Why Europe said no to Mylotarg for children with leukaemia
The EMA has declined to extend gemtuzumab ozogamicin to children with newly diagnosed acute myeloid leukaemia, a use the FDA approved in 2020. The reasoning is a lesson in paediatric trial design.
In brief
The EMA has declined to extend Mylotarg to children with newly diagnosed AML, citing uncontrolled data and no survival gain. The FDA approved it in 2020.
- On 16 October 2026 the EMA published its decision not to extend Mylotarg to children aged 1 to under 18 with newly diagnosed AML.
- The main study gave Mylotarg to all 515 children, so it could not isolate the drug’s effect.
- The FDA approved the drug for children from one month of age in June 2020, on the same kind of event-free survival evidence.
The European Medicines Agency has declined to recommend extending Mylotarg (gemtuzumab ozogamicin) to children aged one to under 18 with newly diagnosed CD33-positive acute myeloid leukaemia (AML). The outcome, published on 16 October 2026 after the CHMP’s October meeting, leaves the antibody-drug conjugate authorised in the EU only for patients aged 15 and over.
The decision puts European and US regulators on opposite sides of the same evidence base. The FDA extended Mylotarg to newly diagnosed paediatric patients aged one month and older in June 2020.
What was submitted
The proposed use combined Mylotarg with mitoxantrone and cytarabine, a different backbone from the authorised adult regimen of daunorubicin and cytarabine. The main study enrolled 515 children with untreated AML, who received one or three doses during first-line induction, with event-free survival as the primary measure.
According to the EMA, the study did not show that Mylotarg improved event-free survival. Because every child received the drug, there was no comparator that could show whether it outperformed no Mylotarg, or how much it contributed to side effects.
The supportive trial
A second study, in 1,063 children and young adults, randomised patients to standard chemotherapy with or without two doses of Mylotarg. Event-free survival improved, but overall survival did not, and the EMA noted that more deaths occurred among patients in complete remission in the Mylotarg group. The agency also cited uncertain outcome assessment, more incomplete data in the Mylotarg arm, and differences in schedule and chemotherapy that limited relevance to the proposed use.
An event-free survival gain without an overall survival gain is a result regulators can read in two directions.
The FDA’s 2020 decision rested on AAML0531, a randomised trial of 1,063 patients aged up to 29. It reported five-year event-free survival of 48% with gemtuzumab ozogamicin against 40% with chemotherapy alone, a hazard ratio of 0.84, and no demonstrated difference in overall survival.
Thin safety data
The EMA also judged the safety evidence too limited. Side effects were collected only for selected, prespecified types and only for 30 days, a period during which many patients still had neutropenia. In a disease where treatment-related deaths matter as much as relapse, that window left open questions.
What changes
Little changes in practice for European patients already in trials: the company told the agency there are no consequences for patients in clinical studies or compassionate use programmes. The product information will be updated with data from the paediatric study, so that prescribers can see the results even though the use is not authorised.
For sponsors, the episode is a reminder that a paediatric package built on uncontrolled add-on data and an event-free survival gain without a survival benefit may satisfy one regulator and not another.
