Lack of assay sensitivity is a problem in randomized controlled trials (RCTs), and large placebo responses are thought to contribute to false-negative studies. This is a systematic review of 94 RCTs assessing the effect of neuropathic pain treatment. After controlling for confounders, the authors found that placebo responses in neuropathic pain trials depend on the particular pain syndrome. Placebo responses were highest in HIV-related neuropathic pain and lowest in postherpetic neuralgia and central pain. Although the study gave no explanation for these differences, it has an important impact on our interpretation of RCTs when comparing drug effects across different pain conditions. It is thus possible that the high failure rate observed in, for example, painful HIV neuropathy, reflects failed trials rather than lack of drug efficacy.
– Written by Nanna B Finnerup
The chemokine receptor 2 (CCR2) modulates neuronal excitability and synaptic transmission and activates spinal microglia. CCR2 is thought to be involved in neuropathic pain mechanisms and preclinical studies suggest that CCR2 antagonists are effective in neuropathic pain. This study is a double-blind, randomized, placebo-controlled study of the efficacy of a novel CCR2 antagonist AZD2423 in patients with painful diabetic polyneuropathy. AZD2423 was well tolerated but there was no effect of AZD2423 in doses up to 150 mg compared with placebo. Thus despite the positive data from experimental models, this study could not confirm an effect of a CCR2 antagonist in painful diabetic polyneuropathy.
– Written by Nanna B Finnerup
This is a very special study, for it is a well-designed, placebo-controlled clinical trial of a newer treatment intervention for lower back pain. It deserves our careful attention, both for what it does reveal and for what we cannot infer from the results. Three outcome measures were utilized: physical functioning (SF-36), Numerical Rating Scale (NRS) and Oswestry Disability Index (ODI). There has been abundant use of these measures; based upon extant literature, the authors state that significant improvements are reflected in an SF-36 increase of 15 points, NRS decrease of 2 points, and ODI increase of 10 points. Their study was designed and powered to detect such changes. A total of 32 patients were allocated to the active treatment and 32 to the sham treatment group; for various reasons five patients were lost from the treatment group and four were lost from the control group. At 6 months after treatment, the actively treated subjects had a mean 15-point increase in the SF-36 score, a 2.2 point decrease in the NRS score, and a 7.4 point decrease in the ODI. All of these values were significantly different from the sham-treated group, at less than the p = 0.05 level. Those who were treated at one level seemed to do better than those treated at two levels, but the groups were too small for significance to be reached. There were few changes in the sham-treated group. So far, so good, but the prudent reader needs to dig a little deeper into the study methods before he or she decides that this is a treatment that should be utilized for his or her patients.
All of the subjects had to have had pain for more than 6 months; back pain had to be greater than leg pain. There could be no history of prior low back surgery, the back pain must have been reproduced by provocative discography, disc height at the suspected level must have been greater than 50% of adjacent levels and an MRI showing evidence of degenerative disc disease at no more than two levels was required. These are rather stringent (and appropriate) criteria for such a study. More importantly, for the determination of whether your patient would fit into the results of this study, the exclusionary criteria must be considered. Excluded were subjects who had a disc bulge or herniation; more than two degenerated discs on MRI; and structural abnormality in the spine at the clinically significant level; any evidence of lumbar radiculopathy; spinal stenosis; significant comorbidities; were covered by workmen‘s compensation; any litigation; depression; ‘psychological issues‘; BMI of >30 kg/m2; or smokers. To use this study as a base for a clinical decision on your patient, you had best be sure that he or she fits the inclusion and exclusion criteria. Otherwise extrapolation to the clinical world is unwarranted. This is emphasized by the author‘s report that 1894 patients had to be screened over a 4-year period in their clinics to find the 64 who initially consented to be in the study. Finally, the authors admit that they did not determine whether or not the subjects could tell which arm of the study they were in. This is a potential confounder.
Finally, although statistical significance was reached at 6 months and marginal clinical significance was also reported, we need to look carefully at what these results say: intradiscal biacuplasty can lower an average patient‘s NRS from seven to five and can lower the SF-36 from 47 to 32, and the ODI from 40 to 33. Would your patient be satisfied with these results for the treatment of low back pain? Would you? It appears to me that intradiscal biacuplasty is a rather impotent treatment method based upon this study. On the other hand, its risks are certainly less than those of an open operation. Where to draw the line is still a matter of debate, but Kapural and his colleagues have performed a very useful study. The issue, once again, is efficacy versus efficiency.
– Written by John D Loeser
This is a retrospective study of clinical data over a 7-year period of chronic nonmalignant pain patients seen at Massachusetts General Hospital. Selection criteria were nonmalignant chronic pain condition, opioids prescribed over three or more clinic visits and information in the patient‘s chart on medical history, opioid dose and pain score. Data from 109 patients were utilized in this study. Over an average span of 704 days, there was no correlation of opioid dose increase or decrease with pain scores. No correlation of response to opioid dosing was seen with type of pain (nociceptive or neuropathic), gender or age. The authors separately analyzed those patients who had an increase of opioid dosage, a decrease in opioid dosage and no change in opioid dosage; in each of these groups there was no correlation with pain score. Although this study has some methodological challenges, it raises an important question: can chronic nonmalignant pain be successfully treated with opioids? Certainly, the lack of correlation of self-reported pain level with dosage of opioid raises the possibility that this is an ineffective therapy that is known to have significant risks to the patient and the community. It also raises the possibility that self-reported pain level is not an adequate outcome measure and that additional outcomes must be considered. Finally, there may be a subset of patients who do favorably respond to opioids, but they do not seem to be identifiable by existing criteria. This will certainly not be the final paper on the topic of opioids for chronic pain, but it does add to our information on this topic.
– Written by John D Loeser
Financial & competing interests disclosure
N Finnerup receives research funding from the Europain Investigational Medicines Initiative, which is a public–private partnership between the pharmaceutical industry and the EU. N Finnerup has also received research funding from Grünenthal, served as consultant for Astellas and Grünenthal, and received honoraria from Norpharma and Pfizer.The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.