Conditioned pain modulation predicts duloxetine efficacy in painful diabetic neuropathy

This study aims to individualize the selection of drugs for neuropathic pain by examining the potential coupling of a given drug's mechanism of action with the patient's pain modulation pattern. The latter is assessed by the conditioned pain modulation (CPM) and temporal summation (TS) protocols. We hypothesized that patients with a malfunctioning pain modulation pattern, such as less efficient CPM, would benefit more from drugs augmenting descending inhibitory pain control than would patients with a normal modulation pattern of efficient CPM. Thirty patients with painful diabetic neuropathy received 1 week of placebo, 1 week of 30 mg/d duloxetine, and 4 weeks of 60 mg/d duloxetine. Pain modulation was assessed psychophysically, both before and at the end of treatment. Patient assessment of drug efficacy, assessed weekly, was the study's primary outcome. Baseline CPM was found to be correlated with duloxetine efficacy (r = 0.628, P < .001, efficient CPM is marked negative), such that less efficient CPM predicted efficacious use of duloxetine. Regression analysis (R2 = 0.673; P = .012) showed that drug efficacy was predicted only by CPM (P = .001) and not by pretreatment pain levels, neuropathy severity, depression level, or patient assessment of improvement by placebo. Furthermore, beyond its predictive value, the treatment-induced improvement in CPM was correlated with drug efficacy (r = −0.411, P = .033). However, this improvement occurred only in patients with less efficient CPM (16.8 ± 16.0 to −1.1 ± 15.5, P < .050). No predictive role was found for TS. In conclusion, the coupling of CPM and duloxetine efficacy highlights the importance of pain pathophysiology in the clinical decision-making process. This evaluative approach promotes personalized pain therapy.