Background: Immune checkpoint inhibitors targeting PD-1 have shown clinical benefit in adults with cancer, but data on these drugs in children are scarce. We did a phase 1-2 study of nivolumab, a PD-1 blocking monoclonal antibody, to determine its safety, pharmacokinetics, and antitumour activity in children and young adults with recurrent or refractory non-CNS solid tumours or lymphoma.

Methods: We did a multicentre, open-label, single-arm, dose-confirmation and dose-expansion, phase 1-2 trial in 23 hospitals in the USA. Eligible patients for part A (dose-confirmation phase) of the study were aged 1-18 years with solid tumours with measurable or evaluable disease (by Response Evaluation Criteria in Solid Tumors [RECIST] version 1.1) regardless of histology. Eligible patients for part B (dose-expansion phase) were aged 1-30 years with measurable disease (by RECIST criteria) in the following disease cohorts: rhabdomyosarcoma, Ewing sarcoma, osteosarcoma, neuroblastoma, Hodgkin lymphoma, non-Hodgkin lymphoma, and melanoma. Patients in part A and were given nivolumab 3 mg/kg intravenously over 60 min on days 1 and 15 of a 28-day cycle in a rolling 6 study design with de-escalation upon dose-limiting toxicities to establish the recommended phase 2 dose. Patients in part B were given the recommended phase 2 dose. The primary outcomes were the tolerability, systemic exposure, maximum tolerated dose, and the antitumour activity of nivolumab at the adult recommended dose in children and young adults. This trial is registered with ClinicalTrials.gov, NCT02304458, with follow-up ongoing and is closed to new participants.

Findings: 85 patients were enrolled between Feb 22, 2015, and Dec 31, 2018, and 75 patients were fully evaluable for toxicity. Median follow-up was 30 days (IQR 27-83). In part A, 13 patients were enrolled and 12 were evaluable for toxicity. There were no dose de-escalations or dose-limiting toxicities and nivolumab 3 mg/kg was confirmed as the paediatric recommended phase 2. 72 patients were enrolled in part B and 63 were evaluable for toxicity. Five (7%) patients in part B had dose-limiting toxicities. The most common overall toxicity was anaemia (35 [47%] of 75 patients; five patients had grade 3 or grade 4) and non-haematological toxicity was fatigue (28 [37%] patients; none had grade 3 or grade 4). Responses were observed in patients with lymphoma (three [30%] of ten with Hodgkin lymphoma and one [10%] of ten with non-Hodgkin lymphoma; all responders had PD-L1 expression). Objective responses were not observed in other tumour types.

Interpretation: Nivolumab was safe and well tolerated in children and young adults and showed clinical activity in lymphoma. Nivolumab showed no significant single-agent activity in the common paediatric solid tumours. This study defines the recommended phase 2 dose and establishes a favourable safety profile for nivolumab in children and young adults, which can serve as the basis for its potential study in combinatorial regimens for childhood cancer.

Funding: Bristol-Myers Squibb, Children's Oncology Group, National Institutes of Health, Cookies for Kids Cancer Foundation.