Do perioperative gabapentinoids decrease postoperative pain?

Welcome back to Depth of Anesthesia. This is a podcast that critically explores our clinical practices. My name is David Hao and I'm an anesthesiologist at the Massachusetts General Hospital. For the first time in depth of anesthesia history, it's a solo episode. That's right, just you and me for this one This episode is actually based on a talk that I gave last year at the third annual Shenzhen Anesthesiology Innovation and Doctoral Forum.

A big thank you to Dr. JP Wong from our department for the opportunity to talk. I presented a review of the evidence around the perioperative use of gabapentinoids, and remember thinking to myself at the time, wouldn't this make for a nice depth of anesthesia episode? Well, here we are. Instead of starting with the case today, we're going to jump right in with the claim, which is that gabapentinoids, which is the class that includes gabapentin and pregabolin, decrease post-operative pain Simple enough. Here now are two questions for you, our listeners, to think about.

What is the level of your agreement with the claim? And what is the level of evidence for what you believe? Alright, a little background on the two drugs we'll talk about today. Interestingly, pregabalin and gabapentin share a similar mechanism in action. Their effects, maybe somewhat surprisingly, are mostly mediated by binding of alpha 2 delta subunits in voltage-gated calcium channels.

This leads to inhibition of calcium influx and subsequent release of excitatory neurotransmitters As a class, gabapendenoids were originally marketed for use in the 1990s as anticonvulsants. So the FDA indications for pregabolin and gabapendin are actually quite limited. In the context of pain, pregabalin has indications for neuropathic pain associated with diabetic peripheral neuropathy, posturpatic neuralgia, fibromyalgia, and neuropathic pain associated with spinal cord injury The only pain-related indication for gabapentin is actually in the management of post-herpetic neuralgia. But in the past decade or so, uh off-label usage of gabapentinoids for both acute nociceptive and neuropathic pain has substantially increased. And perioperative use of gabapentinoids in particular has been fairly ubiquitous and often embedded in our clinical pathways such as enhanced recovery after surgery protocols.

Excitement for the potential value of gabapentinoids first started gaining momentum sometime in the early 2000s. One of the first studies in this space was a randomized double-blind placebo-controlled study conducted by Dirks and colleagues and published in Anesthesiology in 2002. The study was titled A Randomized Study of the Effects of Single Dose GABAPentin v. Placebo on Postoperative Pain and Morphine Consumption after mastectomy. So in this study, they took 70 patients who are undergoing unilateral radical mastectomy and randomized them to either receive a single dose of 1200 milligrams of oral gabapentin or placebo one hour before surgery.

Everyone was then given a morphine PCA with a 2.5 milligram dose and a lockout time of 10 minutes for four hours post-operatively. And what they found was that total morphine consumption was reduced from 29 milligrams in the placebo group to 15 milligrams in the gabapentin group. And again, this is for the duration of the PCA, which is four hours post-op. The differences between pain at rest at two and four hours post op were not statistically significant, but pain during movement was reduced from forty one millimeters in the placebo group to twenty two millimeters in the gabapentin group at two hours post op. and from thirty-one to nine at four hours post op. The authors actually commented, quote, This is the first clinical study to demonstrate an analgesic or anti-hyperalgesic effect of gabapentin in somatic postoperative pain.

End quote. There's a interesting accompanying editorial by Doctor Gilron, which actually shares, I think, some interesting insights to perspectives from the time. The editorial is titled Is Gabapentin a Broad Spectrum Analgesic and published in Anesthesiology in 2002. I quote from the article, Data from Dirks et al. support a favorable therapeutic profile in postoperative patients, suggesting that gabapentin may be a welcome addition to the anesthesiologist's pharmacopoeia of coanalgesics. such as NSAIDs and local anesthetics. However, before considering the routine use of gabapentin in post-operative pain, further studies are needed to 1, replicate the findings of Dirk's et al. 2.

Evaluate the safety and efficacy of postoperative multidose gabapentin administration. And 3. Determine the optimal duration of use and dose of gabipentin in postoperative pain. End quote. What a perfect transition for us to start discussing what the next 20 years have actually yielded.

We'll start with a systematic review and meta-analysis published in 2015 in the British Journal of Anesthesia. That explores the impact of pregabalin on postoperative pain scores in opioid consumption. And they looked at whether those outcomes differ according to things like pregabalin dose or frequency of administration. The authors searched for any randomized control trials that compared pregabalin to control in patients undergoing surgery, and a total of 55 studies comprising over 4,000 patients were included in the final analysis. with 49 of these studies investigating acute pain. Of the total 4,155 patients, about 55 % which is two thousand two hundred and seventy received pregabalin and the other forty-five percent or a thousand eight hundred eighty five patients served as controls.

For the different dosing regimens, the authors looked at subgroup analyses at 2 hours and at 24 hours. At two hours, they looked at the individual doses of pregabalin given before surgery. So this was either less than or equal to 75, between 100 and 150, or 300 milligrams. And at 24 hours, they looked at the dose and frequency for the same three dose levels, as well as single versus multiple dosing at each dose level I won't go too much into detail on the subgroup analyses for pain scores, but as usual, I'll provide a citation to the manuscript and fortunately it's also an open access article. So let's talk about the primary outcomes.

In terms of pain scores at two hours, pool results showed a statistically significant reduction in pain scores at rest. and during movement with a mean difference of negative 0.81 and negative 0.94 respectively. All VAS scores were converted to a 0 to 10th scale by the way. For the pain scores at twenty-four hours, pooled results showed a statistically significant reduction in scores at rest, with a mean difference of negative zero point three eight, and during movement at negative zero point four seven. For opioid consumption, there was a statistically significant reduction with a mean difference of negative 2.09 mg of morphine equivalence at 2 hours at negative eight point two seven milligrams of morphine equivalents at twenty four hours.

For the subgroup analysis at two hours, opioid consumption was reduced in the one hundred and one fifty and three hundred milligram dose groups compared with placebo. but not with the 75 milligram dose level. And in subgroup analysis at 24 hours, all doses of pregabalin reduce 24 hour opioid consumption compared with placebo. when given either as a single preoperative dose or as multiple doses. But no differences were found between giving either a single dose or multiple doses at any dose level or between the dose levels. For the secondary outcomes, I'll mention specifically the issue of side effects, sedation, severe sedation, and dizziness at 24 hours. and visual disturbances up to twenty-four hours were significantly more common in pregabalin-treated patients.

So the relative risk of sedation was 1.46 Severe sedation was two point one one, dizziness was one point three three, and visual disturbances was three point five two. So in terms of takeaways, we have a systematic review and meta-analysis that demonstrates statistically significant reductions in pain scores at rest. with a mean difference of about 0.8 at 2 hours and 0.4 at 24 hours. Pain scores during movement, a mean difference of about 0.6 at 2 hours and 0.5 at 24 hours. and opioid consumption, a mean difference of about 2 mg of morphine equivalence at