Introduction. Invasive pulmonary aspergillosis (IPA) remains a major clinical issue in Acute Leukemia (AL) patients (pts) and can affect the outcome of hematological treatment. Acute Myeloid Leukemia (AML) is considered the most frequent hematological condition predisposing to IPA; however, the introduction of posaconazole prophylaxis has significantly reduced IPA during AML induction. On the other hand, incidence of IPA among Acute Lymphoblastic Leukemia (ALL) pts seems to be increasing. In order to confirm these epidemiological data and to analyze the impact of IPA on the outcome of AL, we analyzed all AML and ALL pts consecutively admitted to our Institution during a 5-year period of observation.
Patients and Methods. From Jul-11 to Jun-16, all consecutive AML and ALL pts admitted to our Institution and treated with intensive chemotherapy were analyzed. IPA was diagnosed according to EORTC/MSG criteria and categorized as possible (poss) and probable/proven (p/p). Anti-mold prophylaxis (posaconazole) was always administered during induction/reinduction only in AML pts. CT scan was performed in case of persisting neutropenic fever for 3 days despite broad-spectrum antibiotic therapy. The effects of age, gender, type of AL, disease risk (ELN 2010 for AML, ESMO 2016 for ALL) and status [complete remission (CR), relapse/refractory (Rel/Refr)], IPA, allogeneic stem cell transplantation (alloSCT) on outcome were evaluated.
Results. During the 5-year period, 175 AL patients were observed (136 AML/39 ALL; median age 55, range 14-74; M/F ratio 6/5). Overall, 28 (16%) IPA were diagnosed, 11 (6.3%) poss and 17 (9.7%) p/p. Frequency was similar in AML (22/136, 16.2%) and in ALL pts (6/39, 15.4%,) but p/p IPA were relatively more frequent in ALL (83%) (1 poss, and 5 p/p) than in AML (54.5%) (10 poss and 12 p/p). Considering the disease phase, IPA occurred in 8% of 175 pts during induction, in 4.4% of 159 pts during consolidation and in 13.5% of 52 pts during reinduction. The frequency of IPA, particularly p/p, observed during induction was lower in AML (10/22, 45%; 5 poss and 5 p/p) than in ALL (4/6, 67%, all p/p). More than half of the IPA observed in AML were recorded in phases different from induction (6 during consolidation and 6 during reinduction). Median time from AL diagnosis to IPA was 2.5 mo (range 0-26) in AML and 1 mo (0-6) in ALL (p=0.089). After a median follow-up of 36 mo (range 4-68), 69 pts died: 12 (17.4%) of refractory AL and concomitant IPA and 2 (2.9%) of IPA while on CR. Thirteen pts with previous IPA underwent alloSCT; 9 (69.2%, 6/8 poss and 3/5 p/p IPA) died, as compared to 24 of 65 (36.9%) pts transplanted without a history of IPA (p=0.03). AL 2-year overall survival (OS) was significantly affected by the development of IPA, with no differences between poss and p/p IPA (no IPA 71.9% vs poss IPA 36.4% or p/p IPA 33.1%; p=0.0003 and <0.0001, respectively) (Fig. 1a). Figures are similar when considering AML and ALL separately (no IPA 66.7% vs poss IPA 40% or p/p IPA 33%, p=0.016 and <0.0042; no IPA 87.5% vs IPA 25%, p=0.0005, respectively for AML and ALL). Notably, after excluding the role of alloSCT by censoring pts at transplant (median follow-up 14.5 mo; range 2-68), poss IPA had no effect, whereas p/p IPA was associated with lower survival (at 2 years: p/p IPA 33.6% vs no IPA 80.6% or poss IPA 85.7%; p=0.0003 and 0.0004, respectively) (Fig. 1b). Multivariate analysis confirmed that developing IPA was an independent risk factor for OS (OR 3.4, CI 1.2-9.4), together with age≥60y (2.19, 1.01-4.74) and having Rel/Refr AL (9.41, 4.34-20.4), whereas achieving CR after 1 induction was protective (0.15, 0.05-0.45). Only p/p IPA was a risk factor for mortality, when excluding the effect of alloSCT (5.06, 1.63-15.65).
Conclusions. These data confirm that IPA still affects a considerable proportion of AL pts and that its frequency has become high even in ALL patients, where it occurs early during induction and mainly as p/p IPA, probably because of the more intensive chemotherapeutic regimens containing dexametasone and of the lack of anti-mold prophylaxis. Therefore, anti-mold prophylaxis should be recommended also in ALL patients during induction. In AL pts not undergoing alloSCT, only those with p/p IPA have a poor prognosis; however having a poss IPA negatively impacts on OS when considering the entire series, including alloSCT pts.
Full conference title:
- ASH 59th (2017)