Background: NSAIDs are widely applied to treat cancer pain and are frequently combined with opioids in combination preparations for this purpose. However, it is unclear which agent is most clinically efficacious for relieving cancer-related pain, or even what may be the additional benefit of combining an NSAID with an opioid in this setting.
Objectives: To assess the effects of NSAIDs, alone or combined with opioids, for the treatment of cancer pain.
Search strategy: We searched the Cochrane Central Register of Controlled Trials (Issue 2, 2002), MEDLINE (January 1966 to March 2003), EMBASE (January 1980 to December 2001), LILACS (January 1984 to December 2001) and reference list of articles.
Selection criteria: Randomized controlled trials and controlled clinical trials that compared NSAID versus placebo; NSAID versus NSAID; NSAID versus NSAID plus opioid; opioid versus opioid plus NSAID; or NSAID versus opioid.
Data collection and analysis: Two reviewers independently assessed trial quality and extracted data. Study authors were contacted for additional information. Adverse event information was collected from trials. Where there was disagreement between reviewers, the opinion of an additional reviewer was sought to resolve the issue.
Main results: Forty-two trials involving 3084 patients were included. Clinical heterogeneity of study methods and outcomes precluded meta-analyses and only supported a qualitative systematic review. Seven of eight papers that compared NSAID with placebo demonstrated superior efficacy of NSAID with no difference in side effects. Thirteen papers compared one NSAID with another; four reported increased efficacy of one NSAID over another. Four different studies found that one NSAID had fewer side effects than one or more others. Twenty-three studies compared NSAIDs and opioids in combination or alone with NSAID/opioid combinations. Thirteen out of 14 studies found no difference, or low clinical difference, when combining an NSAID plus an opioid versus either drug alone. Comparisons between various NSAID/opioid combinations were inconclusive. Nine studies assessed the association between dose and efficacy and safety. Four papers demonstrated increased efficacy with increased dose, but no dose-dependent increase in side effects within the dose ranges studied. Study duration ranged from single dose studies performed over six hours to crossover studies lasting six weeks; however the majority of studies were of less than seven days duration.
Authors' conclusions: Based upon limited data, NSAIDs appear to be more effective than placebo for cancer pain; clear evidence to support superior safety or efficacy of one NSAID over another is lacking; and trials of combinations of an NSAID with an opioid have disclosed either no difference (4 out of 14 papers), a statistically insignificant trend towards superiority (1 out of 14 papers), or at most a slight but statistically significant advantage (9 out of 14 papers), compared with either single entity. The short duration of studies undermines generalization of their findings on efficacy and safety of NSAIDs for cancer pain.