Therapy-Related Myeloid Neoplasms Following Autologous Stem Cell Transplantation: The Prevalence of Chip Mutations at Time of Transplantation — a Single Center Experience

Introduction: Therapy-related myeloid neoplasms (tMN) are high-risk conditions evolved after exposure to a number of agents, including cytotoxic therapy, and include myelodysplastic syndrome (MDS), acute myeloid leukemia (AML) and myeloproliferative neoplasms (MPN). As such cytoreduction as part of autologous stem cell transplantation (ASCT) increases the risk of developing tMN. Importantly, both the use of ASCT and the incidence of tMN are rising. The recent characterization of clonal hematopoiesis of indeterminate potential (CHIP) has, in preliminary reports, been shown to increase the risk of developing de novo hematological disease as well as tMN. We hypothesized that patients with non-myeloid primary disease who develop tMN after ASCT, had detectable myeloid mutations at time of transplantation, and that these may represent a risk factor in the development of tMN. This study characterizes tMN patients previously subjected to ASCT and investigates whether CHIP mutations are present in hematopoietic stem cells at time of ASCT.

Methods: The cohort of this observational study consists of patients treated with ASCT at the Department of Hematology, Aarhus University Hospital (Denmark) from 1989 to 2016. Cases were identified via the Danish Pathology Registry and all tMN diagnoses were verified by the same experienced hemopathologist (GBK). Only cases from patients with non-myeloid primary disease, who were diagnosed with tMN after being treated with ASCT (minimum latency being 90 days) were included. 36 cases with available leukapheresis products were identified out of a cohort of 1130 patients.

Samples collected from leukapheresis prior to ASCT were subjected to targeted next-generation sequencing (NGS), using the commercially available panel "Myeloid Solution" (Sophia Genetics, Saint Sulpice, Switzerland), covering 30 genes relevant for myeloid neoplasms, joined with a bioinformatics pipeline from Sophia Genetics. Samples from 31 patients have been subjected to NGS. Variants residing in or within ± 25 nucleotides of coding exons and with coverage >5000 at the variant site were reported. Indels present in polynucleotide stretches were excluded.

Data for continuous variables age, latency to tMN and survival were analyzed as one sample from a normal distribution based on the Students t-test. Normality was assessed via Q-Q plot. Estimates are reported with a 95% confidence intervals (Stata, version 15.1, StataCorp LLC, TX, USA).

Results: The cohort consisted of 25 males (80.7%) and 6 females (19.3%), with an estimated median age at ASCT of 58 years (CI 95% 54;63). Estimated median time to tMN was 3.7 years (CI 95% 2.5;5.4) and estimated median survival after tMN diagnosis was 132 days (CI 95% 71;246). At time of tMN diagnosis, 14 patients had a poor risk karyotype and 9 patients had intermediate risk karyotype (hereof 4 normal karyotype). Karyotype was not evaluated in 8 patients.

CHIP mutations were detected in stem cell enriched leukapheresis products from 21 patients (67.7%). Of these, we found multiple mutations in 14 patients (66%) and in one patient as many as 6 CHIP mutations. Point mutations were frequently found in the DNMT3A gene and was present in 16 out of 21 patients (76%). Six patients had more than one DNMT3A mutation, one of which had 5 separate DNMT3A mutations. Other mutations detected were TP53 (6/21), TET2 (5/21), ASXL1 (4/21), EZH2 (1/21), WT1 (1/21), JAK2 (1/21), NRAS (1/21), HRAS (1/21), BRAF (1/21), CSF3R (1/21), SF3B1 (1/21), ZRSR2 (1/21), CALR (1/21), SRSF2 (1/21).

Conclusion: We found that CHIP mutations can be detected at time of ASCT in patients being treated for non-myeloid diseases. We hypothesize that presence of CHIP mutations at ASCT may predict the development of tMN and as such serve as a biomarker in this setting. We speculate that high-dose cytotoxic therapy may provide an evolutionary advantage for hematopoietic clones containing CHIP mutations. On the other hand, we cannot rule out that the cytoreduction administered prior at ASCT may be a main contributor to the tMN development. To address this as well as the development of tMN in the post-ASCT phase, a nested case-control study will be necessary.