Estimated IQ Systematically Underestimates Neurocognitive Sequelae in Irradiated Pediatric Brain Tumor Survivors

Abstract

While cranial irradiation is often essential to attain a cure for pediatric brain tumors, it is associated with adverse neurocognitive outcomes, primarily manifested as declines in full-scale IQ (FSIQ). This decline results primarily from a decline in processing speed, a component of FSIQ. However, estimated IQ (EIQ) is frequently used in research and clinical settings but does not incorporate processing speed. We hypothesized that EIQ systematically underestimates neurocognitive sequelae in irradiated pediatric brain tumor survivors. We treated 185 pediatric brain tumor patients with proton radiation therapy. All patients had at least 1 neuropsychological evaluation at baseline and/or 1 or more follow-up evaluations with sufficient data to calculate both FSIQ and EIQ. The Wechsler Intelligence Scales were used to calculate FSIQ and EIQ for each patient, and mixed linear models were used to assess disparities between FSIQ and EIQ. At baseline, EIQ was 2.2 points (95% confidence interval 1.2-3.2 points) higher on average than FSIQ (P<.001). The median follow-up period was 26months. The disparity between EIQ and FSIQ persisted and worsened over time (P=.012), with FSIQ losing on average 0.4 points/year (95% confidence interval 0.01-0.8 points/year) relative to EIQ. The disparity at baseline varied with sex and age. EIQ systematically underestimates the neurocognitive sequelae of children treated with cranial radiation therapy. FSIQ is much more likely than EIQ to identify neurocognitive deficits, allowing for appropriate interventions, as well as academic services and accommodations. Thus, EIQ should have a very limited role in both clinical and research settings for this population.