Impact of Early Hospitalization, Medical Comorbidities and Allogeneic Stem Cell Transplantation on Survival in Acute Myeloid Leukemia Patients Treated with Venetoclax Based Therapy - a Single Institution Experience from an Academic Center

Introduction: Bcl-2 inhibitor Venetoclax (Ven) in combination with hypomethylating agents (HMA) has improved response rates (RR), and median overall survival (OS) in Acute Myeloid Leukemia (AML) patients who are elderly and/or ineligible for intensive therapies. However, observed outcomes in patients treated in off-trial setting may be inferior (Winters, et al 2019). While disease markers that predict poor RR are well described, patient factors that could impact the inferior outcomes in the real-world setting have not been fully understood.

We investigated the impact of incremental age, early hospitalization, comorbidity burden, along with other disease and treatment variables, on survival.

Patients and methods: After IRB approval, all consecutive patients with AML and high-risk myelodysplastic syndromes (MDS) who were treated with Ven based regimen between 2017 and 2021 were reviewed. Descriptive statistics are reported. Kaplan-Meir analysis was used to estimate survival, and cox-proportional hazards was used to investigate survival association with variables. Overall survival (OS) was calculated from the time of initiation of venetoclax to the date of last follow up or death. Disease related data, effect of Incremental age, number of medical co-morbid conditions, unplanned early hospitalization (EH) rates within 30 days (EH30) and 90 days (EH90), causes of hospitalization, transfusion needs, and other treatment related adverse events were captured. Factors that impacted EH and overall survival were investigated.

Results: Baseline characteristics are outlined in Table 1. A total of 89 subjects with AML and MDS-EB2 who received Ven based therapy were identified. All but 3 received HMA/Ven combination. Forty-nine (55%) received HMA/Ven in first line (FL) and 40(45%) in second line (SL) or later, of whom 12(30%) were treated in post- transplant (ASCT) relapse. Median age was 72 and 64, and median follow up was 21 months (95% CI, 9,27) and 22 months (95% CI, 9,32), respectively, for FL and SL. By ELN criteria high risk patients were 77% and 57% in FL and SL cohorts, respectively.

Sixty percent and 40% patients had EH30 and 52%, and 35% patients had EH90 in FL and SL cohorts, respectively. Presence of ≥2 comorbidities significantly predicted EH90 (p=0.04). Most common cause of EH was neutropenic fever and sepsis. In 20 patients where data was available, median pRBC units received were 6(range 1-32), and 10(range 0-50), and median platelet units were 8(range 0-42), and 1 (range 0-73), in FL and SL cohorts, respectively.

Table 2 summarizes survival outcomes by important variables. Median OS is 6 months (95% CI, 4,10), and 8 months (95% CI, 4,11) for the entire FL and SL groups. Within the SL group, median OS was significantly worse at 4 months (95% CI 0,8) Vs 11 months (95% CI, 4,13) for those who previously received an ASCT vs those who did not (p=0.02).

In FL patients, increasing age (10 vs 5 months, p=0.03) and poor therapy response (11 vs 2 months, p<0.0001) had significant impact on survival, while in SL group TP53mut (13 vs 4 months, p=0.04) and poor response (8 vs 2.5 months p=0.04) were significant predictors of poor OS. On multivariate analysis (MVA), only TP53mut (p=0.04), and response (p=0.003) retained significant impact on OS in both FL and SL cohorts. While EH90 and presence of ≥2 comorbidities also trended towards lower survival, they did not reach statistical significance. Incremental age had no impact on OS in MVA.

Following remission, ASCT Vs no-ASCT resulted in an improved survival, with a median OS of 15 months (95%CI 0,19) vs 5 months (95% CI 3,9), p=0.09; and NR (95% CI 3, NR months) Vs 6 months (95% CI 4, 8), p=0.04, In both FL and SL cohorts respectively.

Conclusions: We comprehensively evaluated disease and patient factors, and their impact on hospitalization, and on OS in AML patients during Ven based therapy. EH and transfusions were common, without an impact on OS. Age had no impact on EH or OS, but comorbidity burden increased EH risk. High risk disease and poor response to therapy resulted in poor OS. ASCT was crucial for prolonged OS in both FL and SL cohorts. One likely cause of inferior OS noted in FL patients could be the lower rate of ASCT. Our results could provide better understanding to optimize Ven based therapy, and improve early treatment related toxicities in the real-world setting. Quality of life and health care utilization burden associated with this regimen needs to be studied systematically.

No relevant conflicts of interest to declare.