NeuPSIG has just published an up to date systematic review on the effectiveness of pharmacotherapy in Lancet Neurology. They have negotiated with the journal to make it available beautifully open access. You can download it for free here.
Neil O'Connell, Brunel University London
This is a comprehensive review, containing 229 trials of the full range of pharmacological agents using robust methods, to synthesize, summarise and make value judgements about the quality of the available evidence. So what are the take home messages?
Using a primary outcome of achieving at least 50% pain relief trial outcomes were described as "generally modest". The number of patients needed to treat with the drug compared to a placebo for one more person to achieve this outcome ranged from a relatively rosy 3.6 (95% confidence interval 3 to 4.4) for tricyclic antidepressants such as amitryptiline, 4.3 (95%CI 3.4 to 5.80 for strong opioids to a less impressive 7.2 (95%CI 5.9 to 9.21) for gabapentin, and 7.7 (6.5 to 9.4) for pregabalin (often sold under the brand-name Lyrica). It’s interesting, at least to me, how much better the older more traditional agents seem to have fared compared on effectiveness to the more modern (and commonly more expensive) agents although the safety and tolerability of gabapentin seems superior.
The spectre of publication bias also raises its head. The reviewers carefully took a number of routes to try to unpick this notoriously difficult issue and estimate that there has been overall a 10% overstatement of treatment effects. Published studies reported larger effect sizes than did unpublished studies. This is not a problem restricted to the field of pain trials. It is a burning issue across the world of clinical trials. It is very important because if we fail to base our clinical recommendations on the totality of relevant evidence (because some data are hidden from us) we are in danger of mis-estimating the benefits and the harms and as a result patients are put at risk. If you think that is pretty important then there are ways that you can help. Check out the All-Trials campaign.
Overall what does this mean? Many drugs are effective but not as effective as we would wish them to be. No pharmacological agent really impresses and for any drug the most probable outcome is failure to produce 50% pain relief. There are various potential reasons for this. The first is that the drugs may only be moderately or marginally effective, another is that neuropathic pain includes quite a mixed bag and our ability to accurately diagnose and to target drugs to specific mechanisms in the clinic is currently fairly poor.
The NeuPSIG review team formulate a number of recommendations for revision of their clinical guideline for managing NP pain, balancing the benefits, harms, costs and strength of the evidence:
- a strong recommendation for use and proposal as a first-line treatment in neuropathic pain for tricyclic antidepressants, serotonin-noradrenaline reuptake inhibitors, pregabalin, and gabapentin;
- a weak recommendation for use and proposal as a second line treatment for lidocaine patches, capsaicin high-concentration patches, and tramadol; and a weak recommendation for use and proposal as third line for strong opioids and botulinum toxin A. Topical agents and botulinum toxin A are recommended for peripheral neuropathic pain only.
This email is also published as a blogpost at www.bodyinmind.org
Finnerup NB, Attal N, Haroutounian S et al. Pharmacotherapy for neuropathic pain in adults: a systematic review and meta-analysis. Lancet Neurol. 2015;14:2:162-73.