Brain correlates of chemotherapy-induced peripheral neuropathy (CIPN) during somatosensory and interoceptive attention: A functional MRI (fMRI) study.

Abstract

e24095 Background: Over half of patients receiving neurotoxic taxane, platinum, or bortezomib chemotherapy experience CIPN—a dose-limiting toxicity involving numbness and pain in the extremities. There are limited biomarkers and treatments for CIPN partly due to lack of knowledge of its pathophysiology. Here, we studied whether CIPN involves changes in the brain’s sensory processing circuitry to inform novel brain-based biomarkers and brain-based treatments for CIPN. Methods: Eleven patients scheduled to receive taxane, platinum, or bortezomib rated CIPN severity (CIPN-20) and completed an fMRI scan before starting neurotoxic chemotherapy and/or 12 weeks later. The 7-min fMRI task involved alternating 20-sec trials of distinct attentional focus: (1) their left fingertips (somatosensory attention), (2) their heart (interoceptive attention), and (3) a flashing word “target” (visual attention; the reference condition). We found an activation cluster in primary somatosensory cortex (S1) comparing somatosensory vs. visual attention and another cluster in primary interoceptive cortex (mid/posterior insula) comparing interoceptive vs. visual attention. We tested for associations between CIPN severity and task-based activations in S1 and the insula. Because this is a pilot study, we present effect sizes (ESs) using Cohen’s d and Pearson’s r, not p-values. Results: The 11 patients were 63±12 years old, 54% women, and 54% had breast cancer with high-quality fMRI data (16 scans; 8 pre-chemotherapy). As expected, somatosensory attention increased activation in S1 compared to visual attention (ES = 1.46), and interoceptive attention increased activation in the insula compared to visual attention (ES = 0.94). Patients with worse CIPN had smaller increases in S1 activation during somatosensory vs. visual attention (ES = 0.38), suggesting reduced S1 activation during somatosensory attention (i.e., failure to activate S1) and/or increased S1 activation during visual attention (i.e., S1 chronically activated). Similarly, patients with worse CIPN had smaller increases in insula activation during interoceptive vs. visual attention (ES = 0.57). Individual differences in S1 or insula activity were not significantly explained by age, anxiety, or problems remembering (all ESs < 0.12). Conclusions: CIPN severity was associated with changes in brain activity during a simple attentional task during fMRI scanning, implicating the brain’s somatosensory and interoceptive processing in CIPN. Future work should test for replication and develop biomarkers and treatments for CIPN that target somatosensory and interoceptive processing, perhaps using this attention task. NCT03021174. Funding: NIH R25CA102618, UG1CA189961, K07CA221931. Clinical trial information: NCT03021174.