INTRODUCTION Pamidronate and other bisphosphonates (bisph) have an anti-nociceptive effect in
animals. In humans, IV pamidronate is analgesic in patients affected by numerous painful
conditions, including cancer bone pain, complex regional pain syndrome (CRPS), ankylosing
spondylitis, and rheumatoid arthritis. The investigators have explored the effect of IV
pamidronate in the management of chronic low back pain (CLBP), a worldwide public health
problem in terms of lost workdays, treatment costs, and suffering.
The study was a randomized, double blind, dose-escalation trial of IV pamidronate. Study
participants were divided among four study phases. Each group received IV placebo or
escalating doses of IV pamidronate.
STUDY DESIGN A phase I-II, randomized, double-blind, placebo controlled, dose finding study
for the treatment of patients with CLBP with IV pamidronate. The study was first conducted at
Beth Israel Medical Center, NY for the first two groups, and was completed at Mount Sinai
Medical Center , NY for the remaining groups.
PRIMARY OBJECTIVES The intent of this pilot study was to determine the optimal IV pamidronate
treatment protocol for CLBP in a Phase III trial. Primary study aims were safety and average
daily pain. Subjects used an electronic diary (LOGPAD) to record their daily adverse events
(AEs) and their baseline and post-treatment average pain on the 0-10 numerical rating scale
(NRS).
STATISTICAL ANALYSES All the analysis were performed on an intention-to-treat basis. Primary
outcomes are LOGPAD change in pain severity and whether a patient was a responder (pain score
dropped ≥2 points or ≥30%). Because there was no statistical difference in these two outcomes
among the 4 placebo groups, all placebo patients were combined (Comb PL) in the subsequent
analyses. Primary analytic tool was mixed model for repeated measures (MMRM), assuming
autoregressive covariance structure (LOGPAD pain and Brief Pain Inventory - BPI
interferences), and Generalized Estimating Equation (GEE) model for categorical outcomes
(Response rates and Patient Global Impression of Change - PGIC). The main objective was to
assess whether the changes in outcome from baseline or the response rates were the same
across the study phases, while adjusting for baseline values and time effect. Least square
means were contrasted. Interaction between time and study phases were also tested to see if
the treatment effect is a function of time. Fisher's exact tests or ANOVA (Kruskal-Wallis)
tests were performed for cross-sectional group comparison, including baseline.