Clinical Features and Preliminary Investigation of PPM1D-Mutated Myeloproliferative Neoplasms

Myeloproliferative neoplasms (MPNs) are characterized by clonal myeloid cell proliferation and include essential thrombocythemia (ET), polycythemia vera (PV), and primary myelofibrosis (PMF). Though mutations in JAK2, CALR, and MPL are the most established drivers of MPN, recent studies have highlighted the role of p53 dysregulation in MPN pathogenesis. Another gene responsible for cell cycle checkpoint responses implicated in various hematologic malignancies is PPM1D. Here, we describe the clinical outcomes of patients with PPM1D-mutant MPNs at a single institution and results of laboratory investigation of the role of PPM1D in the pathogenesis of MPNs.

Data was obtained from clinical next generation sequencing (NGS) performed at our institution between 2018 to 2024. We identified 19 individuals with MPNs with a PPM1D mutation either at diagnosis or on subsequent testing. All patients were white, 58% (n=11) were female, and the median age at diagnosis was 58 years (IQR, 50,72). Median follow up was 117 months (IQR 25, 185). The distribution of diagnoses was 42% (n=8) ET, 26% (n=5) PV, 16.3% (n=3) CMML, and 16.3%(n=3) MPN-NOS/MDS; notably no PMF diagnoses were present. Of the 19 individuals 42% (n=8) were not tested for a PPM1D mutation at diagnosis, 37% (n=7) had a PPM1D mutation at diagnosis and, 21% (n=4) acquired one later.

Of the 16 non-CMML cases, 81% (n=13) had a JAK2 mutation and 6% (n=1) had a CALR mutation. 2 patients (1 PV and 1 ET) did not have a classical driver mutation present. 32% (n=6) progressed to myelofibrosis with a median time of 129 months (IQR, 94,177). 37% (n=7) progressed to AML with a median time of 116 months (IQR 56, 177). 74% (n=14) received hydroxyurea initially, 11% (n=2) received HMA containing therapy and 11% (n=2) were observed or treated with phlebotomy. One patient who had CMML-2 with high-risk features was treated with azacitidine and sabatolimab at diagnosis. Of note, many patients initiated treatment prior to the advent of targeted therapies and had favorable responses to hydroxyurea monotherapy for many years, but ultimately 63% (n=12) required a change of initial therapy, 4 due to worsening disease, 2 due to progression to myelofibrosis, 6 due to progression to AML, and 2 due to transitions of care. Overall mortality was 37% (n=7), with 4 dying of AML, 2 of bone marrow transplant complications and 1 of a non-oncologic cause.

To further interrogate the role of PPM1D mutations in MPN pathogenesis, we utilized our mouse model of PPM1D activating mutations in which a c-terminal truncating mutation (T476*) is conditionally introduced via Cre-recombinase into the endogenous Ppm1d locus (reference PMID 37595362). Using the Jak2V^617F-fl/+ mice, in which V617F is conditionally introduced into Jak2 upon Cre expression, we bred Ppm1d^+/+;Jak2^V617F/+;Vav-Cre, Ppm1d^T476*/+;Jak2^+/+;Vav-Cre, and Ppm1d^T476*/+;Jak2^V617F/+;Vav-Cre animals (reference PM1D 20541703). We transplanted bone marrow from these mice into lethally irradiated recipient animals and monitored hematologic engraftment over approximately 12 weeks. As expected, mice that received BM from Ppm1d^T476*/+;Jak2^+/+ showed no significant hematologic abnormalities, whereas those that received BM from Ppm1d^+/+;Jak2^V617F/+ displayed a polycythemia. However, there was no difference in the degree of polycythemia between the Ppm1d^+/+;Jak2^V617F/+ and the Ppm1d^T476*/+;Jak2^V617F/+ groups, which suggests that in these mouse models a truncating Ppm1d mutation does not enhance or alter the Jak2-driven polycythemia phenotype.

Our study presents the largest reported cohort of PPM1D-mutated MPNs to date. Our clinical data suggests that although uncommon in MPN, PPM1D mutations may be associated with more aggressive phenotypes as evidenced by the relatively high rates of transformation to MF and AML. In contrast, an activating Ppm1d mutation did not enhance the polycythemia phenotype of Jak2-driven disease in mouse models. However, these animal models may not fully recapitulate human disease, and we did not interrogate other phenotypes such as transformation. Thus, expanded clinical cohorts to further refine the relationship between PPM1D mutations and MPN disease phenotypes coupled with additional mechanistic studies of PPM1D in MPN pathogenesis are required. These findings underscore the complexity of MPNs and highlight avenues for future exploration in precision medicine for patients with PPM1D mutations.