Anti-tumor necrosis factor antagonists in the treatment of low back pain and radiculopathy: a systematic review and meta-analysis

Low back pain, with or without radiculopathy, is an important cause of disability and economic expenditure. However, many patients are not achieving optimal pain control with existing medications. Tumor necrosis factor antagonists (anti-TNFα) could be an alternative drug treatment. Systematic review the efficacy and safety of anti-TNFα in the treatment of low back pain with or without radiculopathy. Inclusion criteria were observational studies with safety as an outcome, and randomized or nonrandomized controlled trial (RCT) studies on efficacy and/or safety of anti-TNFα drugs on low back pain. Exclusion criteria included patients with auto-immune conditions or osteoporosis. Studies were assessed independently by 2 authors regarding inclusion/exclusion criteria, risk of bias, clinical relevance, quality, and strength of evidence (GRADE approach). Of the 1,179 studies retrieved, all duplicates were excluded and then the inclusion/exclusion criteria was applied. One observational study (n = 143) and 11 RCTs remained (n = 539): 8 for etanercept (n = 304), one for adalimumab (n = 61), one for adalimumab and etanercept (n = 60), one for infliximab (n = 40) and one for REN-1654 (n = 74). Only 3 etanercept and 2 adalimumab studies showed statistically significant pain relief when compared to placebo. There was no difference in the overall incidence of adverse effects when comparing anti-TNF-α and placebo. Despite the statistically significant effect, this meta-analysis has important limitations, such as high heterogeneity and high use of outcome imputation. There is low evidence that epidural etanercept has a low-to-moderate effect size when compared to placebo for pain due to discogenic lumbar radiculopathy (5 studies, n=185), with a standardized mean difference = -0.43 (95% confidence interval [CI] -0.84 to -0.02).There is moderate evidence that epidural etanercept does not have a higher adverse effects incidence rate when compared to placebo for discogenic lumbar radiculopathy (5 studies, n = 185) with a relative risk (RR) = 0.84 (95% CI 0.53 to 1.34).There is moderate evidence that anti-TNFα does not have a higher adverse effects incidence rate when compared to placebo for low back pain (10 studies, n= 343) with an RR = 0.93 (95% CI 0.56 to 1.55).We strongly suggest that anti-TNFα continue to be studied in experimental settings for the treatment of low back pain. We cannot currently recommend this therapy in clinical practice. New research could shed some lig