Changes in the Mutational Landscape in Acute Myeloid Leukemia Relapse Under Allogeneic Immune Pressure

Acute myeloid leukemia (AML) is a heterogenic disease harboring different cytogenetic abnormalities and mutations. The combination of intensive conditioning chemotherapy and the allogeneic graft with its unique graft-versus-leukemia effect provides selection pressure for surviving leukemia cells.

The present analysis is a single centre retrospective cohort study, analysing the mutational landscape of 30 adult patients (pts) who were diagnosed with AML and underwent allo-HCT at our center between June 2016 and January 2022 and suffered from hematologic relapse until December 2023. For each pt two myeloid panels (Illumina TrueSight® Myeloid NGS Panel with 54 genes) were analyzed: one at primary diagnosis or before allo-HCT and one at the time of relapse. The vast majority of AMLs had an adverse risk profile (76.7%) according to ELN 2022, whereas only 16.7% were intermediate risk and 6.7% favourable risk. We found the distribution of the groups to be well balanced: 23% acquired new mutations in relapse, whereas 26% lost one or more of their known mutations. Also, 26% acquired new mutations, but also lost known mutations and in 26% the mutational landscape did not change upon relapse. NRAS, RUNX, TP53 and DNMT3A were the most common mutations that were present in the relapse situation (n=7 / 6 / 5 / 4 pts, respectively), whereas targetable mutations included FLT3-ITD (n=3; 1 with gain of mutation and 2 with persistence) and IDH1 (n=2; 1 with gain of mutation and 1 with persistence). Furthermore, we found that mutations in NRAS (5 pts), KRAS (3 pts) were most often acquired in relapse, but also RUNX, ETV6, FLT3-ITD, DNMT3A, ASXL1, PTPN-11, TP53, BCOR, IDH1, IDH2, ABL, CBLB and BRAF were acquired in 1 - 2 pts each. Interestingly, there were also mutations that never occurred de-novo in relapse: TET2, NPM1, U2AF, CEBPA, JAK2, FLT3-TKD, SFB1, GATA2, KIT, KMT2A-MLLT3, STAG2, MLL-PTD, CBL, RAD21, SRSF2, PHF-6 and BCORL1. Regarding the loss of existing mutations, we found that TET2 and FLT3-ITD were most often lost in the relapse situation (4 pts and 3 pts, respectively).

Next, we analyzed the progression free survival (PFS) and the survival of pts who had newly acquired mutations (“gain”) during relapse, compared to pts who did not have newly acquired mutations (“no gain”). Pts who had simultaneous loss and gain of different mutations were counted in the “gain” group. We did not observe any differences in terms of PFS following allo-HCT (P=0.8, HR 0.91, 95% CI 0.44 - 1.88). Interestingly, regarding the overall survival (OS) following primary diagnosis or transplantation we observed a survival benefit in pts who did not experience the acquisition of new mutations (primary diagnosis: P= 0.2, HR 1.75, 95% CI 0.73 - 4.19; transplantation: P= 0.28, HR 1.61, 95% CI 0.67 - 3.83), although this did not reach significance. To understand whether early relapse did impact the mutational landscape in our pt cohort, we analyzed pts who suffered from relapse within the first 6 months following transplantation (early relapse; 11 pts) and pts with relapse at a later time point (19 pts; late relapse). In early relapse, we observed that acquisition of new mutations was most frequent (37%, 4/11 pts), whereas loss and acquisition occurred in 18%. In pts with late relapse the mutational landscape was different with only 16% (3/19) experiencing an acquisition of new mutations and 32% (6/19) with newly acquired mutations and loss of known mutations. Regarding the PFS, we observed a trend towards a shorter relapse-free time period when pts acquired new mutations compared to pts who did not acquire new mutations (P=0.27, HR 0.77 95% CI 0.45 - 6.43), suggesting that the occurrence of new mutations might accelerate the dynamic of relapse. However, we did not find these differences in pts who had relapse later than 6 months following allo-HCT (P=0.73, HR 0.68 95% CI 0.27 - 1.73). Regarding the OS of the pts, pts with early relapse had a much shorter survival probability compared to pts with relapse at later time points, again highlighting the dynamic of the disease.

In summary, we found a broad range of mutational changes in the relapse situation compared to primary diagnosis. Interestingly, gain of new mutations was accompanied with a trend towards a worse survival. These data warrant further investigations in a larger cohort. Moreover, our findings highlight the importance of repeated testing of pts with relapse using next generation sequencing.

Wäsch:Janssen, Sanofi: Research Funding; Amgen,BMS/Celgene, Janssen, Kite/Gilead, Novartis, Pfier, Sanofi: Consultancy; Abbvie,Amgen, BMS/Celgene, Janssen, Kite/Gilead, Pfizer, Sanofi: Honoraria. Wehr:Takeda: Honoraria, Other: travel grant; MSD: Honoraria, Other: travel grant; Jazz: Honoraria, Other: travel grant. Zeiser:Incyte: Consultancy, Honoraria; Medac: Honoraria; Neovii: Consultancy; Sanofi: Honoraria; Ironwood Pharmaceuticals, Inc.: Consultancy; Novartis: Consultancy, Honoraria; Mallinkrodt: Consultancy, Honoraria.