Adding nelarabine to adult T-ALL therapy did not improve survival in UKALL14

Adding a single cycle of the T-cell leukaemia drug nelarabine to standard chemotherapy did not improve outcomes for adults with T-cell acute lymphoblastic leukaemia, according to final results from the T-cell arm of the UKALL14 trial [s1]. The drug was safe and well tolerated, but it did not move the measure the trial was built to move — a clean, if deflating, answer to a question the field had hoped would go the other way.

Nelarabine is already approved for T-cell acute lymphoblastic leukaemia (T-ALL) that has relapsed or stopped responding, where it can clear disease but carries a reputation for nerve toxicity. The appeal of UKALL14 was to test whether giving it earlier, as part of first-line treatment, might stop relapses before they start. That is a different and higher bar: a drug can shrink refractory disease yet still fail to change the long-run odds when added up front.

What the trial did

UKALL14 (NCT01085617) was a phase 3 randomised controlled trial sponsored by University College London — an academic study, not a company registration [s1][s2]. Its T-cell arm enrolled adults aged 25 to 65 and randomly assigned them 1:1, at trial entry, to standard of care or to standard of care plus nelarabine 1.5 g/m² given on days 1, 3, and 5 after a second induction course [s1]. The primary endpoint was event-free survival (EFS) — the share of patients alive and free of relapse or treatment failure. From 2011 to 2018, 143 patients were randomised: 75 to standard of care and 68 to standard of care plus nelarabine [s1].

Because this was a randomised comparison rather than a single-arm study, it can isolate what the added drug did, and the honest reading is that it did very little.

What it found

Event-free survival was 58.2% with standard of care and 62.7% with the addition of nelarabine — a gap too small and too uncertain to count [s1]. The hazard ratio was 0.86, with a 95% confidence interval from 0.52 to 1.43 and a P value of 0.57, meaning the result is entirely compatible with no benefit at all [s1]. The three-year cumulative incidence of relapse was likewise near-identical: 31% with standard of care versus 28.9% with nelarabine [s1].

Relapses, when they came, came fast and were hard to salvage. Forty of 44 relapses — 90.9% — occurred within two years of randomisation, and median survival after relapse was just 5.7 months, with no difference between the arms [s1]. On the safety side, the news was better: severe adverse events and non-relapse mortality did not differ between groups, and crucially there was no excess of grade 3-4 neurotoxicity, the side effect that had most worried clinicians about using nelarabine earlier [s1].

How to read it

This is a negative trial, and it is worth saying plainly rather than hunting the subgroups for a silver lining. The drug was added to a backbone that already works reasonably well, the numerical difference in EFS pointed the right way, and the confidence interval was wide — all of which can tempt a reader into reading hope into the numbers. But a 95% interval that runs from a 48% reduction in events to a 43% increase is not evidence of benefit; it is evidence that the trial could not distinguish nelarabine from no nelarabine [s1].

What the study does establish is reassuring on toxicity: one consolidation cycle at this dose did not bring the neurological harms that limit nelarabine in the relapsed setting [s1]. That matters for how the drug is positioned, even though it does not rescue the efficacy result. The speed and lethality of the relapses — most within two years, median post-relapse survival under six months — also underline why adult T-ALL remains one of the harder leukaemias to treat, and why a modestly active drug added to induction was never going to be enough on its own [s1].

For context, related coverage has examined a trial testing whether the antibody blinatumomab can replace chemotherapy in childhood acute lymphoblastic leukaemia and the FDA approval of a new bispecific antibody in follicular lymphoma.

What to watch

The useful next questions are about sequencing and selection: whether nelarabine earns its place in specific high-risk subgroups, whether measurable-residual-disease testing can identify the patients most likely to relapse, and whether newer immunotherapies change the calculus for adult T-ALL altogether [s1]. For now, UKALL14 says that bolting one cycle of nelarabine onto standard induction does not improve survival — a result that, negative as it is, saves future patients from a treatment that would not have helped.

This article describes research and is not medical advice. Decisions about leukaemia treatment belong with treating haematologists.

Sources

• One cycle of nelarabine given after induction does not improve outcome in adults with T-cell precursor ALL: UKALL14 — HemaSphere, 1 October 2026
• UKALL14 trial registration (NCT01085617) — ClinicalTrials.gov