Once-daily OROS hydromorphone ER compared with placebo in opioid-tolerant patients with chronic low back pain

Objective: This multicenter, double-blind, placebo-controlled study using a randomized withdrawal design evaluated the efficacy and safety of once-daily OROS hydromorphone ER in the treatment of opioid-tolerant patients with chronic moderate-to-severe low back pain (LBP). Main outcome measures: The primary efficacy assessment was mean change in pain intensity based on patient diary Numeric Rating Scale (NRS) scores from baseline to final visit of the 12-week double-blind phase. Secondary endpoints included mean change from baseline to each visit in patient diary NRS scores; and office NRS scores; time to treatment failure; Patient Global Assessment; rescue medication use; and Roland Morris Disability Questionnaire total scores. Clinical Trial Registration: ClinicalTrials.gov NCT00549042. Results: For the primary outcome measure, hydromorphone ER significantly reduced pain intensity compared to placebo (p<0.001). Median diary NRS score change from baseline to endpoint was significantly lower for hydromorphone ER (0.2 units) compared to placebo (1.2 units). A significantly higher proportion of hydromorphone ER (60.6%) vs. placebo (42.9%) patients had at least a 30% reduction in diary NRS pain score from screening to endpoint (p<0.01). Hydromorphone ER was well-tolerated, although 60 (13%) discontinued during the enrichment phase for adverse events and more active (9, 6.7%) than placebo (4, 3.0%) patients discontinued treatment for adverse events during the randomized phase. Conclusions: These results provide evidence for the efficacy and safety of hydromorphone ER in opioid-tolerant patients with chronic moderate-to-severe LBP. Potential limitations include the shortened dose-conversion/titration phase, limiting the daily allowable dose of hydromorphone ER to 64 mg, and the allowance of limited rescue medication throughout the entire double-blind phase. Other trial design elements such as the use of an enrichment phase and the inclusion of only opioid tolerant patients may limit the generalizabilty of these results.