Recombinant Interferon Alpha-2a (r-IFNα) is a potent immunomodulating agent, which has been used off-label for the treatment of polycythemia vera (PV) for more than three decades and has been demonstrated to induce high rates of clinical, hematological and molecular responses. Only few studies have compared efficacy and safety of r-IFNα vs. hydroxyurea (HU), which is considered first line therapy for PV patients > 60 years in most parts of the world. However, recent studies have provided encouraging results for the treatment of PV with r-IFNα compared to HU irrespective of age (R. Hoffmann 2016; H. Gisslinger 2018).


To examine the difference in efficacy and safety of low-dose r-IFNα in PV patients ≤ 60 or > 60 years of age compared to HU > 60 years of age.


Ninety newly diagnosed or previously phlebotomized PV patients only (WHO 2008) were enrolled in the DALIAH trial (NCT01387763). All patients provided written informed consent. Patients ≤ 60 years were randomized (I:I) to r-IFNα-2a (Pegasys®) or to r-IFNα-2b (PegIntron®) whereas patients > 60 years were randomized (I:I:I) to either r-IFNα-2a, r-IFNα-2b or to HU. The starting dose of r-IFNα-2a and r-IFNα-2b was 45 or 35 µg/week, respectively. The HU dose was 500 to 2000 mg/day. Patients randomized to r-IFNα who presented with major thrombosis or platelets > 1500 109/L received HU from inclusion and until normalization of the platelet count. Efficacy assessment consisted of the clinicohematological and the molecular response rates by intention to treat analysis (ITT) using the European Leukemia Net (ELN) 2009 criteria. JAK2V617F analysis was performed by qPCR. Groups were compared by Fisher's Exact Test.


Three-year analysis was available in all but five patients (n=85) at time of abstract submission (Table 1). The analysis was performed after a median of 36 months (range: 33-39 months). The median treatment dose was 684 mg/day (IQR: 131 - 942) for HU, 51 μg/week (IQR: 30-90 μg/week) and 54 μg/week (IQR: 30-66 μg/week) for r-IFNα-2a age ≤ 60 and > 60, respectively, and 41 μg/week (IQR: 29-45 μg/week) and 36 μg/week (IQR: 31-37 μg/week) for r-IFNα-2b age ≤ 60 and > 60.

The overall clinicohematological response rate (ORR) was 68% (13/19) for HU, 42% (14/33) for r-IFNα ≤ 60 years and 39% (13/33) for r-IFNα > 60 years. The partial clinicohematological response rate (PHR) and the complete clinicohematological response rate (CHR) was 53% (10/19) and 16% (3/19) for HU, 9% (3/33) and 33% (11/33) for r-IFNα ≤ 60 years and 9% (3/33) and 30% (10/33) for r-IFNα > 60 years. Neither the ORR, CHR nor the PHR was significantly different between the three groups. Maintenance of CHR from first occurrence to data analysis after 36 months was 11% (2/19) for HU, 21% (7/33) for r-IFNα ≤ 60 years and 18% (6/33) for r-IFNα > 60 years.

Forty-seven JAK2V617F positive patients were available for molecular response analysis after 36 months of therapy. A partial molecular response (PMR) was detected in 21% (4/19) of HU treated patients and in 24% (7/29) of r-IFNα treated patients ≤ 60 years and 18% (6/33) of r-IFNα > 60 years. Notably, 7% (2/29) of the r-IFNα treated patients ≤ 60 years obtained a complete molecular response (CMR). The median JAK2V617F reduction from baseline was 38% (IQR: 31-63%) for HU, 79% (IQR: 59-92%) for r-IFNα ≤ 60 years and 73% (IQR: 49-97%) for r-IFNα > 60 years. There was no statistically significant difference in the PMR between groups.

Discontinuation of treatment for any reason after 36 months of therapy was 21% (4/19) for HU, 52% (17/33) for r-IFNα ≤ 60 years and 45% (15/33) for r-IFNα > 60 years. Toxicity related discontinuation was 5% (1/19) for HU and 30% (10/33) for both r-IFNα ≤ 60 and > 60 years. Grade 3-4 AEs occurred in 32% (6/19) of HU treated patients, 27% (9/33) in r-IFNα treated patients ≤ 60 years and in 42% (14/33) r-IFNα treated patients > 60 years. SAEs were reported in 21% (4/19) for HU, 9% (3/33) for r-IFNα ≤ 60 years and 24% (8/33) for r-IFNα > 60 years. The numbers of grade 3-4 AEs as well as SAEs were comparable between groups.


After 36 months of therapy CHR was non-significantly higher in PV patients treated with r-IFNα compared to HU by ITT, irrespective of age. Also, maintenance of CHR was longer for r-IFNα. However, ORR was non-significantly higher for HU. PMR was almost similar between the three groups but the median JAK2V617F reduction was greater for r-IFNα. Toxicity related discontinuation from study therapy was higher for r-IFNα compared to HU.


Stentoft:Bristol-Myers Squibb: Research Funding; Merck Sharp&Dohme: Research Funding. Hasselbalch:Novartis: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.