Optic Neuritis: A Rare Complication of Primary Varicella Infection

Brief Reports 515 His vancomycin-resistant enterococcal bacteremia had been preceded by treatment with iv vancomycin, imipenem, and ciproflox-acin for prior infections. He again had severe neutropenia and mucositis. He never developed signs or symptoms of sepsis. He received no specific therapy, and subsequent blood cultures were negative for enterococci. His neutropenia resolved after 37 days, and his leukemia went into remission. In September 1995 the patient's leukemia relapsed, and he again received chemotherapy. A rectal surveillance culture once again yielded the same E. faecium isolate. After two infections (Candida albicans fungemia 10 days before isolation of vancomycin-resistant enterococci [VRE] in blood cultures and Staphylococcus epidermidis bacteremia 20 days before isolation of VRE in blood cultures), blood cultures again yielded the identical E. faecium isolate. Even though his vascular catheter insertion sites were changed, E. faecium bacteremia persisted and he had severe neutropenia and mucositis. On the fifth day of bacteremia, he developed septic shock with hypotension and respiratory and renal failure. He remained bacteremic throughout the course of therapy with doxycycline and novobiocin (the organism was susceptible to these two antibiotics). He was still severely neutro-penic at the time of his death 3 days later. Although vancomycin-resistant enterococcus has become a major nosocomial pathogen, little is known about its epidemiology. Leuke-mia patients offer a unique opportunity to study the epidemiology of VRE since they are admitted to the hospital numerous times and (at our institution) since rectal surveillance cultures are performed weekly while the patients are hospitalized. In our patient's case, he was colonized with the same vancomycin-resistant enterococcus clone over a 2-year period despite extended periods (>3 months) outside of the hospital and remission of his leukemia. Of 22 follow-up rectal surveillance cultures, all but one yielded VRE. At our institution, we maximize the chance of isolating VRE by using selective media (colistin nalidixic acid agar supplemented with defibrinated 5% sheep blood, vancomycin [10 /..tg/mL], and amphoter-icin B [1 pg/mL]) for surveillance cultures, which may explain the high proportion of positive cultures. However, the potential for long-term carriage suggests that all cancer patients with a history of coloni-zation with VRE should be isolated during every hospitalization. Our patient had multiple risk factors for infection due to VRE. He consistently became bacteremic with VRE after induction chemother-apy that caused neutropenia and mucositis, which in turn necessitated the use of therapeutic antimicrobial agents. He had received multiple antibiotics (including oral vancomycin) …