Abstract
Introduction: Incidence of fungal infections is reducing in the last years due to wider use of effective antifungal prophylaxis. On this basis, we evaluated clinical characteristics and outcome of patients (pts) with hematological malignancies (HMs) and fungal bloodstream infections (FBSI).
Patient and Methods: This retrospective/prospective study gathered consecutive documented FBSI observed among HMs diagnosed between January 2011 and June 2015 in 19 Italian Hematology Departments that refer to SEIFEM (Sorveglianza Epidemiologica Infezioni Fungine Emopatie Maligne) group.
Results: We collected overall 100 patients including retrospective pts and those observed in the first six months of the prospective study with FBSI among 16 centers; further 3 centers reported no cases of FBSI. Regarding patients' characteristics male/female ratio was 1, median age was 55 yeras. (IQR 18-88). Two-third of FBSI were detected in AML (43 - 43%) and NHL (27 - 27%); the remaining pts were affected by ALL (9 - 9%), MM (6 - 6%), MDS (5 - 5%), MDS/MPN (5 - 5%), CLL (3 - 3%) and HL (2 - 2%). Thirty-five pts had FBSI at the onset of HM or after the first induction, 47 after treatment for refractory/relapsed disease, 13 when in remission, 28 pts during transplant procedures (17 from allogeneic donor, 11 from autologous cells). Fifty-nine pts were receiving antifungal prophylaxis at FBSI breakthrough: 26 posaconazole, 16 fluconazole, 6 itraconazole, 3 amphotericin B, 3 caspofungin, 1 voriconazole, and 4 combination of amphotericin B/azoles; 4 pts were treated with secondary prophylaxis after previous fungal infection. Eighty-nine pts had a central venous catheter. Eighty-four pts presented a neutrophils count < 500/mmc for a median time of 7 days before FBI onset (0-70 d); 50 pts received steroids and other 17 immunosuppressive treatment for allo-HSCT.
Yeasts were the most common agent detected. Candida spp. represent the most represented yeast, counting for 77% of all infections; (21 albicans and 56 non albicans); 8% of FBI were due to 8% Geotrichum, 3% Trichosporon and 2% Rhodotorula, Molds were rare but not infrequent: 8% were caused by Fusarium and 1% by Scedosporium.
Three patients died before starting any antifungal. Fifty-two received echinocandins (49 caspofungin and 3 anidulafungin), 22 liposomal amphotericin B (L-AmB), 15 azoles (7 fluconazole, 3 posaconazole, 3 voriconazole and 2 itraconazole) and 8 pts combo therapy (5 posa+L-AmB, 2 Caspo+L-AmB, 1 vori+caspo).
Thirty-eight pts died within 30 days from FBSI diagnosis, 28 (74%) of whom due to infection. Among these 12 (43%) suffered from AML (1 in induction, 1 in CR, 10 had refractory/relapsed disease), 2 (7.2%) from ALL (all with refractory/relapsed disease), 7 (25%) from NHL (2 in induction, 2 in CR, 3 with refractory/relapsed disease), 4 (14%) from refractory/relapsed MM, 1 (3.6%) from MDS in remission, 2 (7.2%) from newly diagnosed MDS/MPN. In 8 pts (21%) for whom we cannot discriminate if death was subsequent to FBSI only or to uncontrolled HMs, fungal isolates were all rare yeast or molds except 2. Mortality was related only with advanced phase of underlying HM (p<0.0001), molds fungemia (p<0.06) and rare fungi isolation (p<0.06). There is no difference in overall mortality rate among pts treated with echinocandins or azole or amphotericin B compounds or combo.
Conclusions: Nowadays FBSI represents a rare complication of HMs, as a consequence of wider availability of effective antifungal prophylaxis. Candidemia still represents the most important cause of FBSI, although about 25% of FBSI are due to rare yeast or molds. Regardless a lowering incidence, the observed mortality remains high even with target treatment.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
