Discussion of PARADIGM (Fathi AT et al., N Engl J Med 2026): a randomized phase 2 trial that compared azacitidine + venetoclax with intensive induction chemotherapy in 172 adults with newly diagnosed acute myeloid leukemia who were *eligible for induction* — that is, precisely the population in which the less intensive regimen was not yet standard. The median age was 64 years and 72% had adverse risk disease according to the ELN 2022 classification.

With a median follow-up of 21.9 months, event-free survival was 14.5 months with azacitidine-venetoclax versus 6.2 months with induction (HR 0.57; 95% CI 0.39–0.84; p = 0.002). Overall survival was 21.5 versus 18.0 months, with no statistically significant difference (HR 0.82; 95% CI 0.52–1.28). The less intensive arm had less toxicity: grade 3 or higher infection in 28% versus 41%, grade 3 or higher hemorrhage in 2% versus 12%, 30-day mortality of 0% versus 3%, no ICU admissions versus 10%, and 12.5 versus 27.3 days of hospitalization in the first month. 60% versus 40% reached transplant.

The point that generates discussion is the design: it is a *phase 2* study, with a one-sided alpha of 0.10 and without type I error control for secondary outcomes — so the difference in access to transplant is a descriptive finding, not a tested comparison. Add a 68% crossover rate among those who failed induction, and it becomes clear why overall survival doesn't differentiate. We discuss what these data change (or don't change) in practice, and what still depends on a phase 3 trial.

Article: https://doi.org/10.1056/NEJMoa2602804

Presentation: Lucas Salviano
Discussion: Carol Marques, Paula Cividanes, Rafael Faust, Luiza Miranda, Israelita Sussuarana, Pedro Moraes, and Vinícius Molla

Chapters:
00:00 Introduction
00:17 Article Presentation
19:53 Discussion