High Molecular Risk Mutations (HMR) Drive Progression of CALR-mutated Myeloproliferative Neoplasm (MPN)

Background: CALR mutations usually portend a favorable prognosis in MPN. However, a subset of patients (pts) has a shorter time to progression despite CALR mutations. We hypothesize that genetic features can identify such pts and allow for timely consideration of definitive therapy with transplantation. In this study, we explore the somatic mutation data of CALR-mutated essential thrombocythemia (ET), primary myelofibrosis (PMF), or post-ET myelofibrosis (post-ET MF) pts to identify risk factors for progression and survival.

Methods: We retrospectively screened all MPN pts seen at our institution between 2015-2024 for a canonical CALR mutation. Data was collected via chart review. HMR were defined as variants in ASXL1, EZH2, SRSF2, IDH1/2, U2AF1 Q157, SETBP1, TP53, or KRAS/NRAS genes. Kaplan-Meier (KM) curve was used for survival probabilities and Cox proportional hazard models were used for univariate analysis.

Results: We identified 130 consecutive pts with CALR-mutated MPN. Thirty-four pts (26%) had primary MF (PMF); 96 pts (74%) had ET of whom 38 (29%) developed post-ET MF. There were 53% men and 26% non-White pts. Median age of initial MPN was 53 (range: 14-83) years (yrs). Median follow-up by reverse KM was ­­­8.1 yrs (range: 0-28). Estimated 10-yr overall survival (OS) from initial MPN was: 96% (95% CI 87-99%) for ET, 92% (95% CI 77-97%) for post-ET MF, 63% (95% CI 44-77%) for PMF. Nineteen pts (15%) progressed to leukemia, and 28 pts (22%) died.

We next analyzed the 72 MF pts [34 (47%) PMF, 38 (53%) post-ET MF], all of whom had broader myeloid somatic mutation panel results available. Forty-four (61%) pts had CALR type-1 (or like) mutation, 39 (54%) pts had CALR variable allele frequency (VAF) >30%, 41 (57%) pts had at least one HMR, and 15 (21%) pts had ≥2 HMR. The most frequent HMR was ASXL1 (n=30, 42%).

Estimates of 5-yr OS and leukemia-free survival (LFS) from the time of MF diagnosis (original diagnosis in PMF and MF transformation in post-ET MF) were 71% (95% CI 60-84%) and 69% (95% CI 58-82%), respectively. The presence of HMR was associated with worse OS (HR 3.03, 95% CI 1.03-8.92) and LFS (HR 3.92, 95% CI 1.35-11.36). Higher number of HMR was associated with worse OS (No HMR: reference; 1 HMR: HR 2.44, 95% CI 0.76-7.85; ≥2 HMR: HR 3.96, 95%CI 1.24-12.66) and worse LFS (No HMR: reference; 1 HMR: HR 2.97, 95% CI 0.95-9.26; ≥2 HMR: HR 5.8, 95%CI 1.85-18.24). Specifically, TP53 (n=9, 13%) or U2AF1 mutations (n=4, 6%) were associated with worse OS (HR 3.72, 95% CI 1.59-8.67 and HR 2.88, 95% CI 0.65-12.78) and LFS (HR 3.24, 95% CI 1.42-7.43 and HR 5.53, 95% CI 1.52-20.09), respectively. ASXL1 mutation appeared to trend towards worse LFS (HR 1.81, 95% CI 0.83-3.96). OS and LFS did not significantly differ by CALR type, VAF >/< 30%, or PMF vs. post-ET MF.

There were 19 pts who progressed to leukemia at a median of 3 yrs (range: 0-14) from MF diagnosis. Eighteen (95%) of these patients had HMR while one had RUNX1 mutation. Of note, all 9 pts with TP53, 2/4 (50%) pts with U2AF1 Q157, and 12/30 (40%) pts with ASXL1 progressed to leukemia. Three of these underwent an allogeneic blood or marrow transplant (BMT) and all 3 are alive at 2 years, 4 years, and 5 years from BMT respectively. Median OS for patients who did not undergo BMT for leukemia transformation was 5.9 months from the time of progression (95% CI 2.7-27.3 months).

Conclusions: Our data supports the hypothesis that the presence of HMR (especially TP53 and U2AF1 Q157) or ≥2 HMR can drive a poor prognosis despite the presence of otherwise favorable CALR driver mutation. Interestingly, we did not see a difference by the type of CALR mutation in our cohort which is different from previously published data (Tefferi et al. Leukemia 2014). This study reflects on importance of full next-generation sequencing testing and timely consideration of BMT if HMR is present in pts with CALR-mutated MPNs.