MPN with multiple driver clones show distinct patterns of therapy response and progression - “MPN double driver”

Abstract

Abstract Myeloproliferative neoplasms (MPN) harboring concurrent mutations in more than one driver gene represent an increasingly recognized yet frequently overlooked subgroup, posing significant diagnostic and clinical challenges. Current guidelines recommend a stepwise diagnostic approach: testing of CALR and MPL only if JAK2 is negative. This sequential strategy inherently risks overlooking cases, as simultaneous mutations in JAK2, CALR, and MPL, although rare, do exist. In published, rather limited datasets, patients typically exhibited older age and elevated platelet counts compared to those with single mutations, with suboptimal responses to conventional therapies necessitating tailored strategies. The true incidence of “double-driver” mutations is likely underestimated. To characterize the molecular landscape and clinical management of “double driver” patients 3 international patient cohorts were investigated: first we analyzed 84,921 samples with suspected diagnosis of MPN. 29,372 individuals (34.59%) were positive for at least one MPN driver mutation (JAK2, MPL or CALR). 439/29,372 individuals had at least 2 (1.5%) and 3/29,372 individuals all 3 driver mutations (0.01%) detectable. In 424 evaluable individuals, M:F ratio was 1.09 with a median age of 73.2 years (range 16.9-96). Leukocyte range was 1.5-117.0 Gpt/l and hemoglobin 6.1-22.0 g/dl, indicating a high variability depending on the clinical phenotype. Patients harboring 2 mutations frequently had elevated platelet counts (median 773Gpt/l; range 22-1,956 Gpt/l). Clone size of the respective driver mutations showed a high variability (CALR: mean 22.2%; range 0.85-62%; JAK2 16.9%, range 0.83-96%; MPL 13.5%, range 0.96-88.8%), while >80% of individuals revealed clone sizes <50%. 57% of cases that had received a complete NGS myeloid panel showed additional co-mutations, most frequently in TET2, SRSF2 and ASXL1, while 43% of the analyzed cases had no further mutation detectable. These results were confirmed in a cohort of the German MPN Study Group Registry, an observational study of MPN patients with >70 participating centers (university & community hospitals and office-based hematologists): out of 4,148 MPN patients with detailed clinical annotation, 54/4,148 (1.28%) were identified with 2 and 6/4,148 (0.05%) with all 3 driver mutations. Analysis of individually selected patient cells (with 2 confirmed drivers) in colony formation assays revealed their bi-clonal nature. The type of driver mutation combination had impact on the clinical phenotype with JAK2/MPL being predominantly diagnosed as ET (45%) and MF (40%), JAK2/CALR-co-mutated cases as ET (53%) and MF (22%), CALR/MPL-mutated cases as ET (60%) and MF (40%) and triple mutated cases as PV (50%) or ET (50%). Overall, patients with 2 drivers showed a higher rate of progression (to myelofibrosis or MPN-blast phase) compared to those with 1 driver mutation (16.7% vs. 9.9%). Of note, JAK2/CALR co-mutated PMF patients showed better overall survival (median OS: JAK2/CALR not reached; JAK2: 5.4 years, CALR: not reached), while JAK2/MPL co-mutation resulted in dismal survival (median OS: JAK2/MPL: 3.8 years; JAK2 5.4 years; MPL 3.2 years). Consistently, clonal competition assays conducted with murine hematopoietic cells over 8 days revealed clonal dominance of JAK2 over CALR (day 8; p<0.0046**) while MPL-mutated cells were dominant over JAK2-clones (day 8; p=0.0115*). Transcriptome profiling of competing cell populations identified distinct inflammatory signatures depending on the driver combination that can be traced to subclonal compartments in murine competition assays in vivo and single cell sequencing. To assess for therapy responses and clinical characteristics in more detail, we investigated 56 “double driver” patients with detailed clinical annotation from 12 international MPN centers and found differential clinical and molecular responses in patients treated with Jak-inhibitor or interferon-alpha therapies, e.g. partial and complete molecular responses of MPL-clones upon IFN-containing regimens. Our data provides insight into a rare, so far overlooked patient population and indicates a need to modify the sequential diagnostic approach. Appearance of MPL and CALR mutations parallel to JAK2 impact on prognosis irrespective of clone size. Differential patterns of therapy response (e.g. to IFN) appear to be relevant for molecular responses and survival.