When Intensification Belongs Upfront vs. Post-Remission: Balancing Cure and Toxicity

A critical question is when in the treatment timeline therapeutic intensification should occur: during induction or post-remission. The panel emphasizes that AML management increasingly requires careful sequencing rather than simply escalating or de-escalating therapy globally. The speakers explore the rationale for early versus delayed intensification, particularly in the context of emerging triplet regimens and targeted therapies.

One of the panelists raises a key concern: although triplets may achieve deep remissions without the toxicity of intensive chemotherapy, some patients may still require transplant. If intensification is deferred until later, at a moment when the leukemia burden is low, patients may tolerate it better, but clinicians risk missing the window to control aggressive disease at diagnosis. In contrast, applying maximal intensification upfront may benefit biologically high-risk patients but carries greater immediate toxicity.

The group reflects on real-world examples in which even patients with favorable-risk features (such as NPM1-mutated AML) sometimes ultimately require transplantation after intensive therapy, reminding the audience that biologically “favorable” does not guarantee long-term cure without consolidation or transplant. They point out that this uncertainty complicates decisions about whether a low-intensity triplet is sufficient on its own or whether certain patients should still receive intensive induction regardless of early response depth.

Panelists also discuss how MRD may eventually help settle this dilemma by allowing for response-adapted strategies. However, MRD-guided treatment in AML is still evolving, and current evidence is insufficient to rely on MRD to decide between induction intensities in most cases.

Although low-intensity induction continues to expand, the decision to escalate therapy, either upfront or later, will remain patient-specific, mutation-guided, and influenced by future clinical trial data.