The plenary sessions at the International Association for the Study of Pain’s 15th World Congress turned from research to clinical considerations with a talk from Nadine Attal, a neurologist and pain specialist at the Hôpital Ambroise Paré, Boulogne, France. Attal gave attendees an early look at proposed new guidelines for the treatment of neuropathic pain, based on an extensive review of current evidence by the IASP Neuropathic Pain Special Interest Group (NeuPSIG). In addition, she previewed a future where neuropathic pain patients will no longer be lumped together by etiology and treated for generic pain, but rather will be sorted by trans-etiological symptoms or signs and treated with rational therapies based on the underlying mechanisms of pain.
Neuropathic pain results from damage to the sensory nervous system and occurs in many diseases, from diabetic peripheral neuropathy to spinal cord injury and stroke. It is a major problem, affecting up to 7 percent of the population, and the hallmarks of spontaneous pain—allodynia and hyperalgesia—are among the most troubling symptoms for pain patients.
What is available to treat neuropathic pain? Right now, the major classes of drugs used are antidepressants, which seem to enhance inhibitory pain modulation systems; antiepileptics to decrease neuronal hyperexcitability in the spinal cord; and opioids for central neuropathic pain. Topical pain treatments such as the lidocaine patch or capsaicin patch are also available, and a new addition to the arsenal is botulinum toxin A.
Over the years, several clinical guidelines have been produced based on systematic reviews and meta-analyses of existing clinical trials for these drugs. But Attal said it was time for an update—the most recent guidelines are four or more years old (e.g., see Attal et al., 2010). Since then, there have been several large trials, new treatments, and advances in techniques for evaluating evidence, all of which prompted the new effort.
The new analysis was based on an extensive systematic review of clinical data using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) methodology that considers the quality of the evidence from individual trials rather than lumping multiple trial outcomes together in a traditional meta-analysis. Importantly, the analysis included data from several unpublished trials. A total of 229 records were analyzed for effect size, tolerability and safety, and patient preferences, among other parameters.
The new recommendations suggest the currently used antiepileptics and antidepressants as first-line treatments. Second-line treatments include combinations of the first-line drugs, plus the opioid tramadol, and topical lidocaine or capsaicin. Third-line treatments include other opioids and, in some limited cases of peripheral neuropathic pain, botulinum toxin A. Overall, the group found not much evidence for the effects of cannabinoids.
Compared to earlier recommendations, the new guidelines downgrade the use of opioids other than tramadol. Topical capsaicin and botulinum toxin are new to the list. The complete data set has been submitted for publication.
The inclusion of unpublished data is an important feature of the new study. There is a strong bias to publish positive trial data, which may lead to an overestimation of the effects of drugs. In the case of pregabalin, a previous meta-analysis put the number needed to treat (NNT, number of patients who would have to be treated to have one patient experience a 50 percent reduction in pain) at between four and six—a moderate result compared to placebo (Finnerup et al., 2010). But the NNT moves closer to eight in the new analysis, when results of four unpublished trials (three of which were negative) were included. And a cumulative meta-analysis of effect size for trials between 1998 and 2013 showed a steadily declining effect size compared to placebo: in two original trials in 1998, the effect size was 1.6; now it sits at 0.7 when high-quality trials and unpublished data are included.
Why might this be? One suggestion is that the placebo effect is getting stronger over time. Trial design can affect the magnitude of the placebo effect, but so far the idea that placebo is higher in recent neuropathic pain trials remains to be confirmed. Attal said she and her colleagues are now working on an analysis to address this question.
Variations in neuropathic pain diagnostic criteria between groups might also contribute. The increasing use of standardized diagnostic algorithms and screening tools in the field is good news, Attal said.
More good news, Attal added, is that the field is seeing a large consensus for the subgrouping of patients based on sensory phenotypes defined either by detailed symptom questionnaires or by quantitative sensory testing (e.g., see Baron et al., 2012; von Hehn et al., 2012). This kind of analysis could help clarify outcomes in clinical trials and allow more personalized medicine. For example, Attal and colleagues are known for developing the Neuropathic Pain Symptom Inventory (NPSI), a questionnaire used to collect detailed symptom profiles. In a recent trial of a chemokine receptor antagonist, although they found no decrease in overall pain scores, there were suggestions of selective changes in several NPSI symptoms of paroxysmal pain and paresthesia/dysesthesia (Kalliomäki et al., 2013), suggesting that individual symptoms could be more sensitive to detecting drug effects.
Attal highlighted another study where she and her colleagues found a correlation between a quantitative sensory measure of thermal sensing and pain reduction by botulinum toxin A in patients with peripheral neuropathy: patients who had poorer heat sensing showed less pain reduction (Ranoux et al., 2008). Attal reported they recently terminated a larger study in which botulinum toxin treatment was effective in patients who showed allodynia, but not in those without. In that study, patients with the best thermal pain perception also responded best to botulinum toxin. The results suggest that the response to botulinum toxin requires patients to have preserved nociceptive function, a result that has been reported for some other drugs, too.
Attal reviewed the growing number of examples like this, where sensory measures or symptom profiles have helped or promised to help pinpoint patients who will respond to different treatments. This kind of analysis, though complex, will help improve trials and treatment, she concluded, by creating a more rational algorithm for neuropathic pain.