AUTHORS
Amulya Bhagirath, Suraj Bansal, BHSc2, Caroline Spaner, MD3, Andy Zeng, MD, PHD4, Jessie Medeiros, PhD5, Andrea Arruda6, Aniket Bankar, MD, MSc, DM, MBBS7, Jose-Mario Capo-Chichi, MD, PhD8, Verna Cheung, BScN, RN MN CON(C)9, Marta Davidson, PhD MD10, Jessica Danquah, CCPA11, James A. Kennedy, MD PhD12, Dawn Maze Sr, MD13, Hassan Sibai, MD, Msc, BSc14, Anne Tierens, MD, PhD15, Hubert Tsui, MD, PhD16, Vikas Gupta, MD FRCP, FRCPath17
AFFILIATIONS
1 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
2 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
3 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
4 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
5 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
6 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
7 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
8 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
9 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
10 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
11 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
12 Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
13 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
14 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
15 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
16 Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
17 Princess Margaret Cancer Centre, Toronto, Ontario, Canada
ABSTRACT
Background: Approximately 10-15% of patients with essential thrombocythemia (ET) and polycythemia vera (PV) transform to post-ET myelofibrosis (PET-MF) and post-PV myelofibrosis (PPV-MF), collectively known as secondary MF (SMF). Patients with SMF share clinical, molecular and cytogenetic features with primary MF (PMF). A variety of clinical and molecular integrated prognostic scores have been described for PMF and have been commonly used in clinical trials for SMF as well. MYSEC-PM is a prognostic score using clinical factors developed for SMF, and more recently a molecular and karyotype integrated score prognostic for survival has been described for SMF (Mora B, et al, Blood, 2026). Comparative prognostic performance of commonly used scores for PMF and SMF is not well studied.
Methods: Using Princess Margaret MPN registry (NCT02760238), we identified patients with SMF in chronic phase with available targeted next-generation sequencing (NGS) studies. Patients with concomitant active second cancer were excluded. We evaluated the comparative prognostic performance of scores developed on PMF datasets (DIPSS, DIPSS-plus, MIPSS70, MIPSS70-plus v2.0) with those developed on SMF datasets (MYSEC-PM, MYSEC-mPM, MYSEC-kmPM). The main end points of interest were overall survival (OS) and transformation to blast phase (BP). The index date was first visit to Princess Margaret. All the clinical and genetic data were linked to index date. OS was analyzed from index date to death with censoring at hematopoietic stem cell transplantation (HSCT) or last follow-up using Kaplan-Meier analysis, Cox proportional hazards models, and Harrell C-index. Cumulative risk of blast phase (BP) transformation was analyzed using cumulative incidence analysis with death as a competing event using Fine-Gray regression.
Results: A total of 230 SMF patients with available NGS were included in the analysis, with 132 (57%) PET-MF and 98 (43%) PPV-MF; median age at initial visit was 66 years (IQR, 59-71) and males were 53%. MPN driver mutations were: JAK2 (65%) CALR (28%), MPL (3%) and 4% were triple-negative. Cytogenetic data were available for 167 (73%). The median follow-up was 45 months (IQR, 16-76), during which 101 (44%) patients died, 31 (14%) transformed to BP, and 53 (23%) underwent allogeneic stem cell transplant.
For OS, the highest numerical discrimination was shown by cytogenetic and molecular integrated systems for SMF (MYSEC-kmPM, C-index 0.76; 95% CI, 0.70-0.82) and PMF (MIPSS70-plus v2.0, C-index 0.76; 95% CI, 0.71-0.82). Molecular integrated system for SMF (MYSEC-mPM) performed similarly despite not requiring cytogenetics (C-index 0.75; 95% CI, 0.70-0.81). DIPSS (C-index 0.71; 95% CI, 0.66-0.77), DIPSS-plus (C-index 0.70; 95% CI, 0.64-0.76), and MYSEC-PM (C-index 0.70; 95% CI, 0.65-0.75) showed moderate discrimination, and MIPSS70 had the lowest of the seven models (C-index 0.63; 95% CI, 0.56-0.69).
In Cox models, the high-risk groups by MYSEC-mPM and MYSEC-kmPM showed the strongest associations, with HRs for death as 16.12 (95% CI, 7.82-33.23; p<0.001) and 18.32 (95% CI, 7.66-43.78; p<0.001), respectively.
For stratification of risk of transformation to BP, patterns emerged by model. High-risk by MYSEC-PM showed borderline association with transformation with sHR 3.54 (95% CI, 0.96-13.08; p=0.06), while high risk in MYSEC-mPM and MYSEC-kmPM models showed significant association with sHRs: 3.84 (95% CI, 1.27-11.56; p=0.02) and 4.05 (95% CI, 1.05-15.67; p=0.04), respectively. For DIPSS, DIPSS-plus, and MIPSS70-plus v2.0, there was no significant association with BP transformation. For MIPSS70, the association was significant by Gray’s test, but Fine-Gray estimates were not ordered across risk groups.
Conclusions: We validate the prognostic significance of MYSEC-mPM and MYSEC-kmPM in our dataset, and extend the clinical utility of these scores in predicting risk of transformation to blast phase. Overlapping 95% CIs between best-performing models prevent definitive conclusions on superiority; however, these results suggest disease biology-enhanced models add prognostic value in SMF.
DISCLOSURES
Suraj Bansal, BHSc: No relevant conflicts of interest to declare.