Pathogenic germ line variants in a prospective multicenter cohort of patients with multiple myeloma

TO THE EDITOR:

Multiple myeloma (MM) is a disease of clonal plasma cells and is the second most common hematologic malignancy.¹ Germ line susceptibility to cancers has primarily focused on solid tumors.^2-4 However, germ line genetic alterations may play a role in the development of MM.^5-8 Previous studies have primarily focused on well-described somatic genetic aberrations in MM.⁹ Identification of germ line predisposition in MM would have important implications for disease monitoring, familial testing for risk prognostication, and possibly treatment decisions. Here, we sought to examine the prevalence of pathogenic germ line variants (PGVs) in patients with MM at our own institution using a broad next-generation multigene sequencing (NGS) panel.¹⁰ 

We prospectively evaluated patients with MM at various stages of disease (newly diagnosed, relapsed/refractory) across the Mayo Clinic enterprise (Mayo Clinic Arizona, Mayo Clinic Rochester, Mayo Clinic Florida) between 1 April 2018, and 31 March 2020. Revised International Staging System (R-ISS), translocations, and copy number alterations (CNAs) were garnered from the electronic health record. High-risk translocations or CNAs were defined as t(4;14), t(14;16), t(14;20), and del(17p) . Patients underwent germ line testing using an 84-gene NGS platform (supplemental Table 1).¹⁰ Other patient selection criteria and sequencing methods were described previously.⁸ 

PGVs were classified as high (relative risk [RR], >4), intermediate (RR, 2-4), or low (RR, <2) penetrance or recessive allele. The diagnostic testing assay is not Clinical Laboratory Improvement Amendments– validated to assess, or return, data on variant allele frequency (VAF) on clinical reports. Internal approximations in the course of variant interpretation protocol provide some context to those reports where relevant. For example, variants with a VAF from ∼9% to ∼35% are reported as “possible mosaic,” variants above ∼35% are reported as heterozygous, and variants under ∼9% are not called or reported. Demographic and clinical characteristics are presented as descriptive statistics. The prevalence of various PGVs and variants of unknown significance (VUSs) are reported. We report the average follow-up time as a measure of survival given that the cohort size is too small for a formal estimate of overall survival by group. Categorical variables in the full data set were compared using the Pearson χ² test. Continuous variables were compared using 1-way analysis of variance. A P value <.05 was considered to be statistically significant. All statistical tests were 2 sided.

We evaluated PGVs included on the 84-gene NGS platform above from peripheral blood samples from the CoMMpass cohort (www.ClinicalTrials.gov identifier: NCT01454297). We filtered for genes that had a moderate or high Clinical Annotation of Variants impact with a VAF of 40% to 60%.¹¹ This study was approved by the Mayo Clinic Institutional Review Board, and all human participants were given a written informed consent.

A total of 74 patients with MM were evaluated in this cohort. Refer to Table 1 for patient characteristics. When evaluating the cohort as a whole (regardless of PGV status), most patients had no or unknown reported history of hematologic malignancy (72.7%) but approximately half of patients (50.7%) had a family history of cancer. Of the 50 patients with International Myeloma Working Group R-ISS stage available, most had stage II (58%), and a minority had a high-risk translocation or CNA (20%). Of the 74 patients, 5 (6.8%) had a PGV, primarily in DNA damage repair genes (Figure 1). Overall, we found a total of 6 PGVs (1 patient with 2 PGVs, 4 patients with 1 PGV). In particular, we found 4 high penetrance mutations, including 2 PALB2 and 2 TP53 mutations, and 2 moderate penetrance PGVs, 1 in ATM and 1 in CHEK2 (Figure 1; supplemental Table 2).

We found no meaningful differences in age, dichotomized race, R-ISS stage, high-risk translocation or CNA, or treatment when comparing rates of patients with PGV, VUS, or negative sequencing results (Table 1). The patient with 2 PGVs had both variants in TP53. This patient had a 1q gain on fluorescence in situ hybridization testing results and no reported family history of cancer. Of note, we were not able to determine whether the TP53 variants in this patient were on the same chromosome. Both patients with PALB2 PGVs did not have any high-risk alterations on fluorescence in situ hybridization testing. One patient with PALB2 PGV had a strong family history of malignancy (esophageal cancer, MM, and other blood related cancers) whereas the other patient with a PALB2 PGV had no family history of malignancy. The patient with a CHEK2 PGV had high-risk CNAs including del(17p) and t(4;14). The patient with an ATM PGV had a family history of solid tumor malignancy (colon, breast, and renal cancers) and monosomy 13. Median follow-up time (Q1, Q3) showed some differences between groups with 37.4 months (34.9, 43.7) in the PGV group vs 42.3 months (30.2, 68.6) in the VUS group and 44.4 months (31.7, 54.9) in the groups with negative sequencing results.

The genes in which PGVs were identified have clinical implications for the patients including screening for solid tumor cancers, cascade family variant testing, and other published management recommendations. Of the 5 patients with PGVs, all had PGVs (inclusive of moderate and high penetrance genes) with clinical actionability such as eligibility for screening to detect solid tumor malignancy. This is important given the observation that up to 11% of cancer patients with PGVs have them discovered only after they have developed a second, preventable primary malignancy.^11,12 They are also indicated for other published clinical management recommendations based on their detected PGVs.^13,14 One such recommendation is cascade family variant testing, which is crucial for preventive surveillance and early detection of cancer in those identified to be at high risk. However, only 1 of the 5 patients with a PGV had family members who chose to participate in the cascade testing.

In this multisite study of patients with MM at various stages of disease, we detected PGVs at a prevalence of 6.8%. Most germ line variants we identified were in DNA damage repair genes and all have well-defined cancer risks and screening guidelines. Interestingly, previous retrospective reports evaluating PGVs in 895 newly diagnosed MM from CoMMpass and a larger cohort of 1161 patients with MM confirmed a similar prevalence to our small cohort (9.9%).^15,16 They also found enrichment in DNA damage repair genes. For validation of our findings, we also evaluated PGVs in genes from our NGS platform in patients from the CoMMpass cohort and found 29 variants with 48 total mutations in 902 peripheral blood samples (3.2% and 5.3%, respectively), which is a similar prevalence (5.3%) to what we observed in our cohort. Median follow-up time was also not significantly different between PGV-positive and PGV-negative patients in the CoMMpass cohort (167.0 vs 118.5 months, respectively; P = .109; supplemental Figure 1).

Here, we show that multigene panel testing can effectively detect PGVs in a substantial group of patients with MM that can have implications for future cancer risk screening both for the patient and their family members. A weakness of our study is our relatively small cohort size. This sequencing panel was also designed for solid tumor hereditary cancer syndromes and did not necessarily contain genes implicated in familial MM.⁹ Although we did not find an association among family history of hematologic malignancy, translocations, and CNAs, this is likely caused in part by the size of the cohort. Larger prospective studies should be conducted on patients with MM at different stages of disease to assess the prevalence of germ line alterations and association with disease characteristics, prognosis, and implications for future cancer screening. It is also important to note that although we cannot definitively report the VAF for the variants identified, analyses per our protocols suggest that there is a low likelihood that any of the pathogenic or likely pathogenic variants from these cases represent examples of clonal hematopoiesis of indeterminate potential; all of these variants would have been observed at a VAF of ∼35% or greater.

Acknowledgments: This study was supported by a Mayo Practice Transformation grant, Mayo Clinic Center for Individualized Medicine, Desert Mountain Members’ CARE Foundation, David and Twila Woods Foundation, and a Faculty Career Development Award from the Gerstner Foundation (N.J.S.).

Contribution: J.E.W.-N., K.L.K., M.A.G., and N.J.S. had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis; N.J.S., A.K.S., L.B., R.F., and K.L.K. were responsible for concept and design; J.E.W.-N., K.L.K., M.A.G., N.J.S., E.D.E., R.L.N., E.J., and B.H. helped with acquisition, analysis, and/or interpretation of the data; J.E.W.-N. drafted the manuscript; K.L.K. and M.A.G. performed statistical analysis; and all authors were responsible for final content of the manuscript.

Conflict-of-interest disclosure: E.D.E. is currently employed and holds stock options in an appropriately traded company, Invitae; and is a current holder of stock options in a privately held company and has membership on entities board of directors or advisory committees of Taproot Health. R.L.N. is currently employed and holds stock options in a publicly traded company, Invitae; is a consultant and a current holder of stock options in a privately held companies, Genome Medical and Maze Therapeutics; has served as a consultant for Pfizer. B.H. has current employment and a current holder of stock options in a publicly traded company, Invitae. A.K.S. has served as a consultant for GlaxoSmithKline (GSK), Sanofi, Janssen, Amgen, and Tempus Health; and has a stock ownership (not including stocks phoned in a managed portfolio) in Tempus Health. N.J.S. has served as a consultant for Janssen Research, Cancer Prevention Pharma, and Recursion Pharmaceutical. L.B. is a consultant for Oncopeptides, Janssen, GSK, Novartis, and Pfizer. R.F. has served as a consultant for Juna, Karyopharm, Bayer, Sanofi, Regeneron, Pharmacyclics, Pfizer, ONCOtracker, Oncopeptides, Novartis, Merck, Kite, Takeda, Janssen, H3 Therapeutics, GSK, Bristol Myers Squibb/Celgene, Amgen, and AbbVie; and has membership on entities board of directors or advisory committees of ONCOtracker, Adaptive Biotechnologies, Caris Life Sciences, and OncoMyx. The remaining authors declare no competing financial interests.

Correspondence: N. Jewel Samadder, Division of Gastroenterology and Hepatology, Mayo Clinic Arizona, 5777 E Mayo Blvd, Phoenix, AZ 85054; email: samadder.jewel@mayo.edu.