Background: Secondary AML (sAML) is characterized by a unique cytogenetic and mutational profile that dictates inferior outcomes. Allogeneic transplantation (HSCT) is an encouraging therapeutic option for sAML, partially overcoming its adverse prognosis. Generalized use of next-generation sequencing (NGS) allows identification of secondary-type signatures, even in patients (pts) presenting with de novo disease (Lindsley C, et al. Blood 2015). The specific role of HSCT among pts harboring secondary-type gene mutations needs further investigation.

Methods:The study aimed to compare the NGS mutation profile and assess its prognostic impact in pts presenting with de novo AML versus sAML undergoing first HSCT between 2010-2023.The two mutation profiles were compared by multiple correspondence analysis and circle plot interaction schemes. Most (91.8%) of the pts were in CR1. Donors were unrelated in 39.9%, matched related in 33.6%, haploidentical in 19.7% and ‘other’ in 6.7%. Univariate analyses were performed using the log-rank test for LFS, OS, and graft-versus-host disease (GVHD)-free, relapse-free survival (GRFS), while RI and NRM were calculated using the cumulative incidence method and compared using Gray's test. Multivariate analysis (MVA), adjusting for potential confounding factors, was performed using a Cox proportional hazards regression model for the main outcomes.

Results: 762 pts were included, 699-denovo AML and 63 with sAML after MDS or MPN in the majority of pts (95.2 %). Follow-up was 2 years (range, 1.9-2.1). Median year of transplant was 2020 (range, 2014-2023). Pts with sAML were older, 61.1 (range, 23.5-73.5) vs. 53.9 (range, 18.5-73.9) years (p<0.001), and 58.7% vs. 48.6% were male (p=0.13). The two groups did not differ in performance status, cytomegalovirus seropositivity, female-to-male combination, and time from diagnosis to transplant. The cytogenetic risk was intermediate in 69.1% and adverse in 26.2%, without a difference between sAML and de novo (69.3% and 25.5% vs. 67.2% and 32.8%, respectively; p=0.1). De novo AML was enriched in NPM1 (20.7% vs. 7.9%, p=0.015) and FLT3-ITD mutations (24.5% vs. 9.8%, p=0.009), whereas mutations in JAK2 (11.5% vs. 1.7%, p<0.001); splicing factors SF3B1 (11.5% vs. 4.9%, p=0.04), SRSF2 (29% vs. 8.9%, p<0.001), U2AF1 (10% vs. 2.8%, p=.012), ZRSR2 (11.6% vs. 1.5%, p=0.002), and TP53 (16.7% vs. 5.7%, p=0.004) were more prevalent among sAML pts. Pts with sAML received BM grafts (9.5% vs. 4%; p=0.047) and reduced intensity conditioning (61.9% vs. 44.2%; p=0.008) more frequently than de novo AML pts. The 2-year LFS and OS rates were 47.5% vs. 59%, hazard ratio (HR)=1.01 (95% CI 0.65-1.57, p=0.95), and 55.5% vs. 68.9%, HR=1.0 (95% CI 0.63-1.59, p=0.99), in sAML vs. de novo AML, respectively. 2-year RI was 35.4% vs. 27.7%, HR=0.86 (95% CI 0.49-1.5, p=0.560), and 2-year NRM was 17.1% vs. 13.2% HR=1.12 (95% CI 0.57-2.45, p=0.62), for sAMLvs.de novo AML. The original disease was the main cause of death in both groups, 58.3% vs. 57.2% of those who died, respectively. The incidence of day 180 acute GVHD grades II-IV was 25.1% vs. 24.2%, and grades III-IV was 5.5% vs. 7.9%, while 2-year total chronic GVHD was 36.6% vs. 38.8%, and of extensive chronic GVHD was 11.1% vs. 13.5%, respectively, both with no significant differences between the two groups. GRFS was 43% vs. 48.4 %, HR=0.99 (95% CI 0.67-1.48, p=0.96) in sAML vs. de novo AML, respectively. Day 30 absolute neutrophil count (> 109/L) was 89.8% vs. 95.7 % and day 60 platelet count (> 20x109/L) was 91.4% vs. 92.5%, respectively.

On MVA, disease status at transplant (CR1 vs. other) had an independent prognostic impact on OS, LFS, RI, and GFRS. TP53 mutation showed a negative prognostic impact on RI and OS. Cytogenetic risk (adverse vs. other) showed a negative impact on all main outcomes (OS, LFS, RI, NRM, and GRFS), and increasing age negatively influenced NRM. An analysis restricted to pts receiving HSCT in CR1 showed that TP53 mutation retained its adverse impact on OS, LFS, and RI; adverse cytogenetics had a negative impact on OS, LFS, RI, and GRFS, and recipient age had a prognostic impact on NRM.

Conclusions: This study confirmed differences in the mutational landscape between de novo and sAML subtypes. Nonetheless, AML presentation type did not show a prognostic impact, whereas the presence of TP53 mutation and adverse risk cytogenetics were the only genetic disease factors showing prognostic impact.

This content is only available as a PDF.
2025
Sign in via your Institution