P-351: Are genomic lesions of Cereblon pathway genes sufficient to explain clinical relapses from IMiD-based treatment in multiple myeloma (MM)?

Introduction: In preclinical CRISPR studies, knockout (KO) of Cereblon (CRBN) or members of its pathway (CRBN-pathway genes, CRBNPGs) are the top perturbations causing MM cell resistance to immunomodulatory thalidomide derivatives (IMiDs), but partial CRBN loss-of-function (LOF) still preserves substantial IMiD responsiveness. Mutations of CRBNPGs were reported in relapses from IMiD±Dex treatment, but it is unclear if they cause complete LOF of CRBNPGs in a way that can explain why MM patients (pts) relapse after IMiD-containing regimens.

Methods: We evaluated the molecular profiling alterations of 42 CRBNPGs (vs. a set of 42 known MM driver genes; defined by prior studies e.g. Ansari-Pour et al Blood 2023) in pts of the MMRF CoMMpass study (IA22) who had genomic or transcriptional analyses in ≥2 paired bone marrow samples before vs. after IMiD-based treatment (baseline, relapse[s]).

Results: Among 179 pts fulfilling inclusion criteria, 32 pts (17.9%) had at least one mutated CRBNPG at any timepoint (ptsmut). Of 25 CRBNPGs mutated in any pt, FAM83F was most frequently affected (n=4; 12.5%), followed by CRBN and IKZF3 (n=3 each; 9.4%). 19 pts (59.4%) had a CRBNPG mutation at baseline; 13 pts (40.6%) developed a new CRBNPG mutation at relapse; and 2 pts (6.3%) had >1 mutation (2 and 5 genes). Increase in variant allele frequencies (VAF) of at least one CRBNPG mutation occurred in 14 pts (43.8% of ptsmut), but VAF at relapse were < 0.5. Interestingly, preexisting CRBNPG mutations were not detectable at relapse in 9 pts (28.1%) or had decline in VAF in 3 pts. Clinical response (≥partial response) to IMiDs was documented in 11/13 pts (84.6%) who developed a new CRBNPG mutation and 11/14 pts (78.6%) with VAF increase, while 2 and 3 pts, respectively, had stable or progressive disease. Out of 19 pts with a mutation present at baseline, 2 (10.5%) were non-responsive to IMiD treatment. Importantly, no evidence for complete LOF for any CRBNPG (e.g., via combination of mutation, downregulation/alternative splicing, DNA copy number loss, fusion, etc) was identified at relapse or baseline. Most pts received IMiDs+Proteasome inhibitor (PI), but in our preclinical CRISPR studies, KO of CRBNPGs did not cause sensitization (or resistance) to PIs, thus genomic results for CRBNPGs conceivably were not skewed by PI use, e.g. via elimination of cells with complete LOF of CRBNPGs. In this set of pts, mutations in known MM driver genes (e.g., KRAS, BRAF, ARID1A, TP53, IRF4 among the most frequently affected) were detected at relapse in 26 pts (81.3% of ptsmut ) with heterogeneous VAF changes at relapse vs. baseline.

Conclusions: A minority of MM pts relapsing from IMiD-based treatments harbor genomic defects of CRBNPGs but typically not complete LOF for any such gene, and new/enriched mutations in known MM driver genes cannot be excluded as alternative explanations, beyond CRBNPGs, for these relapses. Additional and more complex, genomic or non-genomic, mechanisms may account for resistance to these regimens.