No abstract

Background:Patients (pts) with acute myeloid leukemia (AML) may be treated with intensive or non‐intensive chemotherapy. While some pts achieve complete remission (CR) after initial treatment, a significant proportion become refractory to initial treatment or relapse after the initial response.Aims:To understand treatment patterns and outcomes in pts with relapsed or refractory AML (RR‐AML) in a real‐world setting.Methods:The Alberta Cancer Registry and local databases of the University of Alberta Hospital and the Tom Baker Cancer Centre were interrogated to identify AML pts with RR‐AML aged ≥18 years treated from January 2013 to December 2016. Pts were considered to have refractory AML if they failed to achieve CR or achieved CR with incomplete count recovery (CRi) after 2 cycles of intensive chemotherapy, 6 cycles of azacytidine, or 4 cycles of low‐dose cytarabine. Based on the treatment regimen received following the diagnosis of RR‐AML, pts were grouped as receiving either intensive therapy (IT), non‐intensive therapy (NIT), or best supportive care (BSC) and were followed from relapse to date of death or last follow‐up.Results:Overall, 572 pts with AML were identified from the database search, 199 (124 males, 75 females) of whom met the eligibility criteria for RR‐AML and were included in this analysis. The median age at diagnosis of RR‐AML was 66.8 years; median follow‐up was 4.7 months. According to the European LeukemiaNet (ELN) 2010 classification, 34 pts (17%) had a favorable risk, 102 (51%) intermediate (Int) risk, 59 (30%) adverse risk profile and 4 (2%) with unknown status. After relapse or refractoriness (RR), 88 pts (44%) received BSC, 46 (23%) received IT with fludarabine, cytarabine + granulocyte colony‐stimulating factor (FLAG) (n = 5), FLAG + idarubicin (n = 29), or another regimen (n = 12), while 65 (33%) received NIT with azacitidine ± another agent (n = 49), low‐dose cytarabine ± another agent (n = 12), or an alternative regimen (n = 4). The unadjusted median overall survival (mOS) for the overall study cohort was 5.3 months from the time of RR with a 12‐month OS rate of 29.6% (95% CI 29.0–30.3). The mOS was 13.8, 9.4, and 2.1 months for IT, NIT, and BSC groups, respectively (P < 0.001) (Figure). The mOS for pts aged <60 years was 8.3 vs 4.5 months for those ≥60 years (P = 0.009). Following RR, 16 (8%) pts received an allogeneic stem cell transplant (ASCT), for whom median survival was not reached, vs 4.5 months in pts who did not undergo transplantation. The mOS for pts with an ELN favorable, Int‐I, Int‐II, and adverse risk profile was 12.4, 4.5, 4.7, and 4.0 months, respectively (P = 0.002). In a multivariable Cox regression model incorporating age, ELN risk group, treatment intensity pre‐RR, number of treatment lines pre‐RR, best response pre‐RR, treatment intensity post‐RR, and ASCT post‐RR, treatment intensity post‐RR (NIT vs BSC: hazard ratio [HR] for mortality 0.32; 95% CI 0.23–0.48; IT vs BSC: HR for mortality 0.26; 95% CI 0.16–0.43) and best response pre‐RR (CR/CRi <12 months vs all others: HR for mortality 0.55, 95% CI 0.33–0.92) were significantly associated with OS.Summary/Conclusion:In a real‐world setting, a high proportion of pts only receive BSC, which is associated with very short survival. Treatment regimen post‐RR is the major predictor of survival, with non‐intensively and intensively treated pts having better outcomes compared to pts who received BSC. However, the overall outcomes of pts with RR‐AML remain poor regardless of treatment, and new therapies are urgently needed.image

e18505 Background: In order to describe the impact of future targeted therapies on treatment outcomes of patients (pts) with relapsed and/or refractory (RR) acute myeloid leukemia (AML), a better understanding of the clinical management pathway in these pts is needed. We therefore evaluated the treatment patterns and associated outcomes in a real-world cohort of pts with RR-AML using a population-based cancer registry and patient medical records. Methods: Pts newly diagnosed wih AML between January 2013 and December 2016, aged ≥ 18 years were identified from the provincial-wide Alberta Cancer Registry (ACR). Data for pts who met the criteria for RR-AML were assessed by hematologists and were extracted from medical records. RR-AML pts were then categorized as: receiving intensive therapy (IT); receiving non-intensive therapy (NIT); or treated with best-supportive care (BSC) following a diagnosis of RR-AML. Results: 572 AML pts were identified from the ACR, of which 199 met criteria for RR-AML and were included in the analysis (124 males, 75 females; median age at diagnosis of RR-AML 66.8 years; median follow-up 4.7 months). In this RR-AML cohort, 26 (13%) pts received ≥ 2 lines of prior therapy. Unadjusted median overall survival (mOS) was 5.3 months, with a 12-month overall survival rate of 29.6% (95% CI 29.0–30.3%) from the time of RR. Following RR, 46 (23%) pts received IT, 65 (33%) pts were treated with NIT, and 88 (44%) pts received BSC, with unadjusted mOS of 13.8, 9.4, and 2.1 months, respectively ( P < 0.001). When stratified by European LeukemiaNet risk classification at diagnosis, unadjusted mOS was 12.4, 4.7, and 4.0 months for favorable risk, intermediate risk, and adverse risk groups, respectively ( P < 0.01). Conclusions: This retrospective, real-world study in Alberta Canada confirms the poor prognosis reported to date in the RR-AML population. Notably, a large proportion of pts received BSC which was associated with dismal survival outcomes. These data also highlight that effective and tolerable alternatives to current treatment options are urgently needed.