Toxicity and Efficacy of Isavuconazole Vs Voriconazole As Anti-Fungal Prophylaxis for Patients with Acute Myeloid Leukemia

Introduction: Triazole agents are preferred for antifungal prophylaxis (AFP) for patients (pts) with acute myeloid leukemia (AML) given their activity against the most common causes of invasive fungal infections (IFI), namely Candida spp. and Aspergillus spp. (Halpern et. al. Blood 2015). However, use of the azole antifungals voriconazole (VCZ) and posaconazole (PCZ) is sometimes limited by their side-effect profiles, including transaminase elevation, visual disturbance, QTc prolongation, and major drug-drug interactions (DDI) due to cytochrome P450 inhibition. Conversely, isavuconazole (ICZ) has comparable efficacy against yeasts and molds, with broader activity against Mucorales, and lesser unfavorable impact of the side effects imparted by VCZ and PCZ. In this study, we sought to compare complication rates and outcomes of ICZ vs VCZ amongst pts with newly diagnosed AML.

Methods: We conducted a single-center, retrospective study of pts with newly diagnosed AML started on AFP with VCZ or ICZ. Patients were followed until death or IFI. Censoring occurred at Day 0 of allogenic stem cell transplant (alloHCT) due to confounding by additional immunosuppressive effects. Chi-square analysis and Wilcoxon rank-sum tests were used for categorical and continuous variables, respectively. Time-to-event analysis from AML diagnosis was performed using Kaplan-Meier method Correlation analysis used point-biserial calculation. Significance was established with a p-value <0.05.

Results: In total, 295 pts were evaluated, with exclusion if AFP was received for antecedent MDS/CMML (n=11), or if there was inadequate information available (n=6). Amongst the 278 pts included, median age was 63 years (interquartile range [IQR] 52-70), with 54.3% being male and 80.6% being white. 63 pts received ICZ and 215 received VCZ as first line AFP. Although pts on ICZ were older (median 66 vs 62 years, p=0.005), there were no differences in patient baseline characteristics including sex, ethnicity, cardiac/pulmonary comorbidity, baseline ANC or CD4 count between groups. DDIs were more common for pts on VCZ than ICZ (18% vs 2%, p=0.002).

Significantly more complications overall were observed for pts on VCZ (n=124, 58%) than ICZ (n=22, 35%, p=0.001), namely visual disturbance (15% vs 0%, p=0.002) and transaminase elevations (15% vs 3%, p=0.017); however, rates of CTCAE ≥grade 3 transaminase elevation were similar (4% vs 3%, p=1.0). Pts starting on ICZ had significantly more issues with insurance approval and cost-prohibition necessitating switching AFP (30% vs 1%, p<0.001). These complications led to 10 ICZ pts spending the majority of time (>50%) on VCZ, 38 VCZ pts spending the majority of time on ICZ, and 12 VCZ pts spending the majority of time on other AFP. This led to a total of 101 and 165 pts predominantly on ICZ and VCZ, respectively.

To reliably evaluate events attributable to individual AFP agents, we then evaluated only pts who were only on a single agent and followed until IFI, death, alloHCT, or last known follow-up (n=129 overall; ICZ, n=41; VCZ, n=88). There were no differences in age, sex, ethnicity, ECOG PS, cardiac/pulmonary comorbidity, or baseline CD4 count between groups. VCZ pts did have a lower baseline ANC (1.2 vs 2.2 x10^3/uL), p=0.012), but there was no difference in days with ANC <0.2 or ANC <0.5 between groups. Pts on ICZ had a higher baseline QTc (452 vs 436 ms, p=0.009), although peak QTc was similar (471 vs 456 ms, p=0.123), as was mean change in QTc (9 vs 18, p=0.285). The rate of cardiac events/arrythmia on AFP were comparable. Although ICZ pts were more likely to receive less-intensive therapy (53.7% vs 18.9% , p=0.0002), the rate of proceeding to alloHCT was similar (36.6% vs 37.8%, p=1.0).

In the larger cohort, the cumulative duration of neutropenia was positively correlated with IFI incidence, with correlation coefficient of 0.26 for days with ANC <0.5 and stronger at 0.45 for days with ANC <0.2. When evaluating pts on a single AFP for >50% of the study period, the rate of IFI was similar between pts receiving ICZ vs VCZ, specifically 2.4% vs 3.3% with proven IFI, 7.3% vs 6.7% with probable+proven IFI, and 15% vs 18.9% with possible+probable+proven IFI (p=0.814)

Conclusions: Our study shows that ICZ is an AFP option that is associated with considerably lesser rates of transaminase elevations, visual disturbances, and DDIs, with similarly efficacious protection from IFI when compared to VCZ for pts with AML.