Interferon matched hydroxyurea for molecular response in a blood-cancer trial

Pegylated interferon-alfa, the drug many haematologists hoped would prove a disease-modifying advance over decades-old hydroxyurea in myeloproliferative neoplasms, did not produce more molecular responses when every randomised patient was counted, according to the Danish DALIAH trial published in eClinicalMedicine [s1]. At 18, 36, and 60 months, the two drugs were statistically indistinguishable on the response measure the trial set out to compare [s1].

The finding matters because of who ran it. DALIAH was an investigator-initiated, open-label trial funded largely by Danish regional and academic research funds rather than by a drug company, registered as a comparison of low-dose interferon against hydroxyurea, and testing a treatment question that commercial trials have mostly approached from the interferon-maker's side [s1][s2]. When the effort came from the clinicians who use the drugs, the honest answer on the primary endpoint was a draw — with an important asterisk.

What the trial tested

Myeloproliferative neoplasms (MPNs) — including essential thrombocythaemia, polycythaemia vera, and myelofibrosis — are chronic blood cancers in which the marrow overproduces blood cells, raising the risk of clots and, over years, progression [s1]. Hydroxyurea is the standard cytoreductive drug; pegylated interferon-alfa has generated excitement because it can drive down the mutated clone, a deeper "molecular" response that some hope translates into better long-term outcomes [s1].

DALIAH randomised newly diagnosed, treatment-naive adults with active MPN across nine centres in Denmark [s1]. Patients older than 60 were assigned 1:1:1 to hydroxyurea, pegylated interferon-alfa-2a, or pegylated interferon-alfa-2b; those aged 60 or younger were assigned 1:1 to one of the two interferon formulations [s1]. The primary endpoint was the proportion achieving a partial or complete molecular response at 18, 36, and 60 months, compared by intention-to-treat between hydroxyurea and the pooled interferon arms [s1].

What it found

Between Feb 7, 2012, and July 6, 2015, 203 eligible patients entered the modified intention-to-treat population, 38 (19%) on hydroxyurea and 165 (81%) on interferon [s1]. In that analysis, molecular response proportions were nearly identical: 19% versus 21% at 18 months (p=1·00), 19% versus 26% at 36 months (p=0·64), and 23% versus 24% at 60 months (p=1·00) [s1].

The picture changed when the analysis was restricted to patients who stayed on treatment. In this per-protocol group, interferon showed higher molecular response beyond three years — 23% versus 56% at 36 months (p=0·01) and 35% versus 67% at 60 months (p=0·03) [s1]. The reason for the gap between the two analyses lies in tolerability: by 60 months, 65% of interferon patients had discontinued, against 37% on hydroxyurea (p=0·0019), and all 13 patients who stopped treatment because of a treatment-related adverse event were in the interferon arms [s1].

Serious toxicity was otherwise broadly similar: grade 3 or higher adverse events occurred in 58% of the hydroxyurea group and 45% of the interferon group (p=0·21) [s1]. The two drugs differed in the character of their side effects — dyspepsia, fever, and urinary infections more common with hydroxyurea; fatigue, influenza-like illness, injection-site reactions, and myalgia more common with interferon [s1].

How to read it

The split between the intention-to-treat and per-protocol results is the whole story. Judged as they would actually be prescribed — started in everyone, tolerated by only some — interferon did not beat hydroxyurea, because so many patients could not stay on it [s1]. Judged among those who tolerated it, interferon drove deeper molecular responses over the long run [s1]. Both statements are true, and per-protocol comparisons are prone to bias because the patients who remain on a harder drug differ from those who quit, so the more conservative intention-to-treat result is the one to lead with [s1].

Two cautions belong here. The trial used 2008 disease criteria that have since been updated, and it is modest in size, especially the 38-patient hydroxyurea arm [s1]. It also measured molecular response, a laboratory marker, rather than clots, progression, or survival — the outcomes patients care about most [s1]. Readers weighing what a laboratory number means for a person can see the same gap in coverage of what a cancer stage actually means.

Why it matters

Interferon has been marketed and advocated as the disease-modifying future of MPN care, and an independent trial showing that its molecular advantage evaporates once real-world tolerability is counted is a needed corrective [s1]. It reframes the choice as one about who can stay on interferon, not whether interferon is simply better. Industry ties are not absent even here — the trial listed some company support, and its senior author reports funding from and advisory roles with several drug firms — but the design and funding were principally academic [s1].

What to watch

The open questions are whether biomarkers can identify at diagnosis the patients likely to tolerate and respond to interferon, and whether the deeper molecular responses seen in tolerant patients eventually translate into fewer clots or less progression [s1]. Longer follow-up and hard clinical endpoints, not laboratory response, will settle that [s1].

This article describes research and is not medical advice. Decisions about treating a myeloproliferative neoplasm are for patients and their treating clinicians.

Sources

• Interferon-α vs hydroxyurea in patients with myeloproliferative neoplasms (DALIAH): a multicentre, randomised, open-label, phase 3 trial in Denmark — eClinicalMedicine, 9 September 2026
• A Study of Low Dose Interferon Alpha Versus Hydroxyurea in Treatment of Chronic Myeloid Neoplasms (NCT01387763) — ClinicalTrials.gov, U.S. National Library of Medicine, last updated 27 April 2022