Chronic low back pain (CLBP) has been linked to maladaptive cortical plasticity in sensorimotor regions, which may contribute to pain persistence through less distinct ("smudged") somatotopic maps of afferent input from the back. However, empirical support for such functional reorganization, especially in the primary somatosensory cortex (S1), is limited. To delineate precise sensory representations of the back in humans and test for altered somatotopic processing in CLBP, we utilized an MR-compatible pneumatic vibration device and applied frequency-specific vibrotactile stimuli at nine thoracolumbar paraspinal sites in 45 patients with CLBP and 41 healthy controls (53 females in total). Representational similarity analysis was employed via a whole-brain searchlight method, and the neural representational patterns were contrasted against theoretical models including a segmental (based on anatomical proximity), simple (upper vs. lower back division) and a random model. In addition, a machine learning classifier was trained to predict afferent input from the upper vs lower back in healthy controls and tested in CLBP patients. Both groups displayed well-organized segmental representations spanning somatosensory, motor, and posterior parietal cortices. Posterior parietal regions exhibited the best model fits, followed by S1. No evidence for differences in representational patterns between groups was found. In CLBP patients, these patterns did not show associations with pain duration, severity and self-reported back perception. The classifier, trained on healthy controls, accurately predicted upper vs lower back afferent input in CLBP patients. These findings indicate preserved cortical maps of the back in CLBP, challenging the hypothesis of sensory cortical reorganization in this condition. Significance Statement Chronic low back pain (CLBP) is often linked to maladaptive reorganization of somatotopic maps in sensorimotor cortices, providing the mechanistic rationale for therapies like sensory discrimination training to “normalize” these maps and alleviate pain. Using frequency-specific vibrotactile stimulation of the back and multivariate fMRI analysis, we found well-organized, segmental cortical representations in both CLBP patients and healthy controls, with no evidence of group differences. Representational patterns showed no associations with clinical measures such as pain duration or intensity. These findings challenge the cortical reorganization hypothesis in CLBP and suggest sensory discrimination training benefits may involve alternative mechanisms.
Dörig et al. (Fri,) studied this question.