Brain mechanisms of pain anticipation in chronic low back pain: Insights into phenotypical responses to mind-body therapy.

Pain perception and emotional processing share common predictive mechanisms influenced by anticipation, expectation, and uncertainty. This study identifies distinct neural phenotypes of pain anticipation in chronic low back pain(cLBP) and examines their clinical and physiological correlates. In 43 cLBP patients, we acquired task-based functional MRI during certain(high/low) and uncertain pain anticipation, quantitative sensory testing, and validated questionnaires at baseline. A subset of patients(n=29) completed the same assessments post-treatment following an 8-week smartphone-based interoceptive attention training. Machine learning classified patients into distinct anticipatory brain phenotypes based on task-based fMRI responses. The study identified two complementary neural phenotypes. The first-the 'How' phenotype-distinguished individuals based on neural mechanisms during certain pain anticipation: those relying on dorsal anterior insula(cognitive-evaluative processing) versus ventral anterior insula(affective-reactive processing). The second-the 'What' phenotype-distinguished individuals based on predictive bias during uncertainty: optimistic, pessimistic, or unbiased pain anticipation. These phenotypes showed significant convergence: individuals using cognitive-evaluative mechanisms predominantly exhibited predictive biases(either optimistic or pessimistic), while those using affective-reactive mechanisms remained unbiased. The cognitive-evaluative group showed enhanced endogenous pain inhibition, stronger brainstem connectivity, and greater reliance on distraction-based coping. Conversely, the affective-reactive group showed enhanced pain facilitation, stronger amygdala connectivity, and less distraction use. Notably, the cognitive-evaluative group demonstrated superior response to an 8-week smartphone-based interoceptive attention training. Our findings reveal distinct anticipatory processing phenotypes that predict treatment response to mind-body therapies. These mechanistic insights identify specific neural phenotypes likely to benefit from interoceptive interventions, advancing precision medicine approaches for chronic pain management. PERSPECTIVE: Chronic low back pain patients exhibit distinct neural phenotypes during pain anticipation: cognitive-evaluative versus affective-reactive processing. These phenotypes converge with prediction biases during uncertainty and predict differential treatment responses to interoceptive training. Brain-based phenotyping enables precision medicine approaches by identifying which patients benefit most from attention-based mind-body interventions.