Cerebral hemodynamics as biomarkers for neuropathic pain in rats: a longitudinal study using a spinal nerve ligation model.

Abstract

Abstract: Neuropathic pain is one of the most challenging types of pain to diagnose and treat, a problem exacerbated by the lack of a quantitative biomarker. Recently, several clinical and preclinical studies have shown that neuropathic pain induces cerebral hemodynamic changes as a result of neuroplasticity in the brain. Our hypothesis in this study is that neuropathic pain leads to cerebral hemodynamic changes over postoperative time in a spinal nerve ligation (SNL) rat model, which has not been longitudinally explored previously. Furthermore, by identifying multiple regional hemodynamic features that are the most distinct between SNL and sham groups, where the sham group underwent only an incision without SNL, it may be possible to classify the SNL group regardless of when the onset of pain occurs. We investigate cerebral hemodynamic changes using dynamic susceptibility contrast magnetic resonance imaging in a rat model up to 28 days after ligating L5/L6 spinal nerves. We trained a linear support vector machine with relative cerebral blood volume data from different brain regions and found that the prediction model trained on the nucleus accumbens, motor cortex, pretectal area, and thalamus classified the SNL group and sham group at a 79.27% balanced accuracy, regardless of when the onset of pain occurred (SNL/sham: 60/45 data points). From the use of the SNL model without prior knowledge of the onset time of pain, the current findings highlight the potential of relative cerebral blood volume in the 4 highlighted brain regions as a biomarker for neuropathic pain.