Classical Philadelphia-negative myeloproliferative neoplasms: focus on mutations and JAK2 inhibitors

The discovery of driver mutations in MPNs has prompted the development of molecularly targeted therapy. JAK2 inhibitors acting on JAK-STAT signaling cascade lead to increased cell apoptosis and reduction of cell proliferation [17]. Among JAK2 inhibitors, ruxolitinib (RUX)-dual JAK1/JAK2 inhibitor has been approved for the treatment of intermediate-2 and high-risk MF [18].

The primary end point of all MF trials was a reduction of 35% or more in spleen volume (SVR) from baseline at 24 or 48 weeks depending on study. The secondary end points included time to a SVR, progression-free survival (PFS), leukemia-free survival (LFS), and overall survival (OS). There have been COMFORT-1 and COMFORT-2 studies which evaluated the efficacy and safety of RUX in MF patients. In COMFORT-1 trial, treatment with RUX resulted in SVR ≥ 35% at 24 weeks in 42% of treated patients if compared with less than 1% in placebo group. In COMFORT-2 trial, SVR ≥ 35% at week 48 was noted for 28% of patients receiving RUX vs 0% in those who received best available therapy (BAT). Spleen reduction was maintained for more than 3 years in RUX patients either in COMFORT-1 and COMFORT-2 trials. The pooled-analysis of both studies has shown that the risk of death was reduced by 30% in RUX group if compared with the control. Moreover, patients who were originally randomized to RUX survived longer than those who crossed over from placebo/BAT to RUX [18‐20]. The COMFORT trials have demonstrated that larger spleen size at baseline was associated with higher risk of death whereas spleen reduction on RUX resulted in better survival. Of note is, that patients with at least 25% spleen reduction on RUX had prolonged survival if compared with those without change in spleen volume [20, 21]. Moreover, it has been shown that RUX alleviated disease symptoms improving patient’s quality of life. An increased inflammatory responses have been demonstrated in patients with MPNs as a consequence of activation of a variety of cytokines. RUX was found to decrease serum cytokine levels reversing the constitutional symptoms thereby leading to clinical improvement [20‐22]. These cytokines are produced by the mutated and non-mutated hematopoietic cells leading to progressive bone marrow fibrosis. Moreover, they seem to play a role in development of extramedullar hematopoiesis [23]. Interestingly, RUX treatment led to reduction of bone marrow fibrosis in 16% of MF patients in COMFORT-II trial [20]. MD Anderson group has reported on the results of bone marrow biopsies at baseline and at 24, 48, and 60 months of RUX. Improvement of fibrosis was demonstrated in 15%, 34%, and 36%, respectively. These results have proved that JAK inhibition modulates the cellular components of bone marrow microenvironment which are thought to be related to fibrosis [24]. Of note is, that only three cases with complete resolution of bone marrow fibrosis on RUX have been reported so far [25‐27]. Long-term treatment with RUX is associated with opportunistic infections [28].

It was demonstrated that RUX altered the JAK2 V617F allele burden in JAK2-positive patients with MF. Moreover, the decreases of allele burden correlated with spleen volume reduction and were inversely associated with disease duration. Only 11% of analyzed JAK2-positive MF patients achieved partial (PMR) or complete molecular response (CMR). Median time to PMR and CMR was 22 and 27 months, respectively. Interestingly, the spleen responses were also demonstrated in patients without allele burden reduction or in those who were JAK2- negative at baseline. Moreover, changes in allele burden did not correlate with improvement of blood parameters, clinical symptoms and bone marrow histopathology [29].

Treatment with RUX in MF can be associated with prolonged survival and reduced risk of death independent of mutation profile. Data from COMFORT-II trial have clearly demonstrated that RUX abrogated the negative impact of high-risk molecular markers. Of note is that these detrimental mutations did not decrease the probability of obtaining spleen reduction or improving clinical symptoms on RUX [30].

Anemia and thrombocytopenia are dose-dependent effects of RUX therapy. Grade 3/4 anemia and thrombocytopenia occurred in 45% and 13% of patients, respectively. One should keep in mind that RUX leads to drop in hemoglobin levels and platelet counts over the first 12 weeks of therapy with subsequent recovery. These RUX-related events require dose reductions or brief treatment interruptions as well as red blood cell or platelet transfusions [31]. Interestingly, the occurrence of RUX-induced anemia does not affect survival [32]. Non-hematologic side effects of RUX are mild and manageable. Of note is that two cases of secondary acute myeloid leukemia were reported in patients receiving RUX [31].

However, it should be noted, that data on RUX efficacy and safety for MF patients provided by Cochrane Systemic Review were not so evident as those quoted above. There was very low or low quality of evidence for the effect of RUX on OS and PFS when compared with placebo or BAT. Similarly, RUX did not decrease the risk of hematologic and non-hematologic toxicity when compared to comparators. The effect of RUX on reduction of spleen size remained uncertain. One should bear in mind that these conclusions were based on COMFORT-I/II industry sponsored trials and included a small number of patients [33].

