In Reply

We would like to thank Drs. van Schalkwyk and Kachhwah et al. for their comments concerning our article.¹  Dr. van Schalkwyk reports that our statement “gabapentinoids are very well tolerated” is “an assertion contradicted by even the most enthusiastic of the three old meta-analyses [we] cite.” Our article cites five, not three, meta-analyses.^2–6  These were not old, but were the most recent meta-analyses on perioperative gabapentinoids available at the time of our writing the review, with two having been published in 2006 and the remaining three in 2007, 2011, and 2012. The five meta-analyses we cited had mildly differing conclusions on the tolerability of gabapentin. One found “the incidence of gabapentin-related side effects (dizziness, lightheadedness, visual disturbance, and headache) was similar in the gabapentin and control groups.”⁵  Another found “gabapentin administration was associated with sedation and anxiolysis (OR = 3.28; CI, 1.21–8.87) but not associated with a difference in lightheadedness, dizziness, nausea, or vomiting.”³  Yet another found that gabapentin “increased the rates of dizziness and sedation in surgical patients” but “most of the reported dizziness episodes were mild and did not affect patient satisfaction.”⁴  A renewed search when writing this letter identifies only one more recent meta-analysis on perioperative gabapentinoids published in 2013, which fortunately specifically addresses how perioperative gabapentinoids were tolerated by patients. It states, as one of the three “Key Points,” “both gabapentin and pregabalin were well tolerated.”⁷  Thus, the meta-analyses we cited and the additional new meta-analysis support that perioperative gabapentinoids are well tolerated. Although the focus of our article was not patient satisfaction, patient satisfaction is a key measure of how well perioperative gabapentinoids are tolerated, which becomes relevant in light of Dr. van Schalkwyk’s concerns.

We identified nine trials in which perioperative gabapentin was assessed alone or in combination, in terms of its effect on overall patient satisfaction. One trial found no difference in satisfaction between gabapentin and control group, supporting that gabapentin is well tolerated,⁸  whereas the remaining eight trials demonstrated statistically significant improved patient satisfaction when gabapentin was used. This suggests that regimens containing perioperative gabapentinoids may be better tolerated than standard higher opioid regimens.^9–16  Indeed, the study by Turan et al. examining the utility of supplementing perioperative epidural use with perioperative gabapentin explicitly states, “despite an increased incidence of dizziness [gabapentin] also increased patient satisfaction.”¹⁶  Similarly, of the four trials we identified that looked at patient satisfaction after perioperative pregabalin, two identified that patient satisfaction was similar in the presence or absence of pregabalin^17,18  and two found that pregabalin actually improved overall patient satisfaction.^19,20  In other words, although it is true that gabapentinoids have side effects, this is not an evidence for Dr. van Schalkwyk’s suggestion that gabapentinoids are not well tolerated. In summary, the existing data suggest that patients tend to be more satisfied when gabapentinoids are used as part of a multimodal perioperative analgesic plan. Our statement that perioperative gabapentinoids are very well tolerated is both accurate and supported by the wealth of data accumulated on this issue.

Dr. van Schalkwyk also reports that we “fail to report accurately the findings of a more recent analysis.” We regret he finds our characterization inaccurate. For the accuracy of the record, the referenced article states: “At least 50% pain relief over 6 hour was achieved by 15% with gabapentin 250 mg and 5% with placebo; giving a RB of 2.5 (95% CI 1.2 to 5.0) and an NNT of 11 (6.4 to 35). Significantly fewer participants needed rescue medication within 6 hours with gabapentin 250 mg than with placebo; NNT to prevent use 5.8.”²¹  In other words, even when given as a single postoperative dose at the bottom of the dose scale, gabapentin is superior to placebo, but is not as good as conventional stand-alone analgesics.

Finally, Dr. van Schalkwyk states “the overall tenor of the review is that gabapentin is substantially effective, both in the management of acute postoperative pain and in the prevention of chronic postsurgical pain. Neither contention is supported by independent analyses.” He is correct about our tenor with regard to statements about the efficacy of gabapentin in acute postoperative pain, but not about whether these are supported by independent analyses. We do believe based on the large number of randomized clinical trials and the multiple independent analyses conducted as the meta-analyses cited above that perioperative gabapentin is substantially effective in reducing acute postoperative pain scores and opioid use in the first 24 h after surgery. Regarding this effect, our tenor has been correctly perceived. Furthermore, we think this is no longer substantially controversial.

Our tenor and opinion are more reserved for both the effect of pregabalin on acute pain and a postulated preventative effect of any gabapentinoid on chronic pain. Data are encouraging but less conclusive for pregabalin than gabapentin in the acute postoperative setting. With regard to effects on chronic pain, based on the existing literature and analyses, we do believe it “considerably more likely than not that gabapentinoids do have a preventative effect with regard to the formation of chronic postsurgical pain.” We think that is different from asserting these drugs are substantially effective for this purpose.

Dr. Kachhwah et al. raise interesting and troubling points about two potential interactions of opioids and gabapentinoids that might generate clinical outcomes not often captured by perioperative analgesic studies of gabapentinoids: opioid abuse and respiratory depression.

We believe Anesthesiology and other journals in the field should encourage submission of follow-up articles documenting long-term follow-up from randomized controlled trials of perioperative gabapentinoids. Such studies should include not only chronic pain outcomes but also long-term opioid use outcomes. Long-term opioid use and/or misuse after surgery is an important clinical outcome with very significant public health implications deserving much more attention. The possibilities that perioperative gabapentinoids might reduce chronic opioid use and misuse after surgery by reducing early postoperative opioid use, or paradoxically might increase chronic opioid use and misuse by making opioids more rewarding, should be actively investigated.

Ultimately, more data are needed to know whether we should be avoiding gabapentinoids in people at high risk for respiratory depression (e.g., sleep deprivation, renal dysfunction, neuraxial opioid use, obstructive sleep apnea, and the elderly, as suggested by Kachhwah et al.), and caution is indeed warranted. However, the statement that pregabalin causes a predisposition to respiratory depression must be interpreted with caution. In the three case reports cited by Kachhwah et al., all patients received perioperative opioids and all were successfully resuscitated as a result of naloxone administration. This suggests that too much opioid was a significant risk in these cases, but does not exclude the possibility that pregabalin might increase the risk of respiratory depression from coadministered opioids. Nonetheless, the high-risk population that Dr. Kachhwah et al. suggest not receive gabapentinoids is exactly the population most at risk from higher opioid doses used in lieu of gabapentinoids. It may ultimately be that these are the people who benefit most from perioperative gabapentinoids due to the opioid-sparing effects. We note that a small study reported elderly patients receiving perioperative gabapentin were less likely to experience postoperative delirium compared with patients who did not receive gabapentin—perhaps, as a result of the opioid-sparing effects of gabapentin.²²  Treatment of patients at increased risk of respiratory depression should be approached with the highest degree of caution with or without the use of perioperative gabapentinoids. Overall, until more data emerge we do support the approach advocated by Kach hwah et al. to screen high respiratory risk patients from routine use of perioperative gabapentinoids.