Summary. We have analysed the incidence and risk factors for the occurrence of invasive fungal infections (IFI) among 395 recipients of an allogeneic peripheral blood stem cell transplantation (PBSCT) from a human leucocyte antigen (HLA)-identical sibling. IFI (n 50) occurred in 46 patients, giving an overall probability of 14%. There were 12 cases of invasive candidiasis (3%), with only one death. Non-Candida IFI occurred in 37 patients (12% probability), mostly invasive aspergillosis (n 32). In multivariate analysis the only two signi®cant variables associated with a higher risk of developing a non-Candida IFI were the development of moderate-to-severe graftversus-host disease (GvHD, P < 0á0001; OR 4á6) and having received steroid prophylaxis for GvHD (P 0á04; OR 2á1). In multivariate analysis the variables associated with a lower overall survival after PBSCT were development of a non-Candida IFI (P < 0á0001; OR 5á6), nonearly disease phase (P 0á0001; OR 1á9), steroid prophylaxis (P 0á02; OR 1á4), moderate-to-severe GvHD (P 0á01; OR 1á6) and cytomegalovirus infection post transplant (P 0á001; OR 1á8). Our results show that non-Candida IFI (in particular aspergillosis) was an important cause of infectious morbidity and mortality after an HLA-identical sibling PBSCT, while invasive candidiasis was rare. Use of steroid prophylaxis and, in particular, the development of moderate-to-severe GvHD post transplant were risk factors for non-Candida IFI. Prophylactic strategies for these infections should thus take into account these risk factors.

Diagnosis of invasive pulmonary aspergillosis (IPA) is often difficult. Recently, the European Organization for Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG) has proposed new criteria for the classification of invasive fungal infections. We have studied the clinical applicability of this classification in 22 patients with hematological malignancies who had IPA at autopsy. While alive, according to the EORTC/MSG criteria, only 2 patients were classified as having proven IPA, 6 as probable, 13 as possible, and 1 was unclassifiable. Of the patients, 64% had no microbiological or major clinical criteria before death. Although the EORTC/MSG criteria are an important step forward in the standardization of definitions used for IPA in clinical research studies, most patients who die with extensive lung disease only reach a level of possible or probable IPA during life, further highlighting that these guidelines should not be used for clinical decision-making.

In this retrospective monocenter study, we analyzed the outcomes of 130 adult hematological patients who developed a proven (n 5 23), probable (n 5 71), and possible (n 5 36) invasive aspergillosis (IA) in a 13-year period. Forty-nine patients (38%) were recipients of an allogeneic hematopoietic stem cell transplantation (AlloHSCT). The main goal of the study was the identification of prognostic factors for 4-month aspergillosis free survival (AFS) and overall survival (OS). IA was identified as the main cause of death in 27/49 recipients of an AlloHSCT (55%) and 28/81 nontransplanted patients (35%). Diagnosis of IA at or before 2000 had a negative impact in both 4-month AFS and 4-month OS in the entire group. In multivariate analysis performed separately for nontransplanted and allo-HSCT patients, five variables (excluding the year of diagnosis) decreased 4-month AFS: (i) impairment of one organ function (OF), (ii) impairment of two or more OFs (two points), (iii) disseminated IA, (iv) neutropenia lasting more than 10 days (non-AlloHSCT group only) or monocytopenia (<0.1 3 10 9 /l) [AlloHSCT group only], and (v) high-dose steroids (nonAlloHSCT group only) or an alternative donor (AlloHSCT group only). According to the number of adverse risk factors, three prognostic subgroups were defined in non-transplanted and alloHSCT patients with good (97% and 78% AFS), intermediate (73% and 32% AFS) and poor prognosis (20% and 11% AFS) of IA [P < 0.01]. In addition, we validated the French and Seattle prognostic indexes for allo-HSCT recipients and the Strasbourg model for all hematological patients with IA. Am. J. Hematol. 84:571-578, 2009. V

This randomized trial suggests that ABLC at 1 mg/kg/d produces less nephrotoxicity than c-AmB, without differences in the incidence of IRAEs and with similar efficacy.