A Retrospective Cohort Study of Interferon-α Therapy in Myeloproliferative Neoplasms; Adverse Events, Thromboembolic Incidence and Causes of Termination of Therapy

Background

In myeloproliferative neoplasms (MPN), Interferon-α (IFN-α) has been shown effective in inducing hematologic and molecular responses and in reducing vascular events. In clinical practice its use is mainly limited by intolerance due to side effects.

Aim

We sought to evaluate the tolerability of IFN-α therapy, the thromboembolic incidence and the causes of termination of therapy in a cohort of MPN, treated outside of clinical trials.

Methods

One hundred patients (M/F 41/59, median age 48 years, range 15-73) with a diagnosis of polycythemia vera (PV, n=47), essential thrombocythemia (ET, n=43) and myelofibrosis (MF, n=10) according to the WHO 2008 criteria, on current or previous treatment with IFN-α (IFN-α-2b, Peg-IFN-α-2b, Peg-IFN-α-2a) were included. The patients, diagnosed 1987-2012, were recruited from 9 centers in Sweden and Norway, and retrospectively analyzed. Hematologic response in PV and ET was assessed according to ELN criteria from 2009. Response to treatment in MF was defined as platelets ² 400x10⁹ /L, white blood counts ² 10x10⁹/L and transfusion independency.

Results

IFN-α treatment characteristics are displayed in Table 1. The median treatment duration for IFN-α was 34 months. Treatment prior to IFN-α had been received by 44 pts including hydroxyurea (n=34), anagrelide (n=19), busulphan (n=2), radioactive phosphorus (n=1), 10 pts having received more than one cytoreductive agent. Complete hematologic response (CR) was observed in 58 pts (PV=28/47, ET=30/43) and partial hematologic response (PR) in 15 pts (PV=2, ET=13). In MF, hematologic response was noted in 8 out of 10 patients. IFN-α related adverse events (AE) were recorded in 76 pts (76/100, 76%) with similar rates between genders (M 30/41, 73%, F 45/59, 76%). AE were generally of low grade. Twenty pts experienced multiple (³ 3) side effects (M/F 6/14), females reporting a total of 96 AE compared to 53 in males. Hematologic toxicity was low with 4 pts presenting with anemia, 4 with leukopenia and 3 with thrombocytopenia. Most common non-hematologic toxicities were fatigue in 30 pts (M/F 11/19), myalgia in 28 (M/F 11/17) and depression in 21 (M/F 4/17), followed by liver function test elevation (n=9), headache (n=9), alopecia (n=8) and skin reaction (n=7). In two pts with autoimmune co-morbidities (rheumatoid arthritis, psoriasis), flare-up of symptoms related to autoimmune activity were seen, leading to discontinuation of therapy. Only one vascular event occurred in a 64 year old woman with PV, in CR since 92 months, who developed a myocardial infarction after 94 months of IFN-α-2b treatment. A total of 43 pts (M 16/41, 39%, F 27/59, 46%) discontinued therapy, of whom 34 (M 13/41, 32%, F 21/59, 36%) due to side effects. The most common cause of discontinuation of therapy due to side effects was depression (15/21), followed by fatigue (12/30) and myalgia (9/28). Nineteen (19/34, 58%) of the pts who discontinued therapy due to side effects were in CR. Discontinuation due to other reasons than side effects were lack of efficacy/progression of disease (n=5), co-morbidities (n=2), CR including molecular response (n=2) and pregnancy (n=1). Out of the 53 pts with ET and MF, 25 were JAK2^V617F mutated and 14 had a CALR-mutation. No significant differences between these two groups were seen regarding side effects or discontinuation rate. Out of the 57 pts remaining on IFN-α, 19 still received IFN-α-2b (19/35, 54%), 7 PegIFN-α-2b (7/12, 58%) and 31 PegIFN-α-2a (31/53, 58%).

Conclusion

In this retrospective cohort study, the treatment discontinuation rate due to side effects was higher than in previous reports. This may be explained by the relatively long median duration of treatment in this cohort, reflecting a poor tolerance of low-grade toxicity over time. Depression was frequent and the most common reported side effect when therapy was discontinued. Gender difference, with females reporting a higher incidence of depression and a larger total burden of AE, was noted. The frequency of thromboembolic events was very low in this IFN-α treated cohort.