Predictors for Response to Ruxolitinib in Real-Life: An Observational Independent Study on 408 Patients with Myelofibrosis

Background: Response to ruxolitinib (RUX), the only JAK1/2 inhibitor commercially available for the treatment of Myelofibrosis (MF) may vary among patients (pts) and is largely unpredictable at therapy start. Therefore, pts' selection is based only on clinical needs.

Aims: To evaluate the impact of pre-treatment clinical/laboratory factors, as well as RUX dose, on response to RUX in a cohort of "real-life" MF pts.

Methods: A multicenter observational study on WHO-defined MF was conducted in 18 Italian Hematology Centers. Data were extracted from a database that included retrospective data on pts treated before January 2015. Subsequently, data were prospectively collected and updated at a 6-month interval.Response to RUX was evaluated according to IWG-MRT criteria.

Results: Between June 2011 and Apr 2016, 408 pts with PMF (54.4%), or postET (27.7%) / post-PV (17.9%) were treated with RUX.

At RUX start, baseline characteristics were (median): age, 68.5 y (range, 26.5-89); ≥65 y, 63.5%; male, 56.4%; hemoglobin (Hb), 10.7 g/dL (7-16.7); transfusion-dependence 27.9%; PLT, 256×10⁹/L (50-1887); PLT <100×10⁹/L, 9.6%; spleen enlargement, 96.6% (spleen length ≥10cm: 64.2%); total symptoms score (TSS), 20 (12-70).International Prognostic Score System (IPSS) was intermediate (intm)-1 (15.7%), intm-2 (47.3%), high (37%).

Molecular data were available for 332 pts (81.4%) and was positive in 81% (JAK2^V617F), 6.3% (CALR), 1% (MPLW515K/L); 2.7% (triple negative). 30 pts (9%) were JAK2^V617Fnegative but did not receive further molecular evaluation.

Karyotype was abnormal in 55 (26%) out of 210 evaluable pts (unfavorable: 8.1%). Median follow-up from MF diagnosis was 3.8 yr (0.3-29.6) and median RUX exposure was 20 mos (3-56.2).

Overall, 152 out of 365 (42%) pts with spleen ≥5cm achieved a spleen response at any time during RUX therapy. At 3 and 6 mos, the response was achieved by 26.6% and 34.4% of evaluable pts, respectively. In univariate analysis, pre-treatment factors negatively correlating with spleen response were: transfusion dependence, platelet count ≤200x10⁹/l, spleen palpable ≥10 cm below costal margin, grade 3 marrow fibrosis, intm-2/high IPSS risk and interval between MF diagnosis and RUX start ≥2y. Three variables remained significant in multivariate regression logistic analysis: large splenomegaly (HR: 2.05, 95%CI: 1.1-3.7; p=0.02), time-interval ≥2y (HR: 1.78, 95%CI:1.0-3.1; p=0.04) and transfusion dependency (HR: 1.95, 95%CI:1.0-3.7; p=0.04). Spleen response significantly correlated with the average RUX dose in the first 12 wks, with pts treated with doses ≥10 mg BID having better response rates (47.3% vs 26.6% if dose <10 mg BID, HR:2.36, 95%CI:1.3-4.3, p=0.005).

360 pts had a TSS >10 at RUX start and 319 (88.6%) achieved a symptom response. In multivariate analysis, factors associated with worse responses were: transfusion dependency (HR: 3.15, 95%CI 1.5-6.4, p=0.001) and a baseline TSS >20 (HR: 6.7, 95%CI 3.2-13.8, p<0.001). RUX titrated dose <10 mg BID during the first 12-wks of therapy negatively correlated with symptoms response (HR: 2.6, 95%CI 1.05-6.7, p=0.037).

Drug-related anemia (acquisition of transfusion dependency or Hb <10g/dl in pts with a previous Hb≥10) was observed in 127/291 (43.6%) evaluable pts. The probability to develop anemia was significantly higher in females (HR: 1.63, 95%CI 1.03-2.57, p=0.036). Notably, anemia was not influenced by RUX 12-wks titrated dose (41.8% in pts with RUX titrated dose <10 mg BID vs41.5% with higher doses).

80 (19.6%) pts discontinued RUX because of: lack/loss of response (28.8%); drug-related toxicity (27.5%, specifically: thrombocytopenia, 16.2%; infection, 6.3%; anemia, 5%); disease progression with/without acute evolution (8.8%); death (13.8%); allogeneic transplant (8.8%); 2^ndneoplasia (3.8%); other unrelated causes (8.5%).

Summary/Conclusion: In a real-life setting, IWG-MRT-defined spleen and symptoms response rates were observed in 42% and 88.6% of evaluable pts, respectively. Disease severity (in terms of transfusion dependency and large splenomegaly) and a delay in RUX start ≥2yr from diagnosis identified pts with lower spleen response rates. Titrated doses <10mg BID significantly correlated with poorer spleen and symptoms responses. Overall, these data point out the importance of an early treatment and of an effective (≥ 10 mg BID) titrated dose in order to achieve better therapeutic results.