-Author name in bold denotes the presenting author
-Asterisk * with author name denotes a Non-ASH member
Clinically Relevant Abstract denotes an abstract that is clinically relevant.

PhD Trainee denotes that this is a recommended PHD Trainee Session.

Ticketed Session denotes that this is a ticketed session.

1882 Phylogenetic Analysis Reveals Distinct Evolutionary Patterns in Multiple Myeloma Patient Derived Xenografts

Program: Oral and Poster Abstracts
Session: 651. Multiple Myeloma and Plasma Cell Dyscrasias: Basic and Translational: Poster I
Hematology Disease Topics & Pathways:
Research, Fundamental Science
Saturday, December 7, 2024, 5:30 PM-7:30 PM

Nathan Becker, MS1*, Enze Liu, PhD2*, Parvathi Sudha, MS2*, Riya Sharma, MS2*, Travis Johnson, PhD3, Aneta Mikulasova, PhD4*, Rafat Abonour, MD2 and Brian A. Walker, PhD2

1Melvin and Bren Simon Comprehensive Cancer Center, Division of Hematology and Oncology, Indiana University School of Medicine, Fishers, IN
2Melvin and Bren Simon Comprehensive Cancer Center, Division of Hematology and Oncology, Indiana University School of Medicine, Indianapolis, IN
3Department of Biostatistics and Health Data Science, Indiana University School of Medicine, Indianapolis, IN
4Centre for Cancer and Biosciences Institute, Faculty of Medical Sciences, Newcastle University, Newcastle Upon Tyne, United Kingdom

Introduction: Multiple myeloma (MM) is characterized by significant genomic heterogeneity and clonal evolution which are important mechanisms of therapy resistance and disease progression. Patient-derived xenograft (PDX) models offer a unique opportunity to dissect the dynamic evolution of MM, including the acquisition of mutations, copy number alterations (CNAs), and genomic instability associated with complex structural variations (SVs). This study utilized the SCID-rab PDX model to investigate clonal dynamics and their implications for treatment resistance in MM.

Methods: CD138+ cells isolated from bone marrow aspirates of patients with multiple myeloma (MM) or plasma cell leukemia (PCL) from the Indiana Myeloma Registry (n=13) were engrafted into SCID-rab PDX mice (n=25). Of the patient samples, three were from newly diagnosed patients, nine were from relapsed refractory patients (RRMM), and one was from a plasma cell leukemia patient (PCL). All patients used in this study had been previously treated with 2-7 lines of therapy. An average of 57 million CD138+ cells were isolated per SCID-rab mouse. Paired patient and PDX samples underwent whole-genome sequencing (WGS) and RNA-sequencing as well as comprehensive multiomics. Single-cell WGS was also available for the patient samples. Mutations, CNAs, and SVs were identified by Strelka, AscatNGS, and Manta, respectively. Tumor subclonal architecture and phylogenetic relationships between related PDX samples and primary patients were reconstructed using CONIPHER using mutation and CNA data. CONIPHER uses Dirichlet clustering to identify candidate mutational clusters and nests the clusters based on evolutionary principles for lineage relationships.

Results: Of the 13 patient samples used to generate PDX mice, four were t(4;14), three were t(11;14), four were hyperdiploid, one was hyperhaploid, and one was t(14;16). Phylogenetic reconstruction revealed distinct evolutionary patterns in the SCID-rab model, including linear and branched evolution. A linear pattern of evolution was observed for seven patients while a branched pattern of evolution was observed for three patients. The progressive acquisition of driver mutations was also observed in five cases, where PDX samples acquired additional driver mutations that were not present or were undetectable in the primary patient sample. For example, in one patient sample there were mutations in FGFR3 and NFKBIA, and the PDX derived from that patient gained an additional clonal FGFR3 mutation and a subclonal ARID2 mutation, highlighting the model's ability to reveal selection for genetic markers associated with disease progression.

Patterns of evolution also differed between mice injected with the same patient sample suggesting mouse specific tumor evolution. In one case, a patient sample was injected into three mice, each of which developed a unique subclonal architecture, derived from clones early in the phylogenetic tree, but complete dominance of one clone was not achieved. One key abnormality found in only one of the three PDX samples was gain1q, another key driver of disease progression.

Additional acquisition of novel CNAs were also seen including in one t(11;14) patient sample which was injected into two mice, of which only one PDX gained a copy of chromosome 11 and subsequently acquired further mutations. In addition, complex SVs were observed in 5 cases and on-going evolution of these was observed. A chromothripsis event was observed on chromosome 17 of one patient characterized by clustered rearrangements, however, analysis of breakpoints in the chromothripsis region of the PDX compared to the patient sample revealed additional interactions with chromosomes 6, 7, and 18 supporting the notion that complex SVs can evolve over time and are a mechanism of genomic diversification that could result in disease progression and treatment resistance.

Conclusion: The SCID-rab PDX model is a powerful tool to investigate genomic heterogeneity and the clonal dynamics within patient samples. Differing patterns of evolution, the acquisition of driver mutations, CNAs, and instability associated with complex SVs were observed. Our analysis indicates that clonal evolution and genomic diversification are on-going and can be modeled in PDXs.

Disclosures: No relevant conflicts of interest to declare.

*signifies non-member of ASH