Pain phenotype as a predictor for drug response in painful polyneuropathy: A retrospective analysis of data from controlled clinical trials

The drugs available for treatment of neuropathic pain have a somewhat disappointing efficacy with many patients left with limited or no effect. Individualized treatment based on phenotype according to presumed underlying pain mechanism(s) has been proposed to improve outcome. We report a retrospective analysis of phenotype-specific effects of several neuropathic pain drugs, which were studied in a series of cross-over, placebo-controlled, clinical trials. The data originate from 7 trials with similar design and outcome recordings, which all had a thorough baseline registration of symptoms, signs and quantitative sensory testing. The latter was used to phenotype patients into subgroups reflecting presumed pain mechanisms. There were a total of 361 patient records distributed over treatments with 4 antidepressants and 4 anticonvulsants. Five of the drugs reduced total pain significantly compared to placebo. Only a few phenotype-specific differences in total pain reduction were found within the investigated drugs. Thus, imipramine reduced total pain 1.84 (CI: 0.02 to 3.67) and pregabalin 0.81 (CI: -0.67 to 2.29) in patients with than without gain of sensory function. Pregabalin showed a better effect in patients with preserved large fiber function with a mean difference in total pain reduction 1.31 (CI: 0.15 to 2.47). No phenotype-specific effects were found for venlafaxine, escitalopram, oxcarbazepine, valproic acid, levetiracetam or St. john's wort. Thus, this post-hoc analysis of 8 drugs with mainly non-selective actions on neuropathic pain mechanisms found limited usefulness of sensory phenotyping in pain as the basis for individualized treatment.