Combination therapy with interferon and ruxolitinib for polycythemia vera and myelofibrosis: are two drugs better than one?

Approximately 35 years ago, recombinant interferon alpha-2 (rIFNα-2) was first reported to control myeloproliferation in essential thrombocythemia,[1] polycythemia vera (PV),[2] and the hyperproliferative phase of primary myelofibrosis (MF).[3] In the ensuing years, these observations have been substantiated in thousands of patients with myeloproliferative neoplasms, indicating that rIFNα-2 is safe and effective for alleviating symptoms, diminishing organomegaly, reducing elevated platelet and white blood cell counts and, in PV, maintaining and controlling the hematocrit[4] and decreasing the level of the JAK2 V617F allele burden.[5] The effect of rIFNα on the molecular markers of essential thrombocythemia and MF have also been noted.[6,7]

The finding of a significant reduction of JAK2 V617F level in a subset of PV patients treated with a long-acting form of interferon owing to its pegylation (PEG- rIFNα-2a)[8] led to more enthusiastic use of this treatment. Subsequently, normalization of marrow morphology and reduction of fibrosis were demonstrated.[9] Interferon monotherapy normalized elevated blood cell counts within a few months, but the major molecular responses and marrow changes required 3-5 years of treatment, especially in MF.[10] Discontinuation of rIFNα in PV could be sustained for a number of years.[3,8] Patients with primary or secondary MF usually require continued treatment.[10]

As with all potent drugs used in treating hematologic cancers, rIFNα is associated with a significant number of side effects, but most commonly, constitutional symptoms such as fatigue, muscle aches, lethargy and, occasionally, fever.[7]

This is presumably due to cytokine effects. Discontinuation rates in rIFNα studies have ranged from 10-30% after 1-2 years,[4,7,8] depending on the dose and frequency of administration of the drug, the severity of associated side effects, and the belief and enthusiasm of the physicians and patients, respectively, regarding its value. Recently, it has been suggested that considering side effects to be all related to dosing, per se, is too simplistic. Danish investigators indicated that chronic inflammation induced by interferon renders patients intolerant or refractory to the treatment.[11] Adding an anti-inflammatory drug with an anti-JAK2 V617F effect, such as ruxolitinib, seemed logical (Figure 1).

Figure 1.

Overcoming inflammatory-resistance to interferon with ruxolitinib.

The study by Sørensen et al., reported in this issue of Haematologica comprised 50 patients with PV or with primary or secondary MF, all of whom were resistant or refractory to rIFNα−2.[12] After the addition of ruxolitinib to the treatment regimen, of 32 PV patients, ten (31%) achieved complete or partial remission; of 18 MF patients, eight (44%) achieved complete or partial remission.

Combination treatment seemed to speed the time to remission, improve blood counts, reduce marrow cellularity and fibrosis and decrease the JAK2 allele burden, all with acceptable toxicity. The dropout rate at the end of 2 years was 6% for PV patients and 32% for MF patients.

These results are most interesting and encouraging, but require confirmation, because it was a single-arm study. Other limiting features, as the authors point out, include the small number of cases, the lack of a dose-finding phase, and the duration of only 2 years of treatment. Detailed molecular reporting of the MF patients in particular would have been of interest, since it is known that initial molecular profile affects response to interferon treatment in early MF.[10] Nevertheless, conceptually, the basis for the use of two potentially synergistic agents with different activities in untreated patients is logical. This then leads to the question of whether or not this combination should be used as initial therapy. The current tendency worldwide is to use hydroxyurea as firstline treatment when needed in PV and to adopt a watchand- wait attitude in treating primary MF. This approach seems illogical since it is not consistent with our concepts of cancer treatment in general, wherein treatment of early cancer yields results superior to those of treating metastatic disease. The reduction and/or elimination of symptoms of disease and avoidance of progression with agents that have a biological basis for use and that are tolerable seems far more rational. This requires testing in the immediate future.