Seizure prophylaxis with valproic acid in pediatric patients with brain tumors Francisco Hélder Cavalcante Félix 2 , Igor Moreira Veras 3 , Cleto Dantas Nogueira 4 , Jony Arrais Pinto Junior 5 , Orlandira Leite de Araújo 2 , Nádia Mendonça Trompieri 2 , Juvenia Bezerra Fontenele 1 1 - Curso de Farmácia, Faculdade de Farmácia Odontologia e Enfermagem, Universidade Federal do Ceará 2 - Hospital Infantil Albert Sabin, Secretaria de Saúde do Estado do Ceará 3 - CRIO - Centro Regional Integrado de Oncologia 4 - Argos Patologia e Faculdade de Medicina da UFC 5 - Departamento de Estatística - Universidade Federal Fluminense [email protected] XIV Congresso Brasileiro de Oncologia Pediátrica — SOBOPE 2014 Introduction V alproic acid (valproate sodium, VPA) is an anti- epileptic drug (AED) widely used for treating sei- zures in children. It has also been used for seizures asso- ciated with brain tumors [1]. Seizure prophylaxis is con- troversial in brain tumor patients. However, a recent me- tanalysis was not conclusive due to the lack of appropriate clinical data [2]. Seizures are more frequent in pediatric patients with a supratentorial tumor but can nonetheless infrequently occur in posterior fossa tumor patients, and their determinants have remained obscure[1]. In our cen- ter, we deal with patients that have inadequate access to medical services, and there have been occasional episodes of seizures in children with posterior fossa brain tumor under our care. Based on these local differences, we con- sidered using prophylactic AED administration to all pe- diatric patients with brain tumors. After the beginning of prophylactic administration we observed a trend towards longer survival in a subset of our patients. In order to study the possible influence of valproate in the survival of pediatric patients with brain tumors we have underta- ken a retrospective cohort study. We also compared this cohort with a historical control from our institution. Due to modifications in referral policies in our state, the admis- sion rate of new patients in our center more than doubled from before 2007 to nowadays. [3]. Methods The institutional review board of our center approved this re- trospective study. We reviewed the charts of patients referred to our institution and diagnosed between January 2000 and December 2008 with primary brain tumors, aged 0–17 years. A member of our team reviewed magnetic resonance images (MRI) for tumor site and cha- racteristics. Some of the images from patients diagnosed before 2006 were not available for review, so we used the description of the radi- ological report to classify the tumors. Usual image parameters (Mac- Donald’s criteria) were used to analyze MRI and judge disease pro- gression [4]. The primary study endpoint for treatment efficacy was time to death due to any cause, from which overall survival (OS) was computed. The primary objective of the statistical analysis was to determine whether the valproate treatment produced a difference in the median overall survival. Survival functions were fitted with the Kaplan-Meyer method [5] and compared with log rank test, expres- sing the results as hazard ratios (HRs) with 95% confidence intervals (CIs). We then performed multivariate analysis with Cox proportio- nal hazards model and the additive hazards Aalen Model to compare the survival function of groups, adjusting for demographic, biomedi- cal, imaging, disease, and treatment characteristics [6, 7]. We con- sidered the use of valproate as a time-dependent covariate in order to adjust the survival model. Pearson’s chi-squared test was used to compare the risk of new-onset seizures between treated and non- treated groups. Data descriptive statistics and statistical calculations were performed on R 2.12 for Mac OSX (R Foundation for Statistical Computing, 2010). Results Figura 1: Global survival estimates of whole cohort (Kaplan-Meier,dashed lines:±95%CI ) and by covariates (Aalen model). Radiation therapy, surgical ressection, and hystological group were statistically significant (p< 0.005) One hundred, sixty five patients were included, 70 fe- male, median age 7.4 years, 34% embryonal tumors, 19% low grade gliomas, 27% brainstem tumors. Median ove- raal survival for the whole cohort was 29 (19-36) months. Median follow-up time was 24 months. Eighty patients received prophylatic VPA. Median OS of patients that received prophylatic VPA was 35 months, and median OS of non-treated patients was 15 months. Two-year OS was 39% (29-51) for non-treated and 66% (56-78) for tre- ated patients. Cox model could not fit the data due to time-dependent variation of VPA treatment. A regression was adjusted with Aalen’s time-dependent model, tes- ting VPA prophylaxis influence on survival probability. Radiation therapy, surgery, and hystological group had statistically significant (p< 0.005) influence on survival probability of the whole cohort of patients, but not VPA treatment. Aalen-model estimated survival probability of both groups was 46% (non-treated) ans 55% (treated) at 2 years and 41 and 46% at 5 years. At the end of the fol- low up time, 69 patients were alive. Twenty-one patients had had seizures of new onset after diagnosis, 12 of them had been treated with VPA prophylaxis. A comparison with Pearson’s chi-squared test yielded a non-significant difference in new-onset seizures risk between treated and non-treated groups. Conclusions Valproic acid, thus, had no effect in survival pro- bability of a heterogenous group of pediatric patients with brain tumors. Seizure prophylaxis with VPA for this cohort of patients was apparently ineffective, so it should not be routinely used for children with brain tumors. References [1] WELLS EM, Gaillard WD, Packer RJ Pediatric brain tumors and epilepsy. Semin Pediatr Neurol 2012;19(1):3–8 [2] TREMONT-LUKATS I, Ratilal B, Armstrong T et al Antiepi- leptic drugs for preventing seizures in people with brain tumors. Cochrane Database Syst Rev 2008;16(2):CD004424 [3]FELIX FH, Trompieri NM, de Araujo OL, da Trindade KM, Fontenele JB. Potential role for valproate in the treatment of high-risk brain tumors of childhood-results from a retrospective observational cohort study. Pediatr He- matol Oncol 2011:28:556–570 [4] MACDONALD DR, Cascino TL, Schold SC Jr, Cairncross JG (1990) Res- ponse criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol 8(7):1277–1280 [5] KAPLAN EL, Meier P. Nonparametric estimationfrom incomplete observati- ons. J Amer Stat Assoc 1958;53(282):457-81. [6] ANDERSEN, P. and Gill, R. Cox’s regression model for counting processes, a large sample study. Annals of Statistics 1982;10:1100-1120. [7] MARTINUSSEN and Scheike, Dynamic Regression Models for Survival Data, Springer (2006). Aknowledgements SOBOPE 2014 - XIV Congresso Brasileiro de Oncologia Pediátrica