Brainstem processing of peripheral punctate stimuli in patients with and without chemotherapy-induced peripheral neuropathy: a prospective cohort functional MRI study

Background: Chemotherapy-induced peripheral neuropathy affects 30% of cancer survivors. Findings from animal studies suggest that brainstem descending inhibitory pathways are important in chronic neuropathic pain. An aberrant descending pain modulation system has been implicated in human neuropathic pain. Whether aberrant descending pain modulation before chemotherapy is associated with development of chemotherapy-induced peripheral neuropathy is unclear. We aimed to assess descending pain modulation systems using functional MRI (fMRI) in chemotherapy-naive patients with cancer to determine whether differences are associated with subsequent development of the neuropathy.

Methods: In this multicentre prospective cohort study, adult patients with cancer and no chronic pain, neuropathy, or risk factors for neuropathy were recruited from the oncology clinic before onset of chemotherapy. After patients had given written informed consent, descending inhibitory pathways were challenged (jittered punctate stimuli 256 mN Somedic von Frey filament) during a 3T fMRI scan, and images analysed with FSL software. Sample size was based on published fMRI estimates. Chemotherapy-induced peripheral neuropathy was diagnosed with the CIPN20 questionnaire.

Findings: 30 patients were recruited (mean age 60·4 years [SD 7·9]). We report a preliminary analysis of the first 12 patients (60·6 [8·3], six women); six had colorectal cancer, four gynaecological cancer, and two lung cancer. Seven patients (three men, four women) developed chemotherapy-induced peripheral neuropathy. After data brain extraction, registration, B0 unwarping, and motion correction, FEAT was used for first and second level analysis. Mean group level comparisons between patients with and without the neuropathy were conducted with mixed-effects analysis (z threshold 2·3, regions considered significant at p<0·05, cluster uncorrected for preliminary analysis) and adjusted for sex, age, and cancer type. Patients with chemotherapy-induced peripheral neuropathy had increased activation in the nucleus cuneiformis and primary somatosensory cortex compared with patients who did not develop the disorder.

Interpretation: These preliminary results suggest that baseline differences exist, before peripheral nerve injury, in the descending pain modulation system of patients who go on to develop chemotherapy-induced peripheral neuropathy. These differences might aid development of biomarkers to guide chemotherapy choices. Limitations of the study include the small sample size for the present analysis, the observational nature of the study, and the possibility of unknown confounders. Strengths include the prospective design in a unique patient cohort and the high sensitivity of fMRI.