Several treatment guidelines advocate for the use of antidepressants for neuropathic pain.41-44 For decades, TCAs have been employed off-label to successfully treat many patients with neuropathic pain states. Early investigations suggested that TCAs were robustly efficacious in managing patients with neuropathy.45-48 Calculated number-needed-to-treat (NNT) values for TCAs were quite low (ie, reflecting that few patients would need to be treated to yield a positive response in one patient compared with placebo), and were comparable to, if not slightly better than, the NNTs generated for anticonvulsants and α2-δ ligands, such as gabapentin or pregabalin.45-48
Unfortunately, early studies involving TCAs conducted many years ago do not meet contemporary standards of methodological rigor; they featured relatively small samples of patients assessed for brief post-treatment intervals with variable outcome measures. Thus, the NNT values obtained in meta-analyses based on these studies may overestimate treatment benefits.49 Further, NNT values derived from meta-analyses tended to combine all drugs within a particular antidepressant class (eg, amitriptyline, nortriptyline, desipramine, and imipramine among the TCAs) employed at diverse doses. Taken together, these limitations raise questions about the results of those meta-analyses.
Subsequent meta-analyses, which employed strict criteria to eliminate data from studies with potential sources of bias and used a primary outcome of frequencies of patients reporting at least 30% pain reduction compared with a placebo-controlled sample, suggest that the effectiveness of TCAs as a class for treating neuropathic pain is not as compelling as once was thought. Meta-analyses of studies employing specific TCAs revealed that there was little evidence to support the use of desipramine,50 imipramine,51 or nortriptyline52 in managing diabetic neuropathy or postherpetic neuralgia. Studies evaluating amitriptyline (dose range 12.5 to 150 mg/d), found low-level evidence of effectiveness; the benefit was expected to be present for a small subset (approximately 25%) of patients with neuropathic pain.53
There is moderate-quality evidence that duloxetine (60 to 120 mg/d) can produce a ≥50% improvement in pain severity ratings among patients with diabetic peripheral neuropathy.54 Although head-to-head studies with other antidepressants are limited, it appears that duloxetine and amitriptyline have comparable efficacy, even though the NNTs for amitriptyline were derived from lower-quality studies than those for duloxetine. Duloxetine is the only antidepressant to receive FDA approval for managing diabetic neuropathy. By contrast, studies assessing the utility of venlafaxine in neuropathic pain comprised small samples for brief durations, which limits the ability to draw clear (unbiased) support for its usefulness.55
Given the diversity of pathophysiologic processes underlying the disturbances that cause neuropathic pain disorders, it is unsurprising that the effectiveness of amitriptyline and duloxetine were not generalizable to all neuropathic pain states. Although amitriptyline produced pain-mitigating effects in patients with diabetic neuropathy and post-herpetic neuralgia, and duloxetine mitigated pain among patients with diabetic neuropathy, there was no evidence to suggest their effectiveness in phantom limb pain or human immunodeficiency virus-related and spinal cord-related neuropathies.35,53,54,56-58
Continue to: Fibromyalgia