[Efficacy of celecoxib, diclofenac, etoricoxib, ibuprofen, ketoprofen and naproxen in low back pain with or without nerve root pain: Systematic review of literature and meta-analysis of randomized controlled trials].

OBJECTIVES

To assess the efficacy of naproxen, ibuprofen, ketoprofen, diclofenac, etoricoxib and celecoxib in low back pain with or without radiculalgia.

METHOD: Systematic review and meta-analysis of randomised controlled trials (RCTs) using the REB method. Eligibility criteria were RCTs assessing the painkiller efficacy of the aforementioned non-steroidal anti-inflammatory drugs (NSAIDs) administered orally versus placebo, in acute and chronic low back pain with or without radiculalgia. The literature search was conducted on Medline, Cochrane Central Register of Controlled Trials, and ClinicalTrial.gov until 31/12/2024. The risks of bias were assessed using the tool Risk of Bias 2 (RoB2) from the Cochrane collaboration. In addition to the qualitative analysis using the REB method, a quantitative analysis was conducted using Review Manager v5.4. The standardised mean differences (SMD) were calculated with a 95% confidence interval (95% CI). Statistical significance was met if the P-value was<0.05.

RESULTS

A total of 11 RCTs were identified: 6 concerned acute low back pain and 5 chronic low back pain. Six were at low risk of bias for pain intensity, and 4 at low risk of bias for functional capacity; 5 were at high risk of bias for pain intensity and 5 at high risk of bias for functional capacity. Three thousand seven hundred eighty-four (3784) patients were included: 2142 suffered from chronic low back pain and 1642 from acute low back pain. The REB analysis concluded "solid evidence of efficacy" for diclofenac in acute low back pain. For the ibuprofen in acute low back pain without radiculalgia, the results were "conclusive to be confirmed". For the naproxen in acute low back pain with radiculalgia, the results were "conclusive to be confirmed". For the other treatments and indications, the conclusion was "lack of evidence".

DISCUSSION: This review has several limitations: a small number of trials that evaluated NSAIDs versus placebo, which limits the interpretation of possible publication bias by the interpretation of funnel plots. More than half of the selected trials aimed to evaluate a different active ingredient or treatment, with the NSAIDs studied primarily used as comparators. Only trials published in French and English were included, which may have introduced selection bias. Furthermore, the results of the RCTs often focused on secondary rather than primary endpoints. Of the 6 NSAIDs studied, only diclofenac showed solid evidence of efficacy on pain intensity in acute low back pain. Ibuprofen and Naproxen showed conclusive evidence requiring confirmation. For the other drugs, there was no evidence of efficacy according to the REB method. For all the analyses, the magnitude of the effect was small and not clinically relevant. It is necessary to conduct new RCTs to confirm the clinical interest of NSAIDs in acute and chronic low back pain.