Page 168 - 2020_02-Haematologica-web

C. Talati et al.
Survival outcomes based on time periods
To account for changes that have occurred over the years in AML treatments and supportive care manage- ment, the entire cohort was divided and grouped accord- ing to the year of treatment initiation: Group A (treatment before 2005, n = 140) and Group B (treatment after 2005, n = 840). A trend towards improved median overall survival (mOS) was noted among the 2 groups but did not reach statistical significance (Group A vs. B, mOS 5.7 months vs. 7.3 months; P=0.051). Baseline characteristics of Group B is provided in Online Supplementary Table S1. When assess- ing for the survival outcomes of HMA-naïve patients from Group B, the difference in mOS due to chosen front-line therapy persisted (P<0.0001) (Online Supplementary Figure S1).
Survival outcomes in patients with previous hypomethylating agent exposure
We also assessed the efficacy of front-line treatments in the small subset of evaluable patients who had previously received HMA for non-AML diagnoses (n=185), focusing on identifying whether a benefit was seen in this sub- group versus high-intensity treatment. Of these 185 patients, 24 patients (13.0%) received HMA subsequently for AML diagnosis, 55 (29.7%) received high-intensity therapy, 32 (17.3%) received low-intensity therapy, and 74 (40%) received supportive care only. We noted similar- ly poor median overall survival among the HMA group (7.8 months), the high-intensity therapy group (5.9 months), and the low-intensity group (5.9 months). However, all three treatment groups had better overall sur- vival than the supportive care group (2.9 months) (P<0.0001) (data not shown). Moreover, multivariate analy- sis of the group also demonstrated the inferiority of sup- portive versus HMA and high-intensity and low-intensity therapy (Table 3). This improved survival versus support- ive care suggested that this subgroup may benefit from some other type of therapy rather than supportive care only.
Responses and early mortality rates
The rate of composite CR (CR and CRi) and 30-day TRM (defined as death within 30 days of treatment initi- ation) were compared among the HMA, high-intensity,
and low-intensity treatment groups (Figure 2). The rate of composite CR was significantly higher in the cohort treat- ed with high-intensity chemotherapy than in the HMA (43.1% vs. 22.7%; adjusted P<0.001) and low-intensity therapy groups (43.1% vs. 7.7%; adjusted P<0.001). Early TRM was significantly lower with HMA treatment at 1.2%, compared with 7.5% with high-intensity chemotherapy (adjusted P<0.01). Among the patients who achieved CR/CRi, we calculated a median relapse- free survival of 10.5 months with HMA versus 9.1 months with high-intensity treatment (P=0.09) and 4.4 months with low-intensity treatment, which was significantly inferior to both HMA (P=0.009) or high-intensity treat- ment (P=0.036). However, it should be noted that the low- intensity subgroup had an extremely small sample size (n=5).
Discussion
Treatment of elderly patients with AML is a therapeutic
challenge for clinicians as the choice of optimal front-line
regimens continues to remain controversial. Here we pres-
ent the results of the largest single institution report of
outcomes amongst AML patients ≥70 years old. Using
pairwise comparisons with propensity score matching,
our results indicated a survival benefit with front-line
HMA compared with high-intensity, low-intensity, or
supportive care therapies. These results confirm and
expand on previous reports that elderly patients with
AML can benefit from treatment over supportive care.2,10,19,20
Clinical trials with the HMA azacitidine or decitabine have previously demonstrated their ability to induce remission and prolong survival in elderly AML patients.16,19,21-23 After adjusting for potential treatment bias between the treatment groups with propensity score matching, we observed a statistically significant overall survival benefit with HMA versus our other treatment groups, with patients treated with HMA having median overall survival of 14.4 months. Our results were compa- rable to the 12.1 months previously observed by Dombret et al.15
Not unexpectedly, high-intensity chemotherapy was
Figure 1. Overall survival (OS) among various front-line therapies for acute myeloid leukemia (AML) in patients ≥ 70 years old. CI: confidence interval; HI: high intensity; HMA: hypomethylating agent; LI: low intensity.
402
haematologica | 2020; 105(2)