Optimal treatment for a chronic infected prosthesis is the removal of infected and necrotic tissue and all the components of the prosthesis with staged revision in conjunction with systemic antibiotics. If this is not possible because of the poor general condition of the patient, because of unacceptable functional results secondary to removal of the prosthesis, or because the patient refuses surgery in an attempt to salvage the infected prosthesis, a reasonable alternative is long-term oral suppressive antibiotic therapy for maintenance of a functioning prosthesis. Prompt recognition with rapid debridement and initiation of antibiotic therapy seems crucial. Our study confirms a favorable outcome of maintenance of functioning prostheses in 86.2% of patients after a mean followup of 5 years. All patients had initial debridement with 4 to 6 weeks of systemic antibiotic therapy. Advanced age did not seem to predict poor outcome. Joint location, duration of symptoms, and the time of onset of infection did not predict success or failure. The overall success rate for Staphylococcus aureus prosthetic joint infection was 69% after a mean followup of 5 years. The ideal regimen and optimal duration of oral suppressive therapy for a favorable outcome is not well-established and needs additional data with prospective multicenter studies.

SUMMARY The CB/Ss LAK strain of inbred hamster was used as a model for studies of infection with Treponema pertenue and of acquired resistance to it. When infected, this strain developed cutaneous lesions which lasted for six to seven months, even in the presence of peak titres of antitreponemal antibody. The rate of appearance and resolution of these lesions varied with the size of the inoculum. The infected hamsters' inguinal lymph nodes increased significantly in weight and teemed with treponemes for several weeks. Animals infected for eight or 10 weeks obtained quick resolution of their lesions by treatment with penicillin and were thereafter resistant to reinfection.

A centrifugation-filtration procedure was developed to expedite the recovery of microorganisms from blood. Fresh whole human blood was inoculated with various aerobic and facultatively anaerobic microorganisms (3 to 18 per ml). The seeded blood was carefully overlaid on a Ficoll-Hypaque gradient (density, 1.114 g/ml) and centrifuged (400 x g) for 45 min at ambient temperature. The entire gradient (plasma, leukocytes, and Ficoll-Hypaque) was removed and filtered through a 0.22-,um membrane filter. The filters were then placed on chocolate agar and incubated at 35°C in humidified air containing 5% CO2. No statistically significant differences were detected between the numbers of microorganisms recovered by filtration and by direct culture of the original inoculum. Most microorganisms were detected within 18 h after filtration. This system has excellent sensitivity and negligible toxicity.

Transient bacteremia associated with percutaneous liver biopsy was studied by pour-plate blood cultures, which were obtained immediately before and after the procedure and 5, 10, 15, and 30 min later in 89 patients. Part of the liver tissue was also cultured in all patients. Histological diagnoses included hepatitis, cirrhosis, cholangitis, fatty liver, granulomata, metastatic liver disease, lymphoma, and miscellaneous disorders. All blood cultures obtained before liver biopsy were sterile. Bacteremia was demonstrable in 12 patients (13.48%). In most of these patients, blood cultures were positive for as long as 15 min after liver biopsy; all cultures were negative at 30 min. Among the bacteria associated with 12 episodes of bacteremia were Escherichia coli, Klebsiella, Bacteroides, enterococci, diphtheroids, Staphylococcus aureus, alpha-hemolytic Streptococcus, and Streptococcus pneumoniae. The patients with positive liver biopsies had a higher incidence of bacteremia (83.3%) than did the patients whose liver biopsies were sterile (8.r%); this difference is stastically significant (P smaller than 0.01). Thus, liver biopsy can be associated with transient bactermia.

The immune mechanism by which hamsters acquire resistance to infection with Treponema pertenue, the causative agent of frambesia, or yaws, has not been elucidated. Serum or cells (spleen or lymph node) obtained from hamsters resistant to frambesial infection were transferred to normal syngenic recipients, who are subsequently infected with T. pertenue. The following parameters were used to measure the ability of immune serum or cells to confer resistance on recipient hamsters to frambesial infection: inhibition of the development of cutaneous lesions, decreased weight, and number of treponemes in the inguinal lymph nodes. This investigation demonstrated that immune serum conferred protection on recipient hamsters infected with T. pertenue. Discontinuation of the administration of immune serum (18 days after frambesial infection) did not result in the development of cutaneous lesions. Since the inguinal lymph nodes contained a sizeable number of treponemes (2.6 x 105), immune serum failed to prevent frambesial infection. Recipients of immune spleen or lymph node cells initially developed frambesial lesions 9 days after infection. The frambesial lesions began to resolve 12 to 14 days after infection and by day 21 had completely regressed. These results illustrated that humoral factors and cells are involved in resistance of the hamster to frambesial infection.

Serum or spleen cells from hamsters immune to infection with Treponema pallidum strain Bosnia A conferred partial protection against syphilitic infection by the same strain on recipient hamsters. Cutaneous lesions did not develop, and in the lymph nodes the average weight and number of treponemes detected were significantly lower than in normal and control hamsters. Treatment of the immune spleen cells with antithymocyte serum and complement abolished their ability to transfer resistance. This is the first direct evidence that thymus-derived cells are involved in resistance to syphilitic infection.

Ten cases of protracted diarrheal illness after the oral administration of lincomycin or clindamycin in standard dosages were observed in previously healthy subjects. An abrupt onset of diarrhea, crampy abdominal pain, fever, and leukocytosis was observed one to 12 days after discontinuation of the drug. Proctoscopic examination revealed erythematous friable mucosa covered with small raised, yellowish-white plaques that were sometimes confluent. Barium contrast studies of the colon demonstrated irregular shaggy mucosa, ulcerations, cobblestone appearance, and thumb printing. Rectal bipsy showed acute inflammation with pseudomembranes with focal or superficial ulcerations. All patients had a protracted course but recovered with supportive management. Follow-up barium enemas and proctoscopy were done on all patients and were normal. A history of diarrhea, fever, and mucosal changes seen on proctoscopy in a patient who has recently received one of these antibiotics should raise the possibility of colitis associated with clindamycin and lincomycin therapy.

The percentage of Ia antigen-bearing (Ia') macrophages was significantly lower in mice infected with Trypanosoma rhodesiense than in normal controls. The degree of difference varied with the source of macrophages and time course of infection. The percentage of Ia+ macrophages isolated from spleens 10 days after infection was 71% of that in the controls, and depletion continued until Ia+ macrophages were almost undetectable 30 days after infection. The rate of depletion was slower in the peritoneal cavity. In contrast, Ia+ macrophages were not significantly depleted from the lymph nodes until 30 days after infection. The ability of macrophages from trypanosome-infected mice to present listerial antigen to sensitized T cells was significantly lower than in controls. Immune T cells had significantly less ability (43% of controls) to incorporate thymidine when exposed to splenic macrophages from infected mice during the early stage of disease. This loss of antigen presentation increased during the course of infection. Peritoneal macrophages also exhibited an early loss of ability to present antigen, but no significant decline occurred thereafter. No significant loss of antigen had occurred in the lymph node macrophages 10 days after infection, but during the later stages of the disease a significant loss was detected. Treatment of macrophages from infected and control mice with anti-Iab serum and complement inhibited their ability to present antigen. Our results demonstrate that Ia+ macrophages and their distribution can influence the ability of infected animals to process-antigens and may therefore account in part for the immunosuppression observed in trypanosomiasis.