Continuous Real-Time Neuropsychological Testing during Resection Phase in Left and Right Prefrontal Brain Tumors

Background: Executive functions are multi-component and are based on large-scale brain networks. For patients undergoing brain surgery in the prefrontal cortex, resection in the anterior prefrontal sites is assisted by continuous monitoring of their performance on several tasks measuring components of executive functions. In this study, we did not test patients during direct cortical stimulation, but during resection itself. We chose tests routinely used to assess executive functions and included them in a protocol for left (LH) and right (RH) hemisphere prefrontal resections. This protocol is meant to be used during real-time neuropsychological testing (RTNT)-an already established monitoring technique. Methods: We retrospectively reviewed a consecutive series of 29 adult patients with glioma in the superior and middle frontal areas who performed the RTNT sequence throughout the resection phase. The testing protocol comprised 10 tests for LH frontal resections and 9 tests for RH frontal resections. Results: RH patients showed a median performance on RTNT with significantly lower scores for visuo-spatial attention and emotion processing (95% Confidence Interval Lower bound of 66.55 and 82.57, respectively, chi 2 (7) = 32.8, p < 0.001). LH patients showed a median performance on RTNT, with significantly lower scores for selective attention and working memory (95% Confidence Interval Lower bound of 51.12, chi 2 (5) = 20.31 p < 0.001) and minimum scores for the same task and for the Stroop test (chi 2 (5) = 17.86, p < 0.005). The delta for accuracy between the first and the last RTNT run was not statistically significant (RH patients: chi 2 (7) = 10.49, p > 0.05, n.s.; LH patients: chi 2 (5) = 3.35, p > 0.05, n.s.). Mean extent of resection was 95.33% +/- 9.72 for the RH group and 94.64% +/- 6.74 for the LH group. Patients showed good performance post- vs. pre-surgery. The greater difference in the number of LH patients scoring within the normal range was found for the symbol-digit modality test (83.3% to 62%), Stroop test (100% to 77%) and short-term memory (84.61% to 72.72%) and working memory (92.3% to 63.63%). For RH patients, the main changes were observed on the clock drawing test (100% to 77.7%) and cognitive estimation (100% to 72.7%). Conclusions: Frontal RTNT offers continuous and reliable feedback on the patients' cognitive status during resection in frontal areas.