Targeted deep sequencing of myeloma in a multi-institutional clinical workflow

Abstract

Abstract Background: Fluorescence in situ hybridization (FISH) remains standard for multiple myeloma (MM) but is limited by probe scope and cell number requirements. Growing evidence and recent IMS/IMWG guidelines underscore the need for sequencing-based assays for comprehensive genomic profiling. PlasmaSeq is a next-generation sequencing (NGS) platform designed to detect copy number alterations, structural variants, and somatic mutations relevant for risk stratification and immunotherapeutic targeting. We evaluated the clinical performance of PlasmaSeq versus FISH in bone marrow (BM) samples from routine diagnostic workflows. Methods 155 BM aspirate samples from MM patients including newly-diagnosed, relapsed and maintenance stages from four academic institutions were analyzed by PlasmaSeq (NCGM, Raleigh NC) in routine clinical testing. One sample was excluded from this analysis due to sample instability. Testing was performed in a central laboratory. CD138 enrichment was performed using a custom optimized method followed by NGS. Sequencing depths varied from 250x – 1000x across all samples. PlasmaSeq integrates copy number, structural, and mutation analyses to determine clonal plasma cell populations and genomic alterations. PlasmaSeq tumor purity (post-CD138 enrichment) was analyzed alongside BM plasma cell content (as available). Cases with PlasmaSeq and FISH, were evaluated for concordance and discordant findings were reviewed. Results DNA input ranged from 1ng/ul to 338ng/ul with the BM plasma cell content ranging from <5% to ~95%, and PlasmaSeq tumor purity values (post-enrichment) ranging from <1% up to >90%, reflecting variable quality and clonality. Clonal plasma cells were detected in 110/154 (71%) samples based on an integrated clonality tool. Of 110 patients, PlasmaSeq identified 45 (41%) Hyperdiploid (HMM), 46(42%) 1q gain, 10(9%) 1pdel , 11(10%) t(4:14), 4(3.6%) t(14:16), 12(11%) t(14:20), 28(25%) t(11:14) and 20(18%) 17p deletion cases. PlasmaSeq identified mutations in MM-relevant genes in 59% (65/110) of samples. RAS mutations—including KRAS, NRAS, and BRAF—were observed in 37% (41/110) of clonal samples, while TP53 mutations were found in 11% (12/110). Variant allele frequencies (VAF) were noted to be distributed across a wide spectrum from 5% to 98%. In 23 of 110 PlasmaSeq-positive cases (21%), FISH was either not performed or unsuccessful; all 23 had clinically significant findings on PlasmaSeq. 87 cases had results from both tests: 57 (66%) were concordant and 30 cases (34%) were discordant, with 25 abnormalities unique to PlasmaSeq. Thirteen cases unique to FISH were ruled out by sequencing, through attribution to off-target FISH probe hybridization based on confirmatory data from PlasmaSeq with 1 patient unexplained for a 1q gain. Of the 25 events detected by PlasmaSeq alone, 13 cases were identified as HMM not tested by FISH, a narrow gain1q event below FISH resolution, t(14;20), t(11;14), t(4;14) translocations not tested by FISH, and a narrow del1p32 missed by FISH, highlighting the sensitivity and specificity of the assay. PlasmaSeq was also able to refine off target changes otherwise not clearly delineated by FISH. For example, a patient with a (11;14) translocation by FISH which proved to be an atypical translocation not involving CCDN1 by PlasmaSeq. Furthermore, 2 samples with low tumor purity (<5% by PlasmaSeq) still reported (14;20) translocations—events not targeted by FISH. Conclusions PlasmaSeq achieved a 100% diagnostic rate in all 110 clonal cases. In 53% (58/110), PlasmaSeq refined FISH results. PlasmaSeq provided actionable in 71% of all real-world BM samples, with varied tumor purity estimates supporting interpretability in low clonal populations. PlasmaSeq showed high concordance with FISH in evaluable samples but revealed additional high-risk variation not captured by FISH, especially in partially tested or low tumor purity samples. The substantial fraction of patients harboring RAS pathway and TP53 mutations also underscores the need for NGS-based profiling. This data also suggests that 250x or deeper sequencing is required to analyze multiclonal plasma cell populations. PlasmaSeq thus provides a quick and reliable sequencing-based method to adequately implement IMS/IMWG recommendations and clearly shows its benefits over traditional FISH based methodologies while providing data for mutations in genes such as TNFRSF17 and GPRC5D in the era of targeted therapy in myeloma.