PPM1D Mutations Are Rare in De Novo and Therapy-Related Acute Myeloid Leukemia

Background: Clonal hematopoiesis of indeterminate potential (CHIP) is an age-related condition characterized by somatic mutations in peripheral blood mononuclear cells (PBMC) of otherwise healthy adults that has been associated with increased risk of developing hematological malignancies. Clonal hematopoiesis has been shown to be present in patients with therapy-related myeloid neoplasms (therapy-related acute myeloid leukemia, t-AML) / therapy-related myelodysplastic syndrome, t-MDS) at the time of their primary cancer diagnosis and before exposure to treatment. Such clones expand under selective pressure from cytotoxic treatment for the primary cancer and can subsequently give rise to overt myeloid neoplasms. Somatic mutations in the gene encoding the TP53-inducible protein phosphatase Mg²⁺/Mn²⁺ 1D (PPM1D) were initially reported in PBMC of patients with solid tumors (breast, ovary, lung) and lymphoma. They are associated with older age and seem to reflect prior exposure to cytotoxic treatment. Moreover, the mutations, all of which are nonsense or frameshift mutations in exon 6, have been described as one of the most recurrent mutations in CHIP and to be frequent in t-MDS.

Methods and results: To resolve this issue and to determine whether clones harboring PPM1D mutations that expand into CHIP after cytotoxic therapy for solid tumors drive leukemogenesis and might be useful as markers to identify patients at risk for t-MDS/t-AML development, we performed PPM1D mutational analysis in 87 patients with de novo AML and in 49 patients with t-AML. As mutations in TP53 as a representative DNA damage response gene are rare in de novo AML, we enriched our de novo AML patient cohort towards poor risk cases with complex karyotypes in order to increase the chance of identifying PPM1D mutations. Patients with core-binding factor AML were excluded from the analysis. Using the published frequency of 15% PPM1D mutations in t-MDS (Lindsley RC et al., N Engl J Med 2017;376(6):536-547) as surrogate for the expected frequency in t-AML, a minimum of 44 t-AML patients was determined to be required to allow for the detection of mutations of PPM1D in t-AML (Chi square with Fisher's exact test for independent groups, α-error 0.05, power 0.8).

We performed focused mutational analysis by targeted Sanger sequencing of PPM1D exon 6 on DNA from bone marrow mononuclear cells or PBMC at diagnosis of de novo or t-AML samples taken prior to treatment initiation. Overall, only one patient with de novo AML (1/87, 1.2%) proved mutation positive. He was diagnosed with AML, FAB M4, at the age of 57 years and harbored a complex karyotype with marker chromosomes in the absence of a TP53 mutation. Unexpectedly, none of the 49 patients with t-AML harbored a mutation in PPM1D.

Conclusion: In this study, we found that PPM1D mutations, which frequently occur in CHIP especially following prior cytotoxic therapy, are uncommon in AML, whether de novo or after prior cytotoxic therapy. These data are in contrast to previous observations on a high frequency of PPM1D mutations in t-MDS samples relative to primary MDS (15% vs. 3%, Lindsley RC et al., N Engl J Med 2017;376(6):536-547). Our findings suggest that while cytotoxic therapy favors the expansion of PPM1D-mutant CHIP clones, possibly even up to the development of t-MDS, mutations in PPM1D seem to be irrelevant for progression to t-AML.