Treatments and Outcomes of Adult Patients with AML in the Real-Life - First Report of the French Observational ALFA-PPP Study

Introduction. AML treatment options have been recently enriched. New agents like gemtuzumab ozogamicin (GO), FLT3 or IDH inhibitors, or CPX-351 have been approved for some patient subsets. Less-intensive options, like the AZA-VEN combination, are used in older/unfit patients. However, while treatments and indications are standardized, their choice is not, depending on often-subjective factors. This makes observational studies so important to describe the real-life management and outcomes of AML patients. Methods. In 2022, the ALFA group initiated a 2-cohort (newly diagnosed, relapsed/refractory) prospective observational study (ALFA-PPP, NCT04777916) aiming to collect clinical data and bio-samples in all AML patients aged ≥18y referred to 27 ALFA centers. Data/sample collection was structured by approved therapies including best supportive care (BSC). Patients treated on a compassionate basis or included in a clinical trial are also registered. Treatment decision-making criteria are prospectively collected. All patients should have a centralized genomic diagnosis (50-gene NGS panel) and molecular/flow MRD follow-up. We report here the initial observations made in the first 648 patients included between April 2022 and September 2023 in the newly diagnosed cohort. The current patient recruitment rate is around 650 new patients per year. Results. There were 340 males and 308 females (median age, 65y [range, 18-100]; ECOG-PS 0/1/2/3/4/NA, 195/295/97/33/8/19; HCT-CI comorbidity index 0/1/2/3+/NA, 54/269/106/216/3; median WBC, 6.9 G/L [IQR, 2.4-26]; median marrow blast percentage, 53% [IQR, 29-79]). The numbers of patients with de novo, post-MDS, post-MPN and therapy-related AML (tAML) were 467 (72%), 54 (8%), 58 (9%) and 69 (11%), respectively. ELN-2022 risk was favorable in 136 (21%), intermediate in 129 (20%), adverse in 340 (52.5%), and not available in 43 (6.5%) patients. In the 595 patients tested, incidences of NPM1, FLT3-ITD, IDH1/2, TP53 and secondary AML (sAML)-like gene mutations were 27%, 19%, 22%, 12% and 48%, respectively. A minority of 172 (26.5%) patients were included into a clinical trial frontline (investigator/company-sponsored trial, 140/32 patients; intensive/less intensive/two-option trial, 142/20/10 patients). They were significantly younger than the 476 (73.5%) patients treated outside any clinical trial (median age, 55 vs 69y, p<.001). In these 476 patients, treatment decision was intensive in 266 (median age, 62y [18-79]; 210 7+3 ± GO or midostaurin, 54 CPX-351, 2 others), less intensive in 185 (median age, 76y [36-100]; 149 AZA-VEN, 18 AZA or low-dose cytarabine alone, 13 AZA-ivosidenib, 5 others), and BSC in 21 patients, while the 4 remaining patients early died before any treatment decision. By investigator declaration, the main reason for selecting a less intensive treatment was advanced age (n=138), AML characteristics/risk (n=19), concomitant diseases (n=16), patient's choice (n=1), or other (n=11), with good statistical correlations with objective data. The objective factors independently associated with the choice of a less intensive option were age ≥65y (p<.001), higher ECOG-PS (p<.001) or comorbidity index (p=.001), but also post-MDS (p=.001) or post-MPN (p<.001) AML, lower WBC (p=.022), poor cytogenetics (p=0.046), and TP53 (p=.003) or IDH1/2 (p=.003) mutation. A total of 89/476 (18.5%) patients received an allogeneic SCT during the first 6 months of follow-up, including 81 intensively and only 8 less-intensively treated patients. In the 185 less-intensively treated patients with a median follow-up of 15 months, median overall survival (OS) was 10.3 months (95% CI, 7.2-12.9) with a 12-month estimate at 45.3% (95% CI, 37.7-52.5). In these patients, multivariate stepwise Cox model evidenced higher ECOG-PS (p=.038), concomitant diseases (p=.003), tAML (p=.046), higher marrow blast percentage (p=.024), sAML-like (p=.048) and TP53 (p<.001) mutation as bad-prognosis factors for OS, while IDH1/2 mutations (p=.001) were associated with longer OS. Conclusions. Awaiting the ELN-2024 recommendations for less-intensive AML therapies, this prospective real-life study confirms their rising place in a European context. Rather than age by itself, survival after less-intensive treatment is governed by multiple independent clinical and genomic factors. Future reports will update these observations on an annual basis.