Background: CPX-351 is a liposomal formulation of daunorubicin and cytarabine, approved for the treatment of newly diagnosed therapy-related acute myeloid leukemia (AML) and AML with myelodysplasia-related changes in adults. Liposomal encapsulation markedly increases the plasma half-life of both drugs and leads to preferential drug accumulation within the bone marrow. CPX-351 has demonstrated reduced non-hematologic toxicities—such as alopecia, gastrointestinal side effects, and hepatic toxicity—while retaining hematopoietic cytotoxicity. Given the well-established cardiotoxicity of conventional anthracyclines, it is important to determine whether CPX-351 is associated with similar, lesser or worse cardiotoxity compared with standard anthracycline therapy used in induction therapy for AML (“7+3”). If lesser toxicity is shown, CPX-351 could potentially be preferred in patients with preexisting cardiac dysfunction or elevated risk for heart failure, and may reduce the need for cardioprotective agents such as dexrazoxane. As with any recently introduced agent that may be used to replace standard therapy, it is crucial to fully characterize any toxicity profile, including with long-term follow-up.

Methods: We previously conducted a pilot study of CPX-351 with or without gemtuzumab ozogamicin (GO) in 25 adults (age <60) with newly diagnosed, FLT3-ITD–negative, favorable- or intermediate-risk AML (Ganzel C et al. BJH 2025). Patients received a single induction course of CPX-351 ± GO. No CPX-351 consolidation was administered. In this study, we also prospectively assessed echocardiographic data—including ejection fraction (EF) and global longitudinal strain (GLS)—at baseline, post-induction, and long-term follow-up.

Results: Of the 25 patients, 15 had EF measured at all three timepoints. The median time from baseline to post-induction ECHO was 1.3 months; from post-induction to follow-up was 20.9 months. The median relative EF change from baseline to post-induction was –4.6%, and from baseline to follow-up was –4.3%.

Three patients (20%) experienced a sustained relative EF decline ≥10% at follow-up. Two of these showed marked early EF decline post-induction (–20.0% and –17.9%) with partial recovery at follow-up, while the third had a persistent decline (–13.0%).

GLS was available in four patients with complete serial ECHOs. The median relative GLS change from baseline to post-induction was +11.6%, indicating early subclinical impairment (i.e., GLS becoming less negative). At follow-up, the median change from baseline was +13.4%. Two patients (50%) showed GLS worsening >15%, reflected by substantial reductions in myocardial strain post-induction (+33.6% and +30.8% relative increases), followed by partial improvement at follow-up (+19.2% and +22.9%). One of these patients also had a relative EF decline ≥10%.

Conclusions:In this cohort of 25 younger adults with favorable/intermediate-risk AML treated with a single CPX-351–based induction (±GO), the median EF and GLS changes were mild and consistent with subclinical cardiotoxicity. Three patients (12%) experienced a sustained relative EF decline ≥10%, and two patients (8%) had relative GLS worsening >15%, including one who met both criteria. These five patients (20%) represent the subset with significant cardiac changes.

In this cohort, both EF and GLS reductions were already evident post-induction in patients with significant cardiac changes. GLS showed partial recovery in both affected patients at follow-up, while EF remained persistently reduced. These findings suggest that GLS and EF may capture overlapping but not identical aspects of anthracycline-related cardiotoxicity. Continued prospective monitoring using both parameters is warranted to better define the cardiac safety profile of CPX-351. Furthermore, in this single-arm study no similar evaluation of cardiac toxicity was made in patients receiving standard “7+3” induction therapy. It is, therefore, impossible to make any statement regarding the cardiac toxicity of CPX-351 relative to “7+3”. For this, a formal prospective comparison would be needed.

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