The Premature Reduction of Venetoclax in an Initial Cycle of Ven-AZA Regimen for Unfit AML Patients Leads to Shortened Overall Survival - a Retrospective Study of Polish Adult Leukemia Group (PALG)

Introduction

The combination of the BCL2 inhibitor venetoclax (VEN) and azacitidine (AZA) has changed the paradigm of treatment for elderly or unfit acute myeloid leukaemia (AML) patients (pts). Indeed, in this hard-to-treat population, treatment with hypomethylating agents (HMAs) has been associated with a complete response (CR) rate of 20–30% and a median overall survival (mOS) of approximately 10 months. By contrast, the VEN-AZA regimen has demonstrated markedly improved outcomes, with response rates around 65% and overall survival of 14.7 months, as evidenced by the pivotal findings of the VIALE-A study.

Translating the results of clinical trials into routine clinical success additionally necessitates the effective management of adverse events (AE) like tumour lysis, and dose modification due to AEs to maintain good treatment outcomes.

Patients

This retrospective, observational study included 254 patients with AML, as classified by the World Health Organization 2016 criteria, who were ineligible for intensive chemotherapy (Tx) and received VEN-AZA at 14 Polish Hematology Departments between September 2021 and January 2024.

The median age of 72 pts was years (range 67-87). The male demographic constituted 53% of the sample. Among the patients, 25% were AML with myelodysplasia-related changes.

Cytogenetic risk categorization according to European LeukemiaNet revealed that 47% of the patients exhibited poor risk, 44% had intermediate risk, and only 9% had favourable risk. Baseline mutations in NPM1, FLT3, TP53, and IDH1/2 were detected in 11.8%, 17.1%, 11.8%, and 13.8% of the patients, respectively. Comorbidities were present in 88% of the patients, with cardiac disease being the most common.

The Kaplan-Meier analysis and log-rank test were used for overall survival (OS) in months (m). To evaluate significant differences between categorical data Fisher's exact test was used.

Result

The composite complete remission (cCR = CR + CRi) was achieved in 154 patients (61.8%) and the mOS was 18 m [95% CI: 12.0-21.4].

Among the molecular subgroups, the highest cCR rate (85%) and mOS (not reached) were observed in patients with IDH1/2^mut which was significantly higher than in unmutated cases (p=0.02, p=0.05, respectively). In contrast, the lowest cCR rate and mOS were found in patients with TP53^mut (cCR 44%, mOS 5.6 m) and were significantly lower than in patients without those (p=0.04, p=0.03, respectively). There was no significant difference in cCR rates and mOS for patients with NPM1 and FLT3 mutations.

Due to variations in the timing of remission assessment during the first cycle which resulted in differences in the duration of VEN administration (14 vs 21 vs 28 days), we analyzed the potential impact of the duration of VEN administration in the first cycle on cCR and mOS.

The study demonstrated a significantly higher cCR and mOS rate in patients who received VEN for 21 (n=69 patients) or 28 days (n=153 patients) during the first cycle compared to those who received it for only 14 days (n=42 patients) for cCR: (69% and 70% vs. 42%, p=0.007) and mOS (21 m vs. 16 m vs. 4.6 m p=0.005, respectively). The analyses did not show significant differences in age, gender, ECOG status, comorbidities, type of mutations and cytogenetic risk between the groups.

During the first three days of the initial VEN-AZA therapy, 31 (12%) pts developed biochemical tumour lysis syndrome (TLS), but only nine of pts (3.9%) progressed to clinical TLS. All 9 patients with clinical TLS experienced renal insufficiency/ acute kidney injury. There were no reported cases of seizures, cardiac arrhythmias, or mortality due to TLS. Two patients required renal replacement therapy, and 7 patients needed escalation of care to the ICU for TLS management. The analyses indicated that a low ECOG status≥2;(p=0.006), IDH1/2^mut (p=0.03), the need for initial cytoreduction therapy (p=0.005), and a creatinine level above 1 mg/dL (p=0.0006) significantly increased the likelihood of developing TLS.

Conclusions

We confirmed that VEN-AZA significantly improves the cCR rate and OS in real-life populations of AML pts who are not eligible for intensive therapy. Nevertheless, the premature reduction of VEN in an initial cycle of VEN-AZA may lead to a lower cCR rate and shortened OS. TP53 mutation results in poorer CR and OS rates in patients treated with VEN-AZA.