EPEN-03. PEDIATRIC INTRACRANIAL EPENDYMOMA: CORRELATION OF SYMPTOMS AND SIGNS AT RECURRENCE WITH OUTCOME IN THE SECOND PROSPECTIVE AIEOP PROTOCOL FOLLOW-UP

Abstract Aims of patient follow-up are: discovering relapse to apply second-line therapy for possible cure, accruing patients in phase 1/2 protocols if second-line therapy is not standardized/curative, evaluating/treating iatrogenic effects. To lessen the patient/family emotional and social economic burdens, we should ideally understand if scheduled radiological follow-up has a better rationale/outcome than the symptomatic relapse. We analyzed the Italian series of 2016 (doi:10.1093/neuonc/now108) comprehending 160 newly diagnosed pediatric/adolescent patients with intracranial ependymoma (EPD) according to relapse status (recurrence at scheduled exam [RECPT]/sign-symptomatic recurring [SYMPPT]). RECPT and SYMPPT were assessed in Cox model as time-dependent variables for overall survival (OS). Differences in characteristics between RECPT and SYMPPT subgroups were assessed by Wilcoxon-Mann-Whitney or Fisher Exact tests. Kaplan-Meier curves depicted event-free survival (EFS) and OS after recurrence according to signs/symptoms. There were 16 SYMPPT /34 RECPT, at a median follow-up of 55 months. No significant differences were found between SYMPPT/RECPT for gender, age, tumor grade/site, shunt, residual disease, radiation boost, type of relapse (local/distant/combined). Time-to-relapse (median:19 months, range 5–104) and treatment adoption did not differ between SYMPPT/RECPT while signs/symptoms depicted an unfavorable factor for OS after relapse (5-year OS: 8% vs 37%). Adjusted model in multivariate analysis confirmed that presence of signs/symptoms was prognostically unfavorable. In order to improve the OS in symptomatic patients, their early identification-in the first two/three years after diagnosis- would be desirable. Perhaps, this could be obtained paying attention to “heuristically detected” subgroup of patients, not just describing general prognostic variables but discovering possible biology differences according to signs/symptoms.