The safety and efficacy of momelotinib (MMB)- JAK1/2 inhibitor for intermediate and high-risk MF were evaluated in two randomized trials: SIMPLIFY-1 and SIMPLIFY-2. The former one compared MMB with RUX in patients who had not been treated with JAK2 inhibitors before. The primary study end point was similar to that of COMFORT trials. A proportion of patients who achieved ≥ 35% reduction in spleen size at 24 weeks was comparable between treated arms: 27% in MMB vs 29% in RUX. MMB was inferior when compared with RUX in terms of symptoms resolution. MMB treatment resulted in a significant decrease of transfusion requirement. Most common side effects of MMB included anemia and thrombocytopenia; 10% of treated patients developed grade ≤ 2 peripheral neuropathy [34]. The SIMPLIFY-2 study compared MMB with BAT in MF patients who had received previous RUX. MMB was not superior to BAT (including RUX) in decreasing spleen size (7% in MMB vs 6% in BAT group). Adverse events were similar between groups except neuropathy which occurred in 10% of MMB-treated patients and 0% in BAT population [35]. Recently, the Mayo Group have presented their results of a 7-year follow-up of MMB treatment in MF patients who were treated between 2009 and 2010 as a part of phase 1/2 study. A vast majority of patients (91%) discontinued treatment after median of 1.4 years, the remaining 9% are still on MMB with maximum duration of treatment exceeded 7 years. Common grade 3/4 adverse events of MMB treatment included thrombocytopenia and liver enzymes increase. Clinical improvements measured as spleen and anemia responses were demonstrated for 57% of MMB-treated patients. Majority of patients improved their quality of life. The presence of ASXL1, SRSF2 and absence of CALR type-1 mutation was associated with unfavorable survival [36].

Pacritinib (PAC) is a kinase inhibitor for JAK2 and FLT3. Data from phase 1/2 trial study have shown that PAC induced spleen reduction with limited hematologic toxicity [37]. The PERSIST-1 study compared the efficacy of PAC (twice and once daily) vs BAT (excluding RUX) in high-risk MF patients. At 24 week SVR ≥ 35% was achieved by 19% patients in PAC group vs 5% in BAT (p = 0.0003) [38]. In the PERSIST-2 study, patients with myelofibrosis and platelet count ≤ 100 × 10⁹/L were randomized to PAC or BAT (including prior RUX). PAC was found to be more effective than BAT in SVR ≥ 35% (18% vs 3%) and led to greater than 50% reduction in total symptom score (25% vs 14%). Of note is, that Food and Drug Administration (FDA) in 2016 placed a full clinical hold on PAC studies following fatal complications including cerebral bleeding and cardiac failures. In January 2017 FDA has lifted its hold on PAC and a new dose-finding study PAC203 was initiated [18, 39]. Patients who benefited from PAC in PERSIST studies were allowed to continue the treatment on compassionate-use basis. Re-initiation of PAC resulted in mild improvement in SVR with low rate of hematologic toxicity. It seems reasonable to offer PAC for patients with MF and pancytopenia, especially thrombocytopenia [40].

Fedratinib (FED) is a JAK2 inhibitor which efficacy was first evaluated in the JAKARTA-1 study. FED at two doses (400 mg and 500 mg) was compared with placebo in 289 patients with intermediate-2 or high-risk MF. Spleen response was achieved in 36% of FED-treated patients at 400 mg, 40% at 500 mg and 1% in the placebo group. Four patients receiving FED at 500 mg daily developed Wernicke encephalopathy (WE) [41]. The JAKARTA-2 study included patients with MF who were resistant to or intolerant to RUX. Spleen response was noted in 53% of patients who were RUX resistant and 63% of patients who were RUX intolerant. However, it should be mentioned that 7% of patients died during the study as a consequence of encephalopathy. It was the reason for study discontinuation [42]. Recently, the analysis of 9 FED trials with 670 patients has revealed that the prevalence of encephalopathy in this population is less than 1%. Based on this data, FDA decided to lift the clinical hold [43].

RUX has been approved for PV patients who are resistant to or intolerant to hydroxyurea based on a phase 3 study (RESPONSE) comparing RUX with BAT. The primary end point was both hematocrit control and ≥ 35% SVR reduction and at week 32. Hematocrit control was observed in 60% of patients in RUX group when compared to 20% of those in BAT arm. At least a 35% SVR achieved 38% and 1% patients, respectively. There was a significant difference in complete hematologic remission (CHR) rate: 24% in RUX vs 9% in BAT. Moreover, the reduction in total symptom score was greater in RUX patients. Side effects were manageable and similar to that presented in COMFORT trials [44]. The RESPONSE trial has also focused on long-term effect of RUX treatment on JAK2V617F allele burden. JAK2-positive PV patients were randomized to RUX or BAT with cross-over to RUX at week 32. It was demonstrated that mean decrease of JAK2 allele burden from baseline ranged from − 12 to − 40% (RUX arm) and − 6 to − 18% (RUX cross-over). CMR and PMR were demonstrated in 3 and 54 patients, respectively. However, one should bear in mind that the reduction in allele burden did not correlate with clinical outcome. These findings were consistent with those seen for MF [45]. Recently, a 4-year follow-up of RESPONSE study has been presented. The median duration of CHR was not reached and 54% of patients remained in CHR. The estimated 5-year OS rates were comparable between arms. Hematologic toxicity improved with RUX continuation as well as non-hematologic adverse events. Infection was the most common side effect of RUX. Rates of thrombotic events were low in both arms [46].

The MAJIC-ET study compared the efficacy of RUX vs BAT for ET intolerant or resistant to HU. Fifty-eight patients were randomized to RUX, however, there was no evidence of superiority of RUX when compared to BAT at 12 months [47]. In contrast, MD Anderson study has demonstrated that RUX treatment was associated with the improvement in ET-related symptoms as well as better and durable platelet, leukocyte, and hemoglobin control. The reduction of JAK2-allele burden was also noted [48].

A phase 2 trial evaluated MMB in patients with PV and ET, however, it was prematurely discontinued due to limited efficacy [49]. The results of studies with JAK2 inhibitors in classical MPNs are shown in Table 2.