Predictors of placebo response in peripheral neuropathic pain: insights from pregabalin clinical trials

Abstract

Background: Greater understanding of factors associated with the high placebo-response rates noted in recent neuropathic pain trials may improve trial design. This study investigated placebo response and its predictors in pregabalin trials in patients with diabetic peripheral neuropathy (DPN) or postherpetic neuralgia.

Patients and methods: Individual patient data from 16 randomized, placebo-controlled, double-blind trials of pregabalin in 3,053 patients with DPN and 1,460 patients with postherpetic neuralgia were pooled (by condition and all together) in order to investigate the placebo response and its predictors. Univariate and multivariate analyses were performed across all 16 trials to identify predictors of change in pain score in patients. Trials with a >2-point mean reduction in pain score at endpoint with placebo were designated high placebo response and were compared with low placebo-response trials (those with a ≤2-point mean reduction) with respect to patient and study characteristics.

Results: Three high placebo-response studies were identified, with all in DPN patients and all conducted postapproval of pregabalin. Younger age, higher mean baseline pain score, longer study duration, higher ratio of patients on active treatment to placebo, and study conducted postapproval were all significantly associated with a higher placebo response (P<0.05). There was a trend towards an increased placebo response in all studies over time without any corresponding change in the response to pregabalin.

Conclusion: Consideration of the factors identified here as contributing to a higher placebo response could help improve the sensitivity and accuracy of clinical trials in patients with neuropathic pain.

Keywords: diabetic peripheral neuropathy; postherpetic neuralgia.

Figure 1

Mean change in placebo response at endpoint, by study week and by study completion date. Notes: Ratio of active treatment to placebo in each trial (total number of patients on active treatment to placebo) for (A) diabetic peripheral neuropathy (DPN) and (B) postherpetic neuralgia (PHN) studies; The solid vertical line separates those studies that commenced preapproval of pregabalin (to the left of the line) and postapproval of pregabalin (to the right of the line); The dashed horizontal line shows the 2-point mean change in pain score at endpoint; Studies with a >2-point change (A0081030, A0081071, A0081081 [DPN patients]) were designated high placebo-response studies; *Indicates trials that were negative (those in which there was no significant difference at endpoint between pregabalin and placebo); The least squares mean change in placebo response by study week for (C) DPN and (D) PHN studies; Study characteristics are shown by color of the line (ratio of active treatment to placebo) and solid or dashed lines (studies conducted pre- and postapproval, respectively); The position of the circles shown with study numbers indicates the last observation carried forward (LOCF) mean change for that study.

Figure 2

Mean change in placebo response at endpoint by study completion date. Notes: Least squares mean change in pain score from baseline at endpoint for all treatment arms by the year the study was completed for (A) diabetic peripheral neuropathy (DPN) and (B) postherpetic neuralgia (PHN) trials.

Figure 3

Regression-tree analysis of placebo response. Notes: Box-and-whiskers plots showing median change in pain score with interquartile range; The first separation is on the status of the patients; Those who discontinued from the study due to lack of efficacy had the smallest placebo response; Of those who did not discontinue from the study due to lack of efficacy, patients who were from studies conducted preapproval of pregabalin had a lower placebo response; Of those from studies conducted postapproval of pregabalin, patients who had a baseline pain score <5.54 had a lower placebo response; Of those from studies conducted preapproval of pregabalin, those from postherpetic neuralgia (PHN) studies had a lower placebo response; Of those from diabetic peripheral neuropathy (DPN) studies, patients from studies with <36 sites had a lower placebo response; Of those from studies with ≥36 sites, patients who remained in the study for <83.97% of its duration had a lower placebo response; The highest placebo response was in patients who did not discontinue from the study due to lack of efficacy, were in studies conducted postapproval, were patients with DPN, were in studies with ≥36 sites, and those who remained in the study for ≥83.97% of its duration.