A fourth implant was partially removed because of ipsilateral shoulder osteoarthritis and subsequent arthrodesis. The most common adverse event was superficial infection of the skin penetration site (15 infections in five patients) followed by skin reactions of the skin penetration site (eight), incomplete fracture at the first surgery (eight), defective bony canal at the second surgery (three), avascular skin flap necrosis (three), and one deep implant infection. The most common radiologic finding was proximal trabecular buttressing (10 of 20 implants) followed by endosteal bone resorption and cancellization (seven of 20), cortical thinning (five of 20), and distal bone resorption (three of 20).

Conclusions: The implant system presented a survivorship of 83% at 5 years and a 38% 5-year incidence of infectious complications related to the skin penetration site that were easily managed with nonoperative treatment, which make it a potentially attractive alternative to conventional socket arm prostheses. Osseointegrated arm prostheses have so far only been used in transhumeral amputations resulting from either trauma or tumor. Their use has not been tested and is therefore not recommended in transhumeral amputations resulting from vascular disease. This method could theoretically be superior to socket prostheses, especially in transhumeral amputees with very short residual humerus in which the suspension of a conventional prosthesis is difficult. Comparative studies are needed to support its potential superiority. Moreover, the radiological findings in this study need to be followed over time because some of them are of uncertain long-term clinical relevance.