Brain network disruption predicts memory and attention deficits after surgical resection of glioma

Surgical resection with adjuvant chemotherapy and radiotherapy are effective treatments to delay brain tumour progression and improve survival. Nevertheless, a large proportion of patients have treatment-induced cognitive deficits that dramatically reduce their life quality. A major problem in basic and clinical neuroscience is that the dispersed and interlocking nature of cognitive circuits makes predicting functional impairments challenging. Here we investigated tumour interactions with brain networks in relation to cognitive recovery after surgical resection and during chemo-radiotherapy treatment. 17 patients with low- and high-grade glioma (aged 22-56 years) were longitudinally MRI-scanned and cognitively assessed using a tablet-based screening tool before and after surgery, and during a 12-months recovery period. Using structural MRI and Neurite Orientation Dispersion and Density Imaging (NODDI) derived from diffusion-weighted images, we respectively estimated tumour overlap and Neurite Density (as an in-vivo proxy measure of axon and dendrite concentration) with brain networks and functional maps derived from normative data in healthy participants. We found that neither total lesion volume nor tumour location based on traditional lobular divisions were associated with memory or attention deficits. However, tumour and lesion overlap with the Default Mode Network (DMN), Attention Network and attention-related regions located in frontal and parietal cortex was associated with memory and attention deficits. This association was above and beyond the contributions of preoperative cognitive status and tumour volume (Linear Mixed Model, Pfdr