Prospective Evaluation of the Impact of Measurable Residual Disease (MRD) By Error Corrected Next-Generation Sequencing (NGS) with CPX-351 in Acute Myeloid Leukemia (AML)

Introduction

CPX-351 treatment significantly improved overall survival (OS) versus 7+3 in secondary AML patients in a Phase 3 study, particularly in patients bridged to allogeneic hematopoietic stem cell transplantation (allo-HSCT) with a median OS that was not reached (Lancet et al., JCO 2018, Uy G et al., Blood Advances 2022). However, there have been minimal data on the impact of molecular MRD and outcomes with CPX-351.

Methods

We prospectively collected bone marrow aspirates from AML patients treated with CPX-351 at Moffitt Cancer Center and Washington University. At each time point (pre-treatment/post-induction/post-consolidation/day+30 post-HSCT/relapse (as applicable)), we utilized a novel, ultra-deep, error-corrected gene panel (Myeloseq-HD^TM Capture Assay) that targets 49 genes that are mutated in myeloid malignancies for the purposes of mutation identification at diagnosis and disease monitoring after therapy (raw sequencing depth >50,000x and consensus read depth >4000x, which allows for detection of previously known variants at variant allele fractions (VAFs) greater than or equal to 0.1%.) Patients with variants detected only in DNMT3A, TET2, and/or ASXL1 (DTA) were considered MRD negative (occurred in 1 pre-HSCT and 2 post-HSCT samples). GraphPad Prism was used for survival analyses and relapse free survival (RFS), event free survival (EFS), and OS were calculated from date of CPX-351 initiation. EFS was defined as either relapse or death. For all endpoint analyses, an MRD cutoff of 0.1% VAF was used.

Results

A total of 48 patients were consented on study and treated with CPX-351 per standard of care/FDA indication with 33 patients having sequential NGS analysis. CR/CRi was achieved in 58% of patients (28/48). With a median follow up of 22.5 months, the median OS of the entire cohort was 15.8 months. Of the cohort, 56% of patients (n=27) were bridged to allo-HSCT with improved OS in transplanted patients compared to non-transplanted patients (22.5 vs 8 months, P=.007). Similarly, 37% of patients were alive at 2 years post-HSCT (n=10) vs 14% without HSCT (n=3). Patients with IDH1/2 and/or NPM1 mutation had improved CR/CRi rates (92% (12/13) vs 55% (11/20), P=.04). No other individual mutation or class of mutations predicted for response. Patients with secondary type mutations as defined by WHO achieved CR/CRi in 62% of cases. Only 3 TP53 mutant patients were treated and 1 achieved CR. Of the 33 patients with serial NGS, 13% of patients with CR/CRi (3/23) achieved MRD negativity at the 0.1% threshold post induction and/or consolidation. Of these, 2 patients did not proceed to allo-HSCT with one patient that is alive and disease free at 47 months (NPM1 mutant). We additionally analyzed NGS cutoffs of 1% and 5% for the entire cohort, but this was not predictive of OS, EFS, or RFS. For patients not proceeding to allo-HSCT, we additionally analyzed 1% and 5% NGS cutoffs, which were not predictive of OS, EFS or RFS. Among mutation-defined subsets, only patients with IDH1/2mutations with clearance of mutation to < 1% had an improved RFS (P=.04) although there was no improvement in OS. The one patient who achieved MRD negativity prior to allo-HSCT was alive and disease free at 38.6 months (ASXL1, IDH2, STAG2, and ZRSR2). NGS clearance cutoffs of 1% and 5% pre-allo-HSCT were not predictive of outcomes. 8 patients had post allo-HSCT MRD measured (Day +30) and 88% (n=7) were MRD negative. The median OS, EFS and RFS were not reached in these patients with improved EFS/RFS versus the entire cohort (P<.05).

Conclusions

In this prospective real-world cohort of CPX-351 treated AML patients, outcomes were similar to the Phase 3 trial with > 50% of patients proceeding to allo-HSCT and with encouraging rates of prolonged survival. Patients with NPM1/IDH1/IDH2 had distinctly favorable outcomes. Molecular MRD clearance to < 0.1% was rare with CPX-351 treatment, suggesting the need for novel frontline combinations to increase molecular MRD negativity. However, patients treated with CPX-351 and bridged to allo-HSCT had very high day +30 MRD negative rates and excellent outcomes (85% without relapse).