Impact of Data Imputation Methodology on Pain Assessment over 24 Hours in a Randomized, Placebo-Controlled Study of Gabapentin Enacarbil in Patients with Neuropathic Pain Associated with Postherpetic Neuralgia

Abstract

Objective: To assess the impact of gabapentin enacarbil on primary and secondary pain endpoints using three data imputation methodologies in a randomized phase II study of adult patients with postherpetic neuralgia.

Methods: The primary endpoint was change from baseline to end of maintenance treatment in mean 24-hour average pain intensity score. Secondary endpoints (daytime/nighttime average pain intensity score, daytime/nighttime current pain intensity score, and daytime/nighttime worst pain intensity score) were based on daily electronic diary assessments. Comparisons of each gabapentin enacarbil dose with placebo were performed using three different statistical methodologies: last observation carried forward, baseline observation carried forward, and mixed-effect model for repeated measures.

Results: Of the 376 randomized patients, 371 were in the intent-to-treat population (gabapentin enacarbil 1,200 mg, 107; 2,400 mg, 82; 3,600 mg, 87; placebo, 95). For mean 24-hour average pain intensity score, there were statistically significant improvements from baseline to end of maintenance treatment for all gabapentin enacarbil groups vs placebo using the three analysis methods. Significant improvements were also observed for all secondary endpoints with gabapentin enacarbil 1,200 mg using the three analysis methods. Most secondary endpoints also showed improvements following treatment with gabapentin enacarbil 2,400 mg or 3,600 mg compared with placebo.

Conclusions: Gabapentin enacarbil (1,200 mg, 2,400 mg, and 3,600 mg) was effective and well tolerated in patients with postherpetic neuralgia compared with placebo, as confirmed by three different and robust statistical methodologies.

Keywords: Data Imputation; Gabapentin Enacarbil; Pain; Postherpetic Neuralgia.