Analysis of Predictive Factors for Early Response to Ruxolitinib in 266 Patients with Myelofibrosis from the Polish Adult Leukemia Group (PALG) Registry

Background: Ruxolitinib (RUX), an inhibitor of JAK1/JAK2 kinases, is the only drug approved for the treatment of myelofibrosis (MF). The main clinical effects of RUX therapy include decrease of splenomegaly and reduction of burden of constitutional symptoms. Importantly, predictive factors influencing response and toxicity of RUX are largely unknown. Since 2017 therapy with RUX has been reimbursed in Poland for patients (pts) with MF, and large proportion of these pts has been followed for drug efficacy and tolerance within registry of the Polish Adult Leukemia Group (PALG). Objectives: The primary objective of the study was to identify baseline factors influencing probability of achievement of early response to RUX in MF. Patients and Methods: We retrospectively analyzed outcome of treatment with RUX after 6 months of therapy within the Polish Ministry of Health RUX reimbursement program in pts with primary MF (PMF) and MF secondary to polycythemia vera (post-PV MF) or essential thrombocythemia (post-ET MF). The inclusion criteria to the RUX reimbursement program were as follows: 1) diagnosis of MF confirmed by a recent bone marrow biopsy, 2) intermediate-2 or high risk IPSS, and 3) presence of splenomegaly and constitutional symptoms. Response to RUX was objectively categorized in regard to splenomegaly and MF symptoms. "Spleen Response" was defined as ≥ 50% reduction of the difference between baseline maximal length of spleen and upper limit of normal (120mm) measured by ultrasonography while "Symptoms Response" was defined as ≥50% reduction of the MF constitutional symptoms as assessed by MPN-SAF TSS score. Influence of potential prognostic factors (age, sex, time from MF diagnosis to RUX treatment, PMF vs post-ET MF and post-PV MF, leukocytosis, number of platelets, hemoglobin level, IPSS, DIPSS, grade of fibrosis in bone marrow, genetic status of MF) on probability of response was tested by univariate and multivariate logistic regression. Results: We enrolled 266 MF pts (56.8% PMF, 26.7% post-PV MF, 16.5% post-ET MF) undergoing treatment with RUX in 14 Polish hematology centers. Pts' median age was 67 years (range 21-86), while 59% of the group were females. Thirty-two% of pts were classified as high risk, 56% as intermediate-2 risk and 12% as intermediate-1 risk according to DIPSS score. JAK V617F mutation was found in 83%, CALR in 12%, MPL in 1.5%, and 3.5% of pts were triple negative. The median time from diagnosis of MF to start of RUX therapy was 30 months (range 0-321). At RUX therapy initiation median spleen length was 210 mm (range 100-300), and median symptoms score was 41 (range 4-93). During first six months the therapy has been continued in all but 26 (9.8%) pts in whom the treatment was withdrawn mainly because of lack of effectiveness. The most common hematologic adverse events included anemia (grade ≥3 in 27% of pts) and thrombocytopenia (grade ≥3 in 8% of pts). Major thrombotic events occurred in 3 pts. Among infectious complication 11 upper respiratory tract infections, 4 urinary tract infections, 3 herpes zoster infections, 3 pneumonias of unknown etiology and 1 case of tuberculosis were reported. Regarding the established categories of response to RUX, at 6^th month of therapy "Spleen Response" was detected in 50% of patients. The only pre-treatment factor significantly associated with the probability of achievement of "Spleen Response" by multivariable analysis was leukocytosis ≥25 G/L (OR 3.57, 95%CI 1.40-9.06, p=.007). In contrast "Symptoms Response" after 6 months of RUX therapy was found in larger proportion of patients accounting for 77% of the studied population. Multivariable analysis revealed that time-interval between MF diagnosis and RUX start (p=.028) and platelets >200 G/L (OR 2.09, 95%CI 1.16-3.77, p=.014) remained significant predictors of "Symptoms Response". Conclusions: In this study we found that in real life setting efficacy of RUX is more promising, especially in terms of symptoms reduction, and better tolerated, as compared to the results of randomized COMFORT studies. Furthermore, our analysis revealed that simple laboratory parameters such as leukocytosis and platelets number may constitute useful predictors of early response to RUX therapy. Finally, since shorter time interval between MF diagnosis and RUX therapy correlated with better "Symptoms Response", introducing RUX earlier during the course of the disease may be more beneficial for pts.