Midostaurin Shows Efficacy in Refractory Multiple Myeloma with TP53 Mutation and Elevated Expression of NTRK1, NTRK3 and AURKA

Introduction

Multiple myeloma (MM) displays enormous genetic complexity and heterogeneity, resulting in deregulated cell signaling, aberrant cell growth and survival. Although, recurrent myeloma alterations are well characterized, their impact on drug response is often unclear. To decipher complex gene-drug interactions, we comprehensively assessed drug responses of 100 MM samples to 142 drugs and compared to genomic, transcriptomic and cytogenetic information, plus clinical outcome of the individual patients. Our results reveal novel insights on i) drug response and resistance mechanisms, ii) genomic and transcriptomic indicators of drug response, and iii) drug repositioning possibilities.

Methods

Bone marrow aspirates were collected from MM patients (n=100; newly diagnosed (NDMM) n=34; relapsed/refractory (RRMM) n=66) and healthy individuals (n=14). CD138-enriched plasma cells were screened against 142 oncology drugs tested in a 10,000-fold concentration range. Whole exome and RNA sequencing were performed for 97 and 78 samples, respectively. To identify pharmacogenomic interactions, we applied differential gene expression profiling, ANOVA and regression analysis.

Results

Overall, the RRMM samples exhibited a larger number of mutations and also showed a larger number of clones as compared to NDMM samples (3.3±0.6 vs 2.7±0.6, p=6e-03). 32% of patient samples showing exquisite sensitivity to signal transduction inhibitors were characterized with high mutation burden (170± 31 compared to other patients 84±13,p=5e-03). Additionally, the gene expression profiles of the highly sensitive samples were associated with cell proliferation, corresponding to previously reported proliferation subtypes (UAMS) with poor prognosis. 14% of the samples that exhibited a multidrug resistant (MDR)phenotype and were resistant to proteasome inhibitors, immunomodulatory drugs and glucocorticoids, exhibited elevated expression of multidrug resistance genes ABCB1 (MDR1) and ABCC3 (MRP3). 28% of the resistant samples were from del(17p) patients. Integrative analysis allowed assimilation of multiple datasets and revealed putative molecular markers associated to individual drugs. NPM1 expression was found to associate with bromodomain inhibitor JQ1 sensitivity, while CCND1 and BMI1 expression correlated with sensitivity to BCL2 inhibitor venetoclax. A hyperdiploid gene expression signature was enriched in glucocorticoid responding patient samples. While RAS pathway mutations (NRAS, KRAS, NF1, and BRAF) were detected in 56% of the patient samples and are recognized as major driver events in MM, MEK/ERK inhibitor efficacy was limited to 55% of samples bearing clonal mutations to these genes. 20% of the patient samples harbored mutations in genes involved in DNA damage repair signaling, namely ATM, ATR, MSH3, BAX, TP53BP1, and TP73, in a mutually exclusive pattern. Patients with these mutations had a high relapse rate, poor overall survival (HR=3.2,95%CI 1.1-9.02), and displayed acquired sensitivity to PI3K-mTOR and HDAC inhibitors. Sensitivity to midostaurin, as observed in FLT3 mutated AML, was detected in a subgroup of RRMM samples (20%) with mutated TP53 and elevated expression of NTRK1, NTRK3 and AURKA .

Summary

Based on their genomic and transcriptomic profiles, we could identify subsets of patients responding to approved drugs and novel compounds such as venetoclax, midostaurin and JQ1 exhibiting potent activity in myeloma. In support of previous studies, we found that alterations in DNA damage signaling pathways contribute to poor prognosis, however, samples with these mutations showed enhanced sensitivity to PI3K-mTOR and HDAC inhibitors. Our study demonstrates that molecular information and ex vivo drug profiling may be useful to develop tailored treatment strategies, identify responding patients and guide treatment decision, especially for relapsed/refractory myeloma patients.