[Clinical results of resection arthroplasty for infected shoulder arthroplasty]

Infection is a rare complication of shoulder arthroplasty. Various therapeutic solutions have been proposed: antibiotics alone, one-stage or two-stage reimplantation, surgical or arthroscopic cleaning without prosthesis removal, scapulohumeral arthrodesis or simple arthroscopic resection. We evaluated the mid-term clinical outcome after resection arthroplasty for the treatment of infected shoulder arthroplasty.

The series included ten infected arthroplasties in ten patients. Mean duration of implantation was two years seven months (range nine months to five years). Bacteriological diagnosis was established from intraoperative articular samples or systematic samples taken during surgical revision procedures: meti-S Staphylococcus aureus strains (n=4), coagulase-negative Staphylococcus (n=5 including three S. epidermidis) Streptococcus mitis (n=1) and Citrobacter koseri (n=1). The mean Constant score before revision was 58 (range 23-77). Subjective patient satisfaction before surgical revision was rated good in six cases, fair in one and poor in three. Surgery associated removal of the implant, complete resection of the cement, resection of the fistular tracts, wide debridement of infected tissues and total synovectomy.

Patients were seen at an average follow-up of three years eight months. The objective functional outcome measured with the Constant score was only fair, 28 points (range 20.6-36), and corresponded to a loss of 29 points compared with the preoperative score. This was explained mainly by lower scores for joint motion, function and muscle force but with persistently satisfactory scores for pain. All patients remained pain-free (daytime and nighttime). Patient satisfaction was rated good for two, fair for five and mediocre for three. Clinical and biological proof of eradicated infection was obtained in all patients.

Infection remains a serious devastating problem for shoulder arthroplasty with an important functional impact. Resection only has a modest clinical effect. Precise identification of the causal germ with institution of adapted antibiotic therapy is required for eradication of the infection. Early diagnosis is probably the most important parameter affecting clinical outcome and surgical options. Functional results after resection arthroplasty are modest. This procedure should be reserved for patients with reduced functional demands. Improved management of the infectious load and reduction of diagnostic delay should help improve functional outcome and favor use of stow-stage procedures for reinsertion.