Introduction

In older patients with acute myeloid leukaemia (AML) not considered suitable for intensive chemotherapy, outcomes are poor. Previous work (Burnett et al, 2007) has established that low dose cytarabine (LDAC) significantly improves overall survival (OS) compared to best supportive care. However, in this evaluation, no complete remissions, with or without count recovery, (CR/CRi) were seen in patients with adverse cytogenetics. With the advent of hypomethylating agents, the attainment of CR is not always a prerequisite for improved OS: with repetitive cycles of therapy clinically significant benefit can be obtained without CR. Studies with azacitidine (AZA) when directly compared to LDAC have not demonstrated significant superiority in MDS (Fenaux 2009) or AML (Dombret 2015). In the latter study, the adverse cytogenetic group had a median number of 2 cycles of LDAC administered, compared to 5 cycles of AZA. The lack of a CR in patients treated with LDAC may cause premature termination of therapy. We therefore evaluated the cumulative experience of 17 years of LDAC as our standard non-intensive therapy in three sequential AML trials: AML14 (1999-2005), AML 16 (2006-2011) and the ongoing LI-1 trial (2011-2016) to gain further insight how outcomes with the same treatment may have evolved and potential predictors of outcome.

Methods

Retrospective analysis of three randomised trials of LDAC versus one of nine comparator arms. Only patients contributing to closed randomisations are evaluated in the analysis. Logistic and Cox regression analyses were used to evaluate predictors of CR, survival, and early mortality. In all cases LDAC was administered as 20mg SC twice daily for 10 days every 4-6 weeks. Eligibility criteria were the same across trials, with the exception of cardiac criteria for ganetespib.

Results

Between Jan 1999 and March 2016, 683 patients, median age 75 (range 51-90) were randomised to LDAC. They were de Novo (62%), Secondary AML (25%) or high-risk MDS (12%); 5% had WHO performance status (PS) 2+; 61% male; median WBC was 7.0 (range 0.5-336 x109/L). Cytogenetic status was known for 492; 27% were adverse risk. Comorbidity status using HCT-CI was collected in AML16 and LI-1; 38% had no recorded comorbidity; 35% a HCT-CI score 1-2, and 27% score 3+. Age, sex, diagnosis and PS were similar over time; AML14 patients had lower WBC, and a greater proportion of adverse risk karyotype patients; LI-1 patients had fewer comorbidities. The median number of cycles administered was 2 in AML14/AML16 and 3 in LI1 (range 1-8 in all trials). With median follow-up for survival of 51.2 months (range 0.1-97.1), median OS was 160 days (1-year OS 27%); the cumulative CR/CRi rate was 19%, the CR/CRi rate in patients with adverse cytogenetics was 14%. Survival outcomes improved over trial (HR 0.86 (0.77-0.96) p=0.01; Figure, Table). There was a significant improvement over trial in 30-day mortality (23% vs 14% vs 13% HR 0.71 (0.54-0.94) p=0.01); beyond day 30 there was also a trend to improving OS (HR 0.90 (0.79-1.0-2) p=0.08). Median time to response was 95 days; fewer than 25% of responders achieved CR/CRi within 60 days. In univariate analyses, CR/CRi was affected by the presence of secondary disease (p=0.0004); age (p=0.05); WBC (p<.0001); PS (p<.0001); adverse cytogenetics (p=0.05), as well as comparator arms available (p=0.05) and the number of arms available (NOAA) (p=0.03). Prognostic factors for survival were WBC (p<.0001); adverse cytogenetics (p=0.03); year of entry (p=0.01) and NOAA (p=0.03). In multivariate analyses the significant prognostic factors for CR/CRi were WBC (p=0.001) and NOAA (p=0.04); for OS these were WBC (p<.0001); PS (p<.0001), and trial (p=0.02).

Conclusions:

We provide updated reference data for LDAC as therapy in this group of patients. Over time, outcomes with LDAC have improved, with a reduction in early mortality, and more courses of therapy being delivered. Responses generally require 3 months of therapy and are observed in patients with an adverse karyotype as well as intermediate risk disease. These results show the need for patience in treatment with LDAC and draw into question the results of trials where duration of LDAC therapy is short. The association between comparator and outcome may also indicate selection, making interpretation of historically controlled or single arm studies in this population difficult.

Disclosures

Copland: Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Conference attendance supported, Research Funding, Speakers Bureau; Celgene: Honoraria, Other: Conference attendance supported, Speakers Bureau; Takeda: Honoraria, Speakers Bureau; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Conference attendance supported, Speakers Bureau; Pfizer: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Incyte: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau. McMullin: Italfarmaco S.p.A.: Consultancy; Shire, Novartis: Honoraria, Speakers Bureau. Ali: Novartis: Honoraria; BMS: Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution