Inpatient Versus Outpatient Hypomethylating Agent Induction for AML As a Predictor for Survival

Abstract

Introduction The hypomethylating agents (HMA) decitabine (dec) and azacytidine (aza) are used in acute myeloid leukemia (AML) for induction therapy in patients with either newly diagnosed or relapsed disease and who are above the age of 60 or for those who would not tolerate intensive induction chemotherapy. As a result, HMA have been widely used at our institution, on both inpatient (IP) and outpatient (OP) services. The decision to administer IP-HMA vs OP-HMA is complex, but IP-HMA is recommended at our institution for neutropenic infections, debility, or risk of tumor lysis. The prognostic significance of this decision, if any, has not been widely explored. We hypothesized that because patients who receive their first cycle of HMA as an IP were at higher risk for complications, they would have a worse 100-day survival than their OP counterparts. Methods This was a retrospective observational study of patients treated at Thomas Jefferson University Hospital from January 2016-January 2020. Inclusion criteria were patients who had a diagnosis of AML, were above the age of 18 when treated, were treated with an HMA during the course of their disease, and were treated at our institution where electronic medical records (EMR) were complete and readily accessible. Exclusion criteria were patients who had died as a direct result of their hematopoietic stem cell transplant (HSCT) and whose records were incomplete or who had been treated at an outside institution. Logistic regression models were used to analyze data. Survival was calculated from day one of the first cycle of HMA received. Results A total of 68 patients were evaluated. Twenty-nine patients received IP-HMA while 39 received OP-HMA. Mean overall survival from HMA initiation was 252 days (95% CI 164-340) in the IP-HMA group and 430 days (95% CI 269-591) in the OP-HMA group. Mean 100-day survival in the IP-HMA group was 78 days (95% CI 65-91) and 93 days (95% CI 87-99) in the OP-HMA group. The average age at HMA initiation in the IP group was 69 (95% CI 64-74) and was 66 (95% CI 62-70) in the OP group. Average Charlson co-morbidity index was 5.5 (95% CI 5.0-6.0) in the IP group and 5.6 (95% CI 5.0-6.2) in the OP group. In the IP group, there were 1 low risk, 7 intermediate risk, and 21 high risk patients according to the European LeukemiaNet (ELN) risk stratification scale. In the OP group, there were 3 low risk, 8 intermediate risk, and 28 high risk patients. Nineteen of the 29 patients in the IP group and 11 of the 39 patients in the OP group received venetoclax in addition to an HMA as part of their therapy. Using a logistic regression model, location of HMA induction was found to have a significant effect on 100-day survival with IP induction being associated with worse survival (odds ratio=3.94; p=0.028) (see figure 1). Co-variates for this model included age at HMA initiation as this was the only confounder found in Wilcoxon rank sum testing to be significant. Other confounders which were tested but not included in logistic regression due to non-significance included race, gender, ELN risk, and Charlson co-morbidity indices. Discussion In this retrospective study, we found that IP-HMA induction was associated with a worse 100 day and overall survival than its OP counterpart. Though we found that co-morbidity indices and ELN risk scores were similar between the two groups, we postulate the reason the IP group had a significantly worse 100 day survival was that being admitted for neutropenic fever, risk of tumor lysis, etc. inherently put them at higher risk for complications like septic shock and hypoxic respiratory failure. Additionally, we found that patients who received HMA as an IP were far more likely to die shortly after their induction, within 100 days and often within 20 days. This would indicate they were potentially poor candidates for chemotherapy and that a palliative approach may have been more appropriate.. This study describes the survival differences seen in patients receiving IP-HMA versus OP-HMA. Knowledge of these results could spawn a more palliative approach for future patients requiring IP-HMA induction. Figure Disclosures Kasner: Otsuka Pharmaceutical: Research Funding; Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees, Research Funding. Palmisiano:AbbVie: Research Funding; Genentech: Research Funding.