Phase II and biomarker study of programmed cell death protein 1 inhibitor nivolumab and metronomic cyclophosphamide in paediatric relapsed/refractory solid tumours: Arm G of AcSé-ESMART, a trial of the European Innovative Therapies for Children With Cancer Consortium.

Abstract Purpose AcSe-ESMART is a European multicentre, proof-of-concept multiarm phase I/II platform trial in paediatric patients with relapsed/refractory cancer. Arm G assessed the activity and safety of nivolumab in combination with metronomic cyclophosphamide +/- irradiation. Experimental design Following a Phase II Simon two-stage design, nivolumab was administered intravenously at 3 mg/kg every 2 weeks of a 28-day cycle, oral cyclophosphamide at 25 mg/m2 twice a day, 1 week on/1 week off. The primary endpoint was objective response rate. Irradiation/radioablation of primary tumour or metastasis could be administered as per physician's choice. Biomarker evaluation was performed by tumour immunohistochemistry, whole exome and RNA sequencing, and immunophenotyping of peripheral blood by flow cytometry. Results Thirteen patients were treated with a median age of 15 years (range: 5.5–19.4). The main histologies were high-grade glioma, neuroblastoma, and desmoplastic small round cell tumour (DSRCT). The safety profile was similar to those of single-agent nivolumab, albeit haematologic toxicity, mainly lymphocytopenia, was commonly reported with the addition of cyclophosphamide +/- irradiation. Two patients with DSRCT and ependymoma presented unconfirmed partial response and prolonged disease stabilisation. Low mutational load with modest intratumour CD3+ T-cell infiltration and immunosuppressive tumour microenvironment were observed in the tumour samples. Under combined treatment, no positive modulation of circulating T cells was displayed, while derived neutrophil-to-lymphocyte ratio increased. Conclusions Nivolumab in combination with cyclophosphamide was well tolerated but had limited activity in this paediatric setting. Metronomic cyclophosphamide did not modulate systemic immune response that could compensate limited T-cell infiltration and the immunosuppressive tumour microenvironment. ClinicalTrials.gov Identifier NCT2813135.