Relapsing intracranial germ cell tumours warrant retreatment.

Background: The optimal therapeutic strategy for relapsing intracranial germ cell tumours (IGCTs) has not been clearly established.
Methods: Relapses of IGCTs, occurring from 01/01/1990 to 31/12/2014, were retrieved from the Societe Française d'Oncologie Pediatrique-TGM 90, 92 and GCT 96 protocols, and from the National Childhood Solid Tumour Registry. Refractory IGCTs were excluded.
Results: Forty-four relapsing IGCTs were identified: 14 were initially treated for histologically proven germinomas (germinoma group), 5 for non-histologically proven germinomas (putative germinoma group) and 25 for non-germinomatous germ cell tumours (NGGCTs) (NGGCT group). In the germinoma group, the 5-year event-free survival (EFS) and overall survival (OS) were 79% (95% confidence interval [CI]: 47-93) and 86% (95% CI: 54-96), respectively. Only one of the 11 patients treated with reirradiation experienced a further relapse. A trend of better EFS was observed for relapses at sites that were not initially involved: 5-year EFS of 100% versus 67% (95% CI: 28-88), p = 0.09. In the putative germinoma group, 4 of 5 patients experienced a further event, leading to 2 deaths. In the NGGCT group, the 5-year EFS and OS were 56% (95% CI: 35-73) and 60% (95% CI: 38-76), respectively. A significant improvement in outcomes after high-dose chemotherapy (HDC) was observed: 5-year OS of 72% (95% CI: 46-87) versus 29% (95% CI: 4-61), p = 0.006.
Conclusion: Relapsing germinomas are highly curable; reirradiation appears to play a key role. Histological proof at initial diagnosis if markers are negative is crucial. Despite inferior outcomes relapsing, NGGCTs can be cured in a significant proportion of cases provided intensive treatment including HDC is applied.
Competing Interests: Conflict of interest statement None declared.
(Copyright © 2020 Elsevier Ltd. All rights reserved.)