AUTHORS
Alessia D'Anna, Postdoctoral Fellow, Meghana Ram2, Junia Vieira dos Santos, MD, PhD3, Alison S. Park, N/A4, Adolfo Aleman, PhD5, Santiago Thibaud, MD6, Hearn Cho, MD, PhD7, Sundar Jagannath, M.D.8, Samir Parekh, MD9, Alessandro Lagana, PhD10
AFFILIATIONS
1 Icahn School of Medicine at Mount Sinai, New York, New York, United States
2 Icahn School of Medicine at Mount Sinai, New York, New York, United States
3 Icahn School of Medicine at Mount Sinai, New York, New York, United States
4 Icahn School of Medicine at Mount Sinai, New York, New York, United States
5 Icahn School of Medicine at Mount Sinai, New York, New York, United States
6 Icahn School of Medicine at Mount Sinai, New York, New York, United States
7 Icahn School of Medicine at Mount Sinai, New York, New York, United States
8 Icahn School of Medicine at Mount Sinai, New York, New York, United States
9 Icahn School of Medicine at Mount Sinai, New York, New York, United States
10 Icahn School of Medicine at Mount Sinai, New York, New York, United States
ABSTRACT
Background: Relapsed/refractory multiple myeloma (RRMM) is shaped by interacting driver lesions, copy-number alterations, and mutational processes, yet rare high-risk events are usually studied in isolation. Prior work in newly diagnosed MM linked 16q deletion with adverse outcome and implicated CYLD and WWOX as candidate tumor suppressors through gene mapping and expression analysis (Jenner et al., Blood 2007). However, whether disruptive CYLD sequence variants carry prognostic significance and whether 16q deletion marks a broader mutational context in RRMM remain unclear. We integrated genomic and outcome data to distinguish gene-level CYLD-associated risk from arm-level 16q biology and define genomic determinants of APOBEC activity.
Methods: We analyzed 349 patients with RRMM using whole-exome sequencing and integrated mutations, copy-number calls, canonical translocations, and clinical outcomes. SBS96 signatures were refit in 292 tumors using a curated reference panel, and APOBEC burden was defined as the combined SBS2 and SBS13 contribution. Broad-arm scans tested associations with log1p APOBEC count after adjustment for TP53 status, observed SBS burden, and global CNA load. Targeted models additionally included MAF/MAFB translocation, hyperdiploidy, 17p deletion, and 1q gain. For 339 samples with matched RNA-seq available, we generated a co-expression network and compared module eigengenes by 16q deletion status using two-sided Wilcoxon tests with BH correction. Sample-matched progression-free survival (PFS) was modeled using Cox regression in 249 tumors with complete clinical and genomic data.
Results: Seven tumors carried predicted high-impact CYLD loss-of-function variants, including nonsense, frameshift, and splice-site variants. PFS was available for four CYLD-altered tumors; all had progression events, with a median PFS of 51 days. CYLD alteration was associated with shorter PFS in univariable analysis (HR=5.62, 95% CI 2.05–15.4, p<10^-3), after adjustment for 16q deletion, hyperdiploidy, TP53 aberration, 17p deletion, and 1q gain (HR=6.52, 95% CI 2.33–18.3, p<10^-3), and after additional adjustment for APOBEC burden (HR=5.49, 95% CI 1.94–15.5, p<0.01). APOBEC burden was itself modestly associated with shorter PFS (HR=1.11, 95% CI 1.01–1.23, p=0.03), whereas 16q deletion alone was not.
We next asked whether arm-level 16q deletion marked a broader mutational-process context. 16q deletion was the strongest positive CNA association with APOBEC burden after adjustment for global CNA load, TP53, and total single-base substitution (SBS) burden (β=0.73, q=0.001). As expected, MAF/MAFB-translocated tumors were APOBEC-enriched. Importantly, 16q deletion remained independently associated with higher APOBEC burden after accounting for this established phenotype and other genomic covariates (β=0.50, q=0.023), whereas hyperdiploidy was independently associated with lower burden (β=−0.47, q=0.023). Among 339 samples with matched RNA data, 16q deletion was associated with 46 co-expression modules (FDR<0.05), 24 of which contained no 16q genes, consistent with trans-associated remodeling beyond direct arm-level dosage. Four of seven CYLD-altered tumors also had 16q deletion; this rare subgroup had a median APOBEC count of 60.9, and two of four were APOBEC-high, suggesting a possible biological connection between CYLD disruption and the broader 16q/APOBEC context.
Conclusions: Building on prior evidence linking 16q loss and CYLD/WWOX dysregulation to adverse outcomes in newly diagnosed MM, our findings extend this biology in RRMM in two directions. Rare disruptive CYLD variants identified an RRMM subgroup with exceptionally short PFS, whereas arm-level 16q deletion defined an APOBEC-enriched genomic state beyond established MAF/MAFB biology. These findings support a mechanistic CYLD/16q/APOBEC axis that may integrate NF-κB dysregulation, mutational processes, and clinical outcome. Orthogonal validation in independent RRMM cohorts and functional models is underway to define this axis as a potential biomarker and therapeutic vulnerability.
DISCLOSURES
Alessia D'Anna, Postdoctoral Fellow: No relevant conflicts of interest to declare.