Histopathological Spectrum and Neuro-radiological Correlations of Childhood Intracranial Brain Tumors in Kenyatta National Hospital and Moi Teaching and Referral Hospital

Abstract

Background: Central nervous tumors are the leading solid tumors in the childhood population accounting for the majority of tumor mortality worldwide. The central nervous system is a specialized system which has on average 130 primary brain tumors. There are more than 100 different histological subtypes of brain tumors with varying incidences over regions. Previously childhood brain tumors (CBT) were uncommon in the African population, however there’s increasing number of cases reported. There’s limited data on childhood brain tumors as well as the histopathological distribution more so of primary brain tumors whose trends are being noted as increasing over time in Kenya. Our study aimed at assessing the spectrum as well as the level of correlation with imaging in diagnosis of brain tumors within the two settings. Objectives: The main objective was to evaluate the histopathological spectrum and neuro-radiological correlations of childhood intracranial brain tumors in KNH and MTRH. Methodology: Study site, design and population: This was a cross-sectional retrospective descriptive study done at KNH and MTRH. The study population was drawn from children who underwent treatment for brain tumors between 2015 and 2017 and whose tissue biopsies (87 cases) were available at the laboratory archives. Data collection and analysis: Patient’s biodata and histology reports including the intraoperative findings from the filed reports in the histopathology department were retrieved. The case blocks were retrieved for histological processing and analysis. Immunohistochemistry (IHC) was done on recommended cases and analyzed. Histopathological evaluation results were entered into Microsoft Excel 2013 and merged with the patients’ biodata and previous histopathology reports using unique patient identification cards. XII Univariate analysis was carried out to describe and summarize the data on the variables including age and gender. Appropriate descriptive statistics were reported depending on the type of data: Frequencies and proportions for categorical data such as tumor type, site, grade, imaging findings and/or measures of central tendency (mean/median/mode) and dispersion (SD/IQR) for patient’s age. Bivariate analysis using chi-square tests was done to evaluate the association between the patients’ gender and age group with brain tumor features. Chi-square statistics and corresponding p-values was reported. Data analysis was conducted at 0.05 level of significance. Main Outcomes: Majority of the affected population were of ages 5-9 years with females at (54%) of the total population. The most affected site was infratentorial compartment (48.3%) with gliomas and medulloblastoma being equally distributed within it (23%). Gliomas and medulloblastoma were the most predominant tumors at 71.3% with gliomas leading at 48.3%. Majority of the gliomas were low grade (69%) with pilocytic astrocytoma being the most common subtype (42.9%). Immunohistochemistry tests were done in all the cases (9%) whose initial and final diagnosis were not tallying. The IHC panel tests conducted had results which all tallied with the final diagnosis after second review. The overall sensitivity for the diagnosis of brain tumors through radiology was at 69.4%. The level of correlation of histopathological to radiological diagnosis was statistically insignificant with P and kappa values of 0.814 and -0.024 respectively. Conclusion: Gliomas and medulloblastomas were the commonest tumors at both centers similar to findings at centers in other studies around the world. Histopathological diagnoses have a high concordance of agreement among various morphologists. The level of correlation between histopathological and radiological diagnosis was high comparable to other findings conducted elsewhere within the country. XIII Recommendations: A standard neuro-radio-pathological proforma is recommended which synchronizes significant clinical, radiological and pathological details within the two departments with a view of ensuring data availability and synchronicity. There’s need to expand the study to other centers in the country to gain the spectrum seen in the country. Some of the cases i.e. 8 cases had their final diagnosis altered after IHC was conducted hence need for follow up of the patients with regards to therapy altercation and prognostication