anticonvulsants

Objectives. To evaluate the effectiveness and comparative effectiveness of nonopioid pharmacologic agents in patients with specific types of chronic pain, considering effects on pain, function, quality of life, and adverse events.

Data sources. Electronic databases (Ovid® MEDLINE®, Embase®, PsycINFO®, CINAHL®, Cochrane Central Register of Controlled Trials, and Cochrane Database of Systematic Reviews) through September 10, 2019, reference lists, data requests, and previous reviews.

Review methods. Randomized controlled trials (RCTs) of nonopioid pharmacologic agents in patients with chronic pain were selected using predefined criteria and dual review. This review focused on seven common chronic pain conditions (neuropathic pain, fibromyalgia, osteoarthritis, inflammatory arthritis, low back pain, chronic headache, sickle cell disease), with effects analyzed at short term (1 to <6 months following treatment completion), intermediate term (≥6 to <12 months), and long term (≥12 months). Magnitude of effects were described as small, moderate, or large using previously defined criteria, and strength of evidence was assessed. Meta-analyses were conducted where data allowed, stratified by duration within each intervention type, using random effects models. We evaluated effect modification through subgroup and sensitivity analyses, including specific drug, dose, study quality, and pain type.

Results. We included 185 RCTs in 221 publications and 5 systematic reviews. In the short term, anticonvulsants (pregabalin, gabapentin, and oxcarbazepine for neuropathic pain, pregabalin/gabapentin for fibromyalgia), serotonin-norepinephrine reuptake inhibitor (SNRI) antidepressants (duloxetine for neuropathic pain, fibromyalgia, osteoarthritis, and low back pain, milnacipran for fibromyalgia), and nonsteroidal anti-inflammatory drugs (NSAIDs) (for osteoarthritis and inflammatory arthritis) were associated with mostly small improvements (e.g., 5 to 20 points on a 0 to 100 scale) in pain and function. Function was not found to be improved with duloxetine for low back pain or pregabalin/gabapentin for neuropathic pain. Moderate improvement in quality of life was seen with duloxetine in patients with neuropathic pain, and small improvements in patients with osteoarthritis, but evidence was insufficient to draw conclusions for other drugs and conditions. While most comparisons of drugs and doses did not identify differences, diclofenac improved pain and function moderately more than celecoxib. In the intermediate term, limited evidence (1 RCT) showed memantine moderately improved pain, function, and quality of life in patients with fibromyalgia; improvements in pain, but not function, were maintained in the intermediate term with duloxetine and milnacipran for fibromyalgia. Other drugs studied, including acetaminophen (osteoarthritis), capsaicin (neuropathic pain), cannabis (neuropathic pain), amitriptyline (fibromyalgia, neuropathic pain), and cyclobenzaprine (fibromyalgia) had no clear effects. Withdrawal from study due to adverse events was significantly increased with nonopioid drugs, with the greatest increase over placebo seen with cannabis. Large increases in risk of adverse events were seen with pregabalin (blurred vision, cognitive effects, dizziness, peripheral edema, sedation, and weight gain), gabapentin (blurred vision, cognitive effects, sedation, weight gain), and cannabis (nausea, dizziness). Dose viii reductions reduced the risk of some adverse events with SNRI antidepressants. In the short term small increases in risk of major coronary events and moderate increases in serious gastrointestinal events (both short and long term) were found with NSAIDs.

Conclusions. In the short term, small improvements in pain and/or function were seen with SNRI antidepressants for neuropathic pain, fibromyalgia, osteoarthritis, and low back pain; pregabalin/gabapentin for neuropathic pain and fibromyalgia; oxcarbazepine for neuropathic pain; and NSAIDs for osteoarthritis and inflammatory arthritis. Improvement in function was not found with duloxetine for low back pain and pregabalin/gabapentin for neuropathic pain. Intermediate- and long-term outcomes were mostly not assessed. Increased incidence of drug class–specific adverse events led to withdrawal from treatment in some patients, suggesting that careful consideration of patient characteristics is needed in selecting nonopioid drug treatments.