Candida dubliniensis Infection, Singapore

To the Editor: We read with interest the letter by Marriott et al. (1) describing the first case of Candida dubliniensis fungemia in Australia. We report the first two cases of C. dubliniensis infection in Singapore. We have been using the API 20C AUX on all yeast isolates from blood, sterile fluids, and tissue to screen for C. dubliniensis since May 2000. To our knowledge, this infection has not been previously reported from a Southeast Asian country. The first patient was a 49-year-old woman with adult polycystic kidney and liver disease; she had mild chronic renal failure and a past history of nephrotic syndrome. She was admitted to the renal unit on May 9, 2001, for management of ascites. A septic workup showed leukocytosis (34.67 x 10 9 /L) with predominance of neutro-phils (95%). Blood culture received on May 31 grew yeast with two slightly different forms, one large and one small. The isolates, investigated separately , were germ-tube positive. The API 20C AUX profile for the larger isolate was 6072114 at 48 hours' incubation (96.2% certainty for C. dublin-iensis) and 6072134 for the smaller isolate (99.8% certainty for C. dublin-iensis). The yeasts grew well on potato dextrose agar at 35°C but poorly at 42°C and 45°C. In addition, electro-phoretic karyotyping with pulsed-field gel electrophoresis showed that both isolates had identical patterns, with eight chromosome fragments, one of which was <1 megabase (Mb), indicating that the two morphologically different strains were the same karyo-typically. The pattern obtained was in keeping with results obtained by Jabra-Rizk et al. (2), with C. dublin-iensis showing a chromosome-sized band of <1 Mb. For C. albicans, by contrast, all bands were >1 Mb. Five control strains—C. dubliniensis (RCPA Microbiology QAP item 2001:2:7A), C. albicans ATCC 90028, and three clinical strains of C. albi-cans—were also run together with the two strains; the results obtained were consistent with those of Jabra-Rizk et al. (2). The MIC of the isolates to flu-conazole by E-test was 0.75 µg/mL, indicating susceptibility. Disseminated intravascular coagulopathy due to sep-sis from a possible ruptured liver cyst developed in the patient. Despite broad-spectrum antibiotics and amphotericin B, hemodialysis, and intensive-care support, she died 5 weeks after admission. Except for C. dubliniensis candidemia, Candida species isolated from the urine and endotracheal secretions (speciation not done), and Acinetobacter bauman-nii (cultured from the endotracheal secretions and femoral catheter tip), no other important microorganisms were isolated. Peritoneal fluid …