Taha Alrifai, Natasha Edwin, Dale L. Bixby, Adam John Hockenberry, Emmanuel Okeke, Brett Mahon, Duane C Hassane, Sari Heitner Enschede, Melissa Larson, Jamile M. Shammo
Background: Myelodysplastic syndromes (MDS) represent a heterogeneous group of malignancies that are characterized by aberrant hematopoiesis and consequently abnormal blood counts. Patients with MDS have an increased likelihood of developing acute myeloid leukemia (AML). Next generation sequencing (NGS) is currently being used to study the genetic landscapes of diverse cancer types in finer detail than has previously been possible, owing to its ease-of-use and widespread availability. Crucially, the accuracy and broad breadth of genomic coverage provided by NGS approaches may permit the study of mutations that are rare in MDS patients, yet well studied in AML or other cancer types. Based on the fact that much less is known about particular rare somatic mutations in MDS, we decided to investigate 6 genes. IDH1 and IDH2 are targetable in patients with AML. Mutations in JAK2 are found in high proportions in patients with myeloproliferative neoplasms (MPNs) or MDS/MPN overlap neoplasms. Finally, other mutations in genes such as BCOR, BCORL1, and MUTYH also occur in MDS patients with appreciable-but generally low-frequencies such that their overall impact on MDS patients is largely unknown.
Methods: We utilized the Tempus LENS platform to retrospectively characterize 236 MDS patients who underwent genomic-testing with the Tempus|xT assay (DNA-seq of 595-648 genes at 500x coverage; full transcriptome RNA-seq; data accessed on 06/28/2021). The patient population was 28% female, with a median age of 71.6 years. Mutations identified included germline and/or somatic single nucleotide variants (though we only considered somatic variants in this study), insertions/deletions and copy number variations (gains defined as ≥8 copies). For a subset of patients processed at Rush University Medical Center (RUMC), we additionally performed a curated clinical investigation into laboratory variables, prognostic data, treatment history and outcomes data.
Results: Within the Tempus MDS dataset, we observed a range of mutation frequencies. In total, 58 patients had mutations in at least one of either: JAK2, BCOR, IDH2, IDH1, BCORL1, and MUTYH. When comparing the observed frequencies against values that have previously been reported, mutations in these 6 genes occurred with higher frequency in our dataset while most other mutations were in accordance with the relative frequencies reported by other sources (Table 1). We observed the largest deviation from expectation for BCOR. Whereas prior reports list this mutation as occurring with a frequency of <5%, greater than 10% of the records we analyzed contained a mutation in either BCOR (n=24, 10.16%) or BCORL1 (n=6, 2.54%). Four records in the dataset had mutations in both BCOR and BCORL1; a total of 26 records (11%) thus contained mutations in one of these two genes. Of these 26 patients, 50% were female and the median age was 70 years. The most commonly observed co-mutations with BCOR/BCORL1 were-in order of frequency-RUNX1 (50%), ASXL1 (23%), DNMT3A (23%), SF3B1 (19%), and STAG2 (19%). Additional data were derived from detailed clinical investigation into a limited subset of 4 patients who were treated at RUMC and who had mutations in either BCOR or BCORL1. Revised International Prognostic Scoring System (IPSS-R) scores were: high/very high (2) and low (2). Three of the four patients displayed complex karyotypes and three of the four progressed to AML.
Conclusions: We identified a higher frequency of BCOR/BCORL1 mutations among patients with MDS than has been previously reported. At the same time, most other mutation frequencies were in relative accordance with existing reports. We observed complex cytogenetic abnormalities in 75% of patients for which we had extensive clinical information. Our data suggest an association between BCOR/BCORL1 mutations and complex karyotypes, a finding that may establish a prognostic role of BCOR/BCORL1 and have potential therapeutic implications. This hypothesis is, however, subject to further investigation and will require larger sample sizes that are likely to become available with the rapidly increasing utilization of next generation sequencing techniques and availability of high-quality, curated clinical information.
VIEW THE PUBLICATION