Cognitive Functions in Repeated Glioma Surgery

Abstract

Low-grade gliomas (LGG) are slow-growing brain tumors infiltrating the central nervous system which tend to recur, often with malignant degeneration after primary treatment. Re-operations are not always recommended due to an assumed higher risk of neurological and cognitive deficits. However, this assumption is relatively ungrounded due to a lack of extensive neuropsychological testing. We retrospectively examined a series of 40 patients with recurrent glioma in eloquent areas of the left hemisphere, who all completed comprehensive pre- (T3) and post-surgical (T4) neuropsychological assessments after a second surgery (4-month follow up). The lesions were most frequent in the left insular cortex and the inferior frontal gyrus. Among this series, in 17 patients the cognitive outcomes were compared before the first surgery (T1), 4 months after the first surgery (T2), and at T3 and T4. There was no significant difference either in the number of patients scoring within the normal range between T3 and T4, or in their level of performance. Further addressing the T1-T4 evolution, there was no significant difference in the number of patients scoring within the normal range. As to their level of performance, the only significant change was in phonological fluency. This longitudinal follow-up study showed that repeated glioma surgery is possible without major damage to cognitive functions in the short-term period (4 months) after surgery.

Keywords: awake surgery; diffuse low surgery; neurocognition; neuropsychological assessment.

Figure 1

Overlap of all the 40 pre-recurrence (T3) patients’ lesions masks. Lesion masks include both the surgical cavity of the first surgery and the lesion regrowth. The number of overlapping lesions is illustrated by different colors coding increasing frequencies (as indicated in the bar code). As shown by the image, the epicenter of the overlap is located at the level of the frontotemporoinsular region. MR images are displayed in radiological convention.

Figure 2

Axial view T1-weighted MRI. Upper image shows frontotemporoinsular glioma before first surgery. (a) Lower images show recurrence before second surgery. (b) Recurrent neoplastic tissue was located both along the walls of and within the previous-tumor cavity.

Figure 3

Performance of the 40 patients with recurrent glioma at T3 and T4 (pre- and post-recurrence surgery, respectively). Red lines indicate the cut-off, below which the patient results were pathological: verbal short-term memory (<4.26 is equal to an equivalent score of 0), working memory (<2.65 is equal to an equivalent score of 0), ideomotor limb apraxia (<53) [25] and oral apraxia (<16) [26], language comprehension (<26.25 is equal to an equivalent score of 0) [27], object naming and action naming (<28 and <26 respectively) [28], and verbal fluency (<16 is equal to an equivalent score of 0) [29]. Patients showing identical scores are represented by overlapping dots.

Figure 4

Mean performance of the 40 patients with recurrent glioma T3 and T4 (pre- and post-recurrence surgery, respectively). Bars represent the standard deviations.

Figure 5

Performance of the 17 patients with recurrent glioma at T1 and T2 (pre- and post-first surgery, respectively) and at T3 and T4 (pre- and post-recurrence surgery, respectively). Red lines indicate the cut-off, below which the patient results are pathological (<4.26 is equal to an equivalent score of 0) [30,31], working memory (<2.65 is equal to an equivalent score of 0) [30], ideomotor limb apraxia (<53) [25] and oral apraxia (<16) [26], language comprehension (<26.25 is equal to an equivalent score of 0) [27], object naming and action naming (<28 and <26 respectively) [28] and verbal fluency (<16 is equal to an equivalent score of 0) [29]. Patients showing identical scores are represented by overlapping dots.

Figure 6

Mean performance of the 17 patients with recurrent glioma T1 and T2 (pre- and post-first surgery, respectively) and T3 and T4 (pre- and post-recurrence surgery, respectively). Bars represent the standard deviations.