Secondary Hematologic Malignancies after Autologous Stem Cell Transplantation for Multiple Myeloma Are Associated with a Distinct Mutational Profile

Background: Patients (pts) diagnosed with multiple myeloma (MM) are at risk for developing secondary hematologic malignancies (SHM). The etiology of SHM-associated genetic alterations (GAs) is unclear. We hypothesized that the GAs present in MM-associated SHMs would have a distinct profile compared to de novo or other therapy related hematologic malignancies.

Aims: For MM pts who develop SHM, compare GAs at Autologous Stem Cell Transplant (ASCT) and at diagnosis of SHM. Assess for presence of previously reported deleterious myeloid GAs and determine if there is clonal evolution from the autograft.

Methods: We retrospectively identified 9 MM pts with SHM post-ASCT. Autograft (auto) cells and SHM Fresh Frozen Plasma Extraction (FFPE) samples underwent whole exome sequencing. GAs with known clinical significance, variant allele frequency (VAF) ≥0.05 or ≤0.9, and high or moderate impact on the gene-encoded protein were included for analysis. From literature review, we identified 89 reported GAs (kmGAs) in myeloid malignancies. Targeted deep sequencing for these mutations was performed to obtain the VAF in both the auto and SHM FFPE samples.

Results: 9 adult pts (age 56-71) with MDS, AML, or ALL were included. 8 auto samples and 9 SHM samples were available. All pts received induction therapy prior to ASCT (44% and 55% with lenalidomide/thalidomide or bortezomib containing regimens, respectively). Lenalidomide maintenance was utilized in 60% of pts.

Whole exome sequencing revealed 118,614 GAs in all samples. 2074 GAs were included. Average mutational burden was similar between the auto and SHM samples. For paired samples (matched auto and SHM samples for each pt), 1173 GAs with kmGAs of GATA2, SETBP1, and ATM were present. GATA2 and SETBP1 were present in 3 and 5 auto samples, and 4 and 6 SHM samples, respectively. SETBP1 and GATA2 were present in paired samples for 3 and 1 pt, respectively.

Targeted deep sequencing revealed significant mutations in SHM samples, but not auto samples, for ABCA12, ASXL1, BCOR, BRAF, EXH2, KDM5A, KMT2A, KMT2D, NOTCH1, PRPF8, TET2, and TP53. The most highly represented mutation was TP53 which was present in 6 pts, followed by KMT2A in 3 pts, KMT2D in 3 pts, PRPF8 in 2 pts, and TET2 in 2 pts. The patient who carried the most significant mutations carried the diagnosis of ALL, harboring 11 genetic mutations in the SHM sample only.

For paired samples, KDM5A, KMT2D, FLT3, SETBP1, ZRSR2, PRPF8, TET2, and TP53 showed mutations in both auto and SHM samples, showing stable or decreased VAFs. GATA2 showed two moderate impact missense mutations, one in a pt with a VAF of 0.58 in the auto sample and 0.40 in the SHM. The other GATA2 variant appeared as a novel mutation in one pt's SHM sample and was also present in two other pts with VAFs of 0.5 and 0.49 in the auto sample increasing to 0.81 and 0.70 in the SHM sample, respectively.

TP53 showed the highest number of variants. Analysis showed 6 high impact variants with VAFs ranging from 0.05-0.80 and 3 moderate impact variants with VAFs ranging from 0.08-0.89. These mutations were represented in both the auto and SHM samples included. There were several TP53 alterations with the most frequent being structural interaction variants, missense variants, and frameshift variants.

Conclusion: This limited cohort demonstrates that mutational profile for pts with SHM is distinct from de-novo myeloid malignancies, and the average mutational burden did not change from pre-transplant to the development of SHM. In this population, TP53, KMT2A, GATA2, and KMT2D represented the most frequent SHM mutations. Targeted deep sequencing revealed that most significant variants were present only in SHM samples suggesting a novel mutation rather than clonal evolution from the auto sample.