Candida parapsilopsis Fungemias in Bone Marrow Transplant Recipients: Implications for Azole Prophylactic Therapy

To the Editor:

Girmenia et al have recently published their experience on breakthrough fungemias caused by Candida sp in patients under fluconazole prophylaxis. Candida parapsilopsis was the most frequent yeast isolated in this setting, but its isolation in blood cultures was not correlated with deep-seated infection in any patient. The investigators conclude the usefulness of fluconazole prophylaxis in the prevention of deep-seated infections caused by fluconazole-susceptible Candida strains, even when breakthrough fungemia develops.1 

The incidence of C parapsilopsis candidemia in our Bone Marrow Transplantation (BMT) unit has also been high: 5.48% of the BMT recipients; in addition, in our experience, the introduction of azole prophylaxis did not decrease its incidence. However, in contrast with the results of Girmenia et al, our data suggest that neutropenic patients with breakthrough candidemias by C parapsilopsis while on therapy with azoles are also at risk to develop invasive candidiasis.

Since 1990, 9 of 203 BMT patients developed C parapsilopsis fungemia: 2 males and 7 females; median age 29, range 15 to 49; 5 autologous and 4 alogenic transplants; 5 from marrow progenitors and 4 from peripheral blood progenitors; 3 chronic myeloid leukemias, 3 lymphomas, 2 acute myeloblastic leukemias, and 1 breast carcinoma. All of them were neutropenic (except 2), had a central venous line, and were receiving parenteral hyperalimentation and empirical broad spectrum antibiotics. The yeast was isolated at a median of 19 days (11 to 45) after transplant.

In our BMT unit, antifungal prophylaxis consisted on nystatin until 1993. Thereafter, a prospective study was started to compare the role of fluconazole and itraconazole as fungal prophylaxis after BMT. Antifungal prophylaxis was maintained until patients recovered from neutropenia. The C parapsilopsis infection rate was similar in fluconazole and itraconazole recipients, and slightly lower, although not statistically significant, than the observed in nystatin patients (Table 1). However, the introduction of prophylaxis with azoles was almost coincidental in time with the beginning of a peripheral blood stem cells transplant (PBSCT) program, and so, the shortening of the neutropenic period after PBSCT may have contributed to explain the slight decrease in the C parapsilopsis infection rate.

A deep-seated infection was observed in all patients, except in the two who were not neutropenic when the yeast was isolated from blood cultures. All patients who suffer candidemia while on prophylaxis with nystatin (5/5) developed invasive infections. The two neutropenic patients who were on antifungal prophylaxis with either fluconazole or intraconazole, also developed deep-seated infection after breakthrough candidemia by C parapsilopsis. An epidemiologic study was performed, but we failed to show any source of infection, contrarily to previous reports.2 3 

Minimal inhibitory concentration (MIC) values from the C parapsilopsis strains isolated by us were found to be higher for fluconazole (median MIC90, 1.56 μg/mL; range, 0.39 to 3.12) than for itraconazole or amphotericin B (median MIC90: 0.09 μg/mL and 0.39 μg/mL, respectively; ranges: <0.09 to 0.09 and 0.09 to 0.78), these fluconazole concentrations being almost double than those observed for C albicans strains.4 The yeast was isolated in the central venous lines of the five patients in whom the catheter could be removed. No patient in whom catheter was not withdrawn survived to the mycosis, probably because of the difficulty of eliminating the yeast because of its high adherence to plastics.5 On the other hand, when plasma itraconazole concentrations were determined in our patients, therapeutic levels were achieved only in half of BMT recipients, as other authors have previously reported.6 

In summary, in our experience antifungal prophylaxis with fluconazole or itraconazole did not seem to diminish the incidence of either candidemia or deep-seated infections caused by C parapsilopsis when compared with nystatin, at least in high-risk neutropenic patients. This lack of efficacy may be explained by the relatively high MIC values of the yeast with fluconazole, its high adherence to plastic surfaces, and the important absorption problems of itraconazole after BMT. So, the possibility of C parapsilopsis–associated deep-seated infections should be kept in mind, even when fluconazole or other azoles are used as prophylaxis in the BMT setting.