Treatment of mycobacterial exit-site infections in patients on continuous ambulatory peritoneal dialysis.

Exit-site infections (ESIs) are frequently due to gram-positive organisms and occasionally to gram-negative organisms. Initial empiric antibiotic therapy is therefore directed against these organisms until culture reports are available. Two cases of ESI associated with Mycobacterium are here reported. The first patient, a 63-year-old man with type 2 diabetes, recently treated for Staphylococcus epidermidis peritonitis, presented with acute purulent drainage at the catheter exit site, accompanied by pain and erythema. No tunnel abscess was identified by ultrasound. Empiric antibiotic therapy was initiated with ofloxacin and vancomycin. A rapid-growing acid-fast bacillus (AFB) noted four days after culture was eventually identified as Mycobacterium fortuitum. Ofloxacin was continued, vancomycin was discontinued, and clarithromycin was added. The ESI initially showed improvement; therapy was therefore continued for several months. However, cultures remained positive for M. fortuitum, and the catheter was removed 5 months after therapy was initiated. The second patient, a 28-year-old woman, presented with severe pain and tenderness at the exit site without erythema or drainage. Empiric therapy with cefazolin, gentamicin, and cephalexin was initiated. Gram-positive cocci and an AFB were identified from the exit-site culture, and antibiotics were initially changed to clarithromycin, trimethoprim/sulfamethoxazole, and ofloxacin. The organisms were subsequently identified as M. chelonae-M. abscessus complex and coagulase-negative Staphylococcus. The patient continued to improve after 3 weeks of antibiotic therapy. However, despite the initial improvement in the ESI, the M. chelonae-M. abscessus complex continued to grow, and amikacin was added intravenously. Despite continued treatment, the ESI did not resolve, and the catheter was removed after 4 months of therapy. Despite unusual exist-site infections with rapidly growing AFBs, both patients continued continuous ambulatory peritoneal dialysis (CAPD) while undergoing treatment for ESI. Catheters were left intact, as improvement was initially seen with no evidence of tunnel infection or peritonitis. Rapid-growing AFB should be considered another possible causative agent for ESI. Two cases of atypical mycobacterial exit-site infection are presented to illustrate the difficulties in managing this complication of peritoneal dialysis. Ofloxacin--or other quinolones--may provide a better spectrum of coverage when choosing empiric therapy in patients presenting with ESI.