Detection of MEN1 resistance mutations in cell-free DNA from acute leukemia patients treated with menin inhibitors

Abstract

Abstract Background: Menin inhibitors have demonstrated promising clinical efficacy in multiple subtypes of acute leukemia, with revumenib being the first approved for relapsed/refractory KMT2A-rearranged leukemias. However, resistance frequently emerges and limits durable remissions. We previously identified somatic MEN1 mutations, detected by the MSK-IMPACT Heme next-generation sequencing (NGS) assay, as a key mechanism of genetic resistance. These mutations impair drug binding but preserve the ability of the mutant menin protein to assemble leukemogenic chromatin complexes and sustain HOX gene expression. Clinical detection of MEN1 resistance mutations relies on targeted NGS of DNA extracted from bone marrow (BM) aspirates. This invasive procedure is logistically more complex and less amenable to serial monitoring than peripheral blood sampling. However, sampling of peripheral blood mononuclear cells has limited sensitivity in patients without high circulating disease burden. We hypothesized that peripheral blood cell-free DNA (cfDNA) can serve as a minimally invasive, scalable, and clinically impactful alternative for detecting emergent MEN1 resistance mutations in menin inhibitor-treated patients. Methods: We prospectively collected peripheral blood samples from acute leukemia patients undergoing menin inhibitor treatment at Memorial Sloan Kettering Cancer Center. Additionally, we obtained BM derived genomic DNA (gDNA) samples from 4 patients with confirmed MEN1 resistance mutations, as identified by MSK-IMPACT Heme. A high-sensitivity research NGS assay we named MSK-ACCESS-MEN1 was designed by including probes to cover all MEN1 exons into the existing MSK-ACCESS Heme assay (Gedvilaite E et al, 2022), a unique molecular indexed ultra-deep sequencing test for cfDNA interrogation in hematologic cancers. Plasma-derived cfDNA was extracted from pre-, on-, and off-treatment samples and sequenced with the customized research panel. MSK-IMPACT Heme testing was performed on matched BM aspirates when available. The detection, sensitivity, and concordance of different assays were evaluated. Results: To validate the performance of the MSK-ACCESS-MEN1 assay, we first analyzed gDNA from 4 different patient BM samples with known MEN1 resistance mutations. MSK-IMPACT Heme identified 8 MEN1 mutations across 4 patient samples, all of which were confirmed by MSK-ACCESS-MEN1, demonstrating 100% concordance. To assess the potential clinical utility of MSK-ACCESS-MEN1, we analyzed plasma-derived cfDNA from 2 patients with prior menin inhibitor therapy who had undetectable or very low-level peripheral blasts and known active BM disease with MEN1 resistance mutations. MSK-ACCESS-MEN1 identified all MEN1 mutations from plasma-derived cfDNA, including 3 different polyclonal MEN1 resistance mutations in a leukopenic pediatric patient with no reported peripheral blasts. We next analyzed a cfDNA sample from a patient with NPM1m AML who received 3 cycles of revumenib and had a BM aspirate which showed 11% blasts by flow cytometry, but without detectable MEN1 resistance mutations by MSK-IMPACT Heme. MSK-ACCESS-MEN1 evaluation of cfDNA from the same day identified an MEN1 M327V resistance mutation with an allele frequency of 1.6%. The patient subsequently had progressive disease, and 4 weeks later the MEN1 M327V resistance mutation was identified in the BM by MSK-IMPACT Heme. Lastly, MSK-ACCESS-MEN1 identified an emerging MEN1 M322I mutation in cfDNA from a patient with KMT2A-rearranged AML who had received 14 cycles of revumenib with an MRD-negative BM remission documented 3 months prior. At the time of cfDNA collection, she remained transfusion independent without evidence of relapse. She subsequently relapsed 3 months later with identification of the MEN1 M322I mutation by MSK-IMPACT Heme from a BM aspirate. Conclusions: Evaluation of plasma-derived cfDNA with MSK-ACCESS-MEN1 enables non-invasive detection of MEN1 resistance mutations in patients treated with menin inhibitors. MEN1 resistance mutations were reliably detectable in cfDNA from peripheral blood in patients without circulating blasts. Importantly, we demonstrate that MSK-ACCESS-MEN1 can identify emerging MEN1 resistance mutations prior to clinical evidence of relapse. Our findings support the incorporation of cfDNA monitoring into clinical protocols for menin inhibitor therapy.