SECOND PRIMARY MALIGNANCY IN MYELOFIBROSIS PATIENTS TREATED WITH RUXOLITINIB

Background: Ruxolitinib (RUX), the first JAK1/JAK2 inhibitor approved for myelofibrosis (MF) therapy, has been associated with the occurrence of second primary malignancies (SPMs), including lymphoid neoplasms and non-melanoma skin cancers (NMSCs). Aims: To investigate incidence, risk factors for SPMs and outcome in RUX-treated MF pts. Methods: We collected 700 MF patients (pts) treated with RUX in 20 European Hematology Centers. Risk factors were identified conducting a time-to-event (SPMs, NMSC and SPMs excluding NMSCs) analysis using the Fine & Gray competing risk regression model with death without SPM as competitive event. Considering SPM as a time-dependent covariate, survival curves were obtained with the Simon and Makuch technique. Results: The characteristics of the MF cohort at RUX start were: median age 68y (24-88); males 58.7%; PMF 53.4%; JAK2, CALR and MPL mutated: 80.3%, 12.7% and 2% of 659 evaluable pts (5% triple negative); prior hydroxyurea (HU) in 430 pts (98.6%), previous neoplasms in 97 pts. Overall, 63.9% and 51.1% of pts had a starting or a cumulative RUX dose > 10 mg BID, respectively; median RUX exposure was 24.6 mos (1-100). 84 pts (12%) developed 92 SPMs after RUX start, with a cumulative incidence of 2.6%, 5.2% and 19.8% at 1, 2 and 5y, respectively. Incidence rate of SPMs was 3.86 per 100 pt-y. NMSCs were the most common SPMs accounting for 44 events in 38 pts (diagnosed in early phase in 85% of pts). Other neoplasms involved urological district (16), lung (11), gastrointestinal tract (5), hematopoietic tissue (1 AML, 1 LNH, 1 CML and 1 Langerhans cell histiocytosis), or others (10). By univariate analysis, male sex (p < 0.001), PLT > 400 X109/L (p = 0.004), Hb ≥ 11 g/dl (p = 0.002) and HU time-exposure > 5y (p = 0.05) were associated with increased incidence of SPMs; in multivariate analysis, male sex, PLT and Hb maintained statistical significance (HR 2.45, CI95% 1.27-4.82, p = 0.01; HR 1.83, CI95% 1.02-3.28, p = 0.04; and HR 2.13, CI95% 1.16-3.98, p = 0.02, respectively). After exclusion of NMSCs, male sex, PLT and Hb confirmed their prognostic value (HR 2.77, CI95% 1.37-5.64 p = 0.005; HR 2.34, CI95% 1.31-4.19 p = 0.004; HR 2.11, CI95% 1.14-3.91, p = 0.02, respectively). Risk factors for NMSC excluding other neoplasms were age ≥ 65 y (p = 0.04), male sex (p = 0.03), secondary MF (p = 0.01), HU time-exposure > 5y (p = 0.004) and previous NMSCs (p = 0.001) in univariate analysis. In multivariate analysis, male sex, HU > 5y and previous NMSCs maintained statistical significance (HR 3.06, CI95% 1.20-7.83 p = 0.02; HR 4.26, CI95% 1.58-11.47, p = 0.004; HR 6.68, CI95% 1.29-34.6, p = 0.02, respectively). Notably, RUX time-exposure ≥1y and starting/cumulative doses > 10 mg BID were not associated with SPMs. Overall, 365 patients (52.1%) discontinued RUX; SPMs diagnosis significantly influenced a higher risk of RUX discontinuation (HR 4.0 95%CI: 2.8-5.7, p < 0.001). In 38 out of 310 deceased pts (12.3%), SPM was the ultimate cause of death, representing the 4th cause after MF progression (25.8%), AML (19%) and infections (13.9%). Patients developing SPMs had indeed an increased risk of death (HR 3.4, 95%CI: 2.4-4.8, p < 0.001) (Fig. 1). Summary/Conclusion: During RUX, the incidence of SPMs increases over time, and has a significant impact on RUX discontinuation and survival. The hallmarks of high oncological risk are male sex, thrombocytosis and prolonged HU exposure, suggesting a closer oncological monitoring in pts with these features. Conversely, RUX time exposure and dose do not influence SPMs and should therefore not be limited for this reason.