Local concentration of antibiotic at the site of infection is a major parameter for its efficiency. However, bone diffusion is poor leading either to their non-use (ex: gentamicin) or the use of high concentration (ex: vancomycin). Local administration could optimize their local concentration combined with lower side effects. We report the clinical experience and pharmacological results of an antibiotic loaded porous alumina used to replace infected bone in 4 patients. Two patients had a destroyed sternum following mediastinitis; one presented a femoral chronic osteomyelitis due to MRSA and one had an infected ankle arthroplasty. The ceramic was loaded with gentamicin in three cases and vancomycin for the ankle infection. Local dosages thanks to Redon's drain and blood samples were performed. Loading was done to protect the device while implanted in an infected area and was combined with conventional antibiotic therapy.Aim
Method
Only 1% of all primary bone tumours are situated in the distal humerus. Destruction of the distal humerus by tumour is rare and reconstruction of the distal humerus is challenging. Because of the amount of bone loss following tumour excision, excision arthroplasty or arthrodesis is impossible and hence some form of reconstruction is usually required. Allograft reconstruction and hemiarthroplasty are uncommon and lead to an unpredictable outcome. Ten patients underwent endoprosthetic replacement of the distal humerus for bone tumours over a thirty one-year period. There were 8 primary and 2 secondary tumours and male to female ratio was 2:3. Average age of the patients was 47.5 years (15–76 years). Mean follow up was 8 years (9 months - 31 years). Four patients required further surgery, three having revision for asceptic loosening and two of these and one other later needing a rebushing. There were no permanent nerve palsies, infections, local recurrences or mechanical failures of the implant. Four patients died of their disease between 12 and 71 months after operation, all with their prosthesis working normally. Average flexion deformity was 15 degrees (0–35) and average flexion of these patients was 115 degrees (110–135). The average TES Score for these patients was 73% (29% to 93%). The activities which the patients found to be no problem (TES score more than 4.5 out of 5) were: brushing hair, drinking from a glass, putting on make up or shaving, picking up small items, turning a key in a lock, doing light household chores and socialising with friends, whilst activities that proved difficult (TES score less than 3 out of 5) were: gardening and lifting a box to an overhead shelf. Pain was not a problem and only 1 of the surviving patients reported ever having to use regular analgesics.
The prosthesis was implanted 15.6 days (mean) after initial trauma and included a metal-backed acetabular implant (except one case) and a bone autograft (except three cases). Complete weight bearing was authorised on day 3 postop for five patients, and at six weeks for ten; at three months for the other three. Mean follow-up was 2.5 years. Thirteen surviving patients were reviewed clinically and radiographically. Five patients who had died were also included in the analysis using data reported by family before death. The Postel-Merle d’Aubigné (PMA) and Harris scores were recorded. Survival curves were plotted. Radiological assessment included bone healing, status of the prosthetic assembly, presence of a lucent line or signs of loosening in the three acetabular zones described by De Lee and Charnley.
Radiologically, all fractures had healed. There were no mechanical complications and no signs of acetabular loosening.
The absence of early deaths and the low rate of intervention-related morbidity in this series of elderly patients is in agreement with a very small number of publications devoted to this topic.