Comparison of Mutation Patterns Between Diagnosis and Relapse in 444 Patients with Acute Myeloid Leukemia Shows High Variability of Stability and Influence On Time to Relapse

In acute myeloid leukemia (AML) molecular mutations are becoming increasingly important as markers for classification, risk stratification and disease monitoring. Frequencies and prognostic impact of most of the currently known mutations have been widely studied during the last years. In contrast, the stability during disease evolution and the role of single markers at relapse is less clear.

To evaluate the pattern of molecular mutations at relapse and their stability during disease progression.

We investigated paired diagnostic and relapse samples in a cohort of 444 adult AML cases (386 de novo, 32 t-AML, 26 s-AML). The cohort was composed of 211 females and 233 males; median age: 62.9 years (range: 18.6–85.2 years). All cases were intensively studied for a wide range of mutations (median: at diagnosis 12, median at relapse 2). The mutations were subdivided into two classes: 1) Disease defining markers such as fusion genes (PML-RARA, RUNX1-RUNX1T1, CBFB-MYH11, DEK-CAN, MLL-translocations, NUP98-fusion genes) and distinct mutations (NPM1, CEBPA, MLL-PTD, and RUNX1). 2) Accompanying mutations (acc mut): FLT3-ITD, FLT3-TKD, NRAS, KRAS, KIT, IDH1, IDH2, ASXL1, DNMT3A, TET2, WT1. Mutations were analyzed by amplicon deep-sequencing, direct Sanger sequencing, conventional PCR, and melting curve analyses. In addition, chromosome banding analysis data was available in all samples.

498 relapses were detected in the 444 patients. At diagnosis the subtypes according to cytogenetics or molecular disease defining markers were as follows: PML-RARA (n=9), RUNX1-RUNX1T1 (n=18), CBFB-MYH11 (n=17), MLL-translocations (n=22), other translocations (n=26), complex karyotype (n=28), MLL-PTD (n=28), NPM1 (n=172), CEBPA (n=22), RUNX1 (n=58), no such marker (n=44). Numbers of mutations detected at diagnosis in median were 2 (range: 0–7). All cases retained their disease defining marker at relapse, thus, these were regarded as stable. In 288/444 (64.9%) cases at least one acc mut was detected (total: 410, median: 1, max: 5). In detail:, FLT3-ITD (n=113), FLT3-TKD (n=40), IDH1 (n=23), IDH2 (n=32), NRAS (n=41), KRAS (n=5), RUNX1 (n=33), and TP53 (n=13), ASXL1 (n=25), TET2 (n=34), DNMT3A (n=29), WT1 (n=14), others (n=8). In contrast to the disease defining mutations that demonstrated 100% stability the acc mut showed much lower stability when the diagnostic/relapse samples were compared. In the 498 available relapse samples 295 (59.2%) informative patterns of acc mut from diagnosis were analyzed. In 161/295 (54.6%) the diagnostic pattern was stable at relapse, in 67 (22.7%) a loss of at least one marker was observed. Further, 28 cases (9.5%) showed gain of at least one additional acc mutation. In 34 cases (11.5%) a shift was observed (complete loss of the diagnostic and gain of a new pattern of acc mut at relapse). Of note, 138/160 FLT3-ITD positive cases (86.3%) retained an FLT3-ITD at relapse. However, different patterns with respect to FLT3-ITD load were observed: in 16 pts (10.0%) the load decreased, in 13 pts (8.1%) it was stable, and in 109 cases (68.1%) it increased which is compatible with a gain of a genetic event.

Between the various disease defining mutation groups there were no relevant differences in time to relapse (TTR) except for the MLL-translocations with shorter TTR compared to all others (median: 8.5 vs. 10.3 months (mo), p=0.006) and the complex karyotypes (median: 8.0 vs. 10.3, p=0.028). In contrast, the acc mut were highly related to TTR: In the total cohort FLT3-ITD+ (n=113) was associated with shorter TTR (7.4 vs. 10.8 months (mo) in FLT3-ITD- cases (n=326), p<0.001) as well as in TET2 mut (8.2 vs. 13.1 mo in TET2 wt). In contrast, mutations in IDH1 or IDH2 (IDH1/2, n=57) were associated with longer TTR (14.5 vs 10.0 mo in IDH1/2 wt (n=150), p<0.001). The same effects of acc mut were observed in subgroup analyses encompassing disease defining abnormalities. e.g. in NPM1 mut and RUNX1 mut.

1) Genetic alterations in AML can be subdivided in stable and unstable markers. Whereas disease defining markers are highly stable between diagnosis and relapse the so-called accompanying mutations showed an unstable pattern in 45.6% of relapses. This instability is even higher (62.7%) if the increase of FLT3-ITD load was taken into account. 2) In addition, the influence of acc mut on the time to relapse is more pronounced than that of the disease defining mutations.

Schnittger:MLL Munich Leukemia Laboratory: Equity Ownership. Alpermann:MLL Munich Leukemia Laboratory: Employment. Nadarajah:MLL Munich Leukemia Laboratory: Employment. Grossmann:MLL Munich Leukemia Laboratory: Employment. Eder:MLL Munich Leukemia Laboratory: Employment. Kohlmann:MLL Munich Leukemia Laboratory: Employment. Fasan:MLL Munich Leukemia Laboratory: Employment. Dicker:MLL Munich Leukemia Laboratory: Employment. Kern:MLL Munich Leukemia Laboratory: Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Equity Ownership.