OA-24: Sequential and persistent loss of BCMA and GPRC5D after bispecific antibodies in multiple myeloma revealed by whole genome sequencing

Introduction: We and others have recently identified biallelic hits in genes coding for immunotherapy targets in Multiple Myeloma (MM) as a mechanism of resistance, such as in BCMA. Similarly, biallelic events in the GPRC5D locus on chromosome 12p have been linked to antigen loss and relapse from GPRC5D targeted therapies. As we are entering an era in which BCMA and GPRC5D targeting agents are approved and available, the question how to sequentially treat patients with anti-BCMA and anti-GPRC5D agents becomes imminent. A more profound knowledge about the selective pressure of sequential immunotherapy on the clonal architecture in MM is key to inform successful sequencing strategies.

Methods: We applied serial whole genome sequencing (WGS, 100x coverage, 2x151 bp, NovaSeq instrument) and immunohistochemistry (IHC) studies in a patient with RRMM who was treated with 10 lines of therapy with literally all available drugs including BCMA and GPRC5D targeting agents.

Results: WGS revealed several illustrating genomic events: First, a homozygous BCMA deletion conferring antigen loss after treatment with a BCMA targeting T-cell redirecting bispecific antibody. This biallelic hit was characterized by monosomy 16 and a small deletion on the remaining allele of 586 kb. IHC confirmed BCMA loss on the protein level. Second, 2.5 years later, the patient was treated with a GPRC5D targeting bispecific antibody and again achieved deep remission lasting for 15-months. At subsequent relapse WGS displayed an acquired 12p deletion of 23.3 Mb including GPRC5D and IHC confirmed GPRC5D loss to underlie relapse. The homozygous BCMA deletion was persistent at this time point, highlighting two sequential evolutionary bottlenecks. Next to the loss of immunotargets, two additional genetic lesions associated to the therapy applied were tracked in this patient: the SBS-MM1 signature, in 12.8% of mutations, which is linked to melphalan exposure, and a clonal truncating mutation in CRBN (p.Gln8Ter), inducing IMiD resistance, that was developed upon 3rd line treatment.

Conclusions: Immunotherapy is an evolutionary bottleneck in the treatment of MM, selecting for antigen loss variants. Sequential mono-immunotherapies may enforce such selection, arguing for multi-specific strategies to circumvent antigen loss. Tumor-intrinsic mechanisms largely explain resistance in our exemplary patient, highlighting WGS of purified MM cells as a promising diagnostic tool to follow-up patients in the era of novel immunotherapies., In our patient the genome could be read as a book in which prior therapies had been noted.