AUTHORS
Taha Iftikhar, MD, Yahan Zhang, MS2, Sandra Mazzoni, DO3, Jack Khouri, MD4, Dongxu Jiang, PhD5, Diana Basali, MD6, Kacey O'Donnell7, Shahzad Raza, MD8, Christy Samaras, DO9, Beth Faiman, PhD, MSN, APN-BC, AOCN®, TCTCN®, FAAN, FAPO10, Faiz Anwer, MD11, Jason Valent, MD12, Babal Kant Jha, PhD13, Louis Samuel Williams, MD14
AFFILIATIONS
1 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
2 Cleveland Clinic Lerner Research Institute, Cleveland, Ohio, United States
3 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
4 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
5 Cleveland Clinic Lerner Research Institute, Cleveland, Ohio, United States
6 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
7 Cleveland Clinic Lerner Research Institute, Cleveland, Ohio, United States
8 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
9 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
10 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
11 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
12 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
13 Cleveland Clinic Lerner Research Institute, Cleveland, Ohio, United States
14 Taussig Cancer Centre, Cleveland Clinic, Cleveland, Ohio, United States
ABSTRACT
Background
AL amyloidosis is the most common subtype of systemic amyloidosis in the US, with dysfunctional proteostasis leading to misfolded light chains without proper activation of degradation pathways. Prior work from our group analyzed a combined 113-patient cohort using RNA-seq-based variant calling, identifying 37 recurrently mutated genes and nominating HERC1 and IGLL5 as candidate AL biomarkers. Most prior AL genomic studies, including this one, lack a contemporaneous comparator across the broader plasma cell dyscrasia spectrum. We addressed this by performing comprehensive DNA-level whole-exome sequencing (WES), both germline and somatic, in a new cohort spanning AL, MGUS, and multiple myeloma (MM), enabling direct within-study comparison and orthogonal validation of prior transcriptomic findings.
Methods
Germline and somatic WES and total RNAseq were performed on 40 patients (AL n=14, MGUS n=7, MM n=15, MGUS+AL n=1, MM+AL n=3; bone marrow derived mononuclear cells. Germline candidates were joint-genotyped, hard-filtered, annotated, and ranked per-cohort across the full exome after excluding ClinVar-benign variants, artifact-prone gene families, and loci with Hardy-Weinberg deviation. Somatic variants were called tumor-only, quality- and rarity-filtered, and cross-referenced against matched germline calls to remove contamination. We then developed a convergent validation framework applied genome-wide: candidates were cross-referenced against a curated, literature-anchored 114-gene panel spanning prior AL and MM sequencing studies, and independently evaluated for concentrated, cohort-specific recurrence across AL, MGUS, and MM.
Results
The median age for AL cohort was 74 (53-82) years with 11 males and 3 females. All patients were biopsy-proven AL, lambda-predominant, with 9/14 AL patients enrolled in a clinical trial of anselamimab. Systematic application of convergent framework across candidate genes in both germline and somatic arms showed MM-restricted patterns: EPHB2 and HIF1A were exclusively present in MM patients, and multiple somatic MM drivers (PIK3CA, IRF4, KMT2A, LTB, RET, PDGFRA, MAX) showed no presence in AL patients. SSH2 was identified as the top-ranked candidate: independently implicated as a recurrently mutated gene. It carried distinct missense variants (p.R896H, p.I472V) in two unrelated AL patients with a fully AL-restricted pattern , zero carriers among MGUS or MM patients across the cohort, after artifact-filtering pipeline and cross-cohort validation. This comprehensive screen also clarified the landscape around previously proposed candidates and confirmed that SSH2's convergent evidence profile was distinct from the broader candidate pool. We additionally observed that ATRIP and KMT2A, despite panel classification as MM-associated genes, were present across AL, MGUS, and MM+AL crossover patients but notably absent from pure MM patients - a hypothesis-generating pattern warranting evaluation in larger cohorts
Conclusion
This study provides a comprehensive, comparative DNA-level analysis spanning AL, MGUS, and MM; literature panel cross-reference paired with systematic cross-cohort recurrence analysis. This framework endorses SSH2 as a robust, AL-specific candidate warranting functional validation, identifies a pattern distinguishing pure MM from the broader AL/MGUS/crossover spectrum meriting further study, and offers a reusable methodology for prioritizing genomic candidates in future AL studies where cohort size limits genome-wide discovery power.
DISCLOSURES
Taha Iftikhar, MD: No relevant conflicts of interest to declare.