McNichol, E., Strassels, S.A., Goudas, L., Lau, J., & Carr, D.B. (2005). NSAIDs or paracetamol, alone or combined with opioids, for cancer pain. Cochrane Database of Systematic Reviews 2005, Issue 2. Art. No.: CD005180.

• Databases searched were Cochrane Central Register of Controlled Trials (CENTRAL) (2nd quarter 2002), MEDLINE (January 1966–March 2003), EMBASE (January 1980–2nd quarter 2002), LILACS (January 1984–December 2001). Investigators also searched the reference lists of relevant publications.
• Search keywords were neoplasm, anti-inflammatory drugs, nonsteroidal, and pain.
• Studies were included in the review if they were published in any language.
• Studies were excluded if they involved animals.

The search retrieved 42 trials. Eight trials compared an NSAID with a placebo, 13 compared one NSAID with another, 23 compared an NSAID with an opioid or an NSAID-opioid combination, and 9 assessed the effect of increasing the NSAID dose. Sixteen of the 42 studies were conducted over seven days or longer, and 11 studies were single-dose studies. No study lasted longer than 12 weeks. Treatments studied included weak opioids, strong opioids, and agonist or antagonists. Nine studies examined dose range. Outcomes measured included differences in pain intensity, pain relief, and incidence and severity of adverse effects. Many studies used a visual analog scale; few studies used validated scales.

The 42 trials reviewed included 3,084 patients.

On the basis of limited data, NSAIDs appear to be more effective than placebo in treating cancer pain. Data to support the use of one NSAID over another, in regard to safety or efficacy, are insufficient. Compared to an NSAID alone or an opiod alone, NSAID-opioid combinations were no more effective or, at most, had a slight statistical advantage. The World Health Organization (WHO) method of cancer pain relief is considered the gold standard. Regarding the first step in the WHO method, the management of mild pain, results of this review support the WHO recommendations: Results strongly suggest that an NSAID alone is superior to placebo and adequate for at least short-term pain relief. Regarding the second step: Evidence is insufficient to refute or support the WHO recommendation to use an NSAID-opioid combination as a means to manage moderate cancer pain. Increasing the dose of an NSAID or adjuvant drug to the maximum acceptable dose may be the better course.

• Heterogeneity precluded meta-analysis.
• Short duration of studies undermines generalizability.
• Studies were too short to demonstrate the safety or efficacy of long-term NSAID use.

Clinicians should be cautious regarding the use of NSAIDs, especially in the population studied. This population takes many prescription drugs, some of which may increase NSAID-related toxicity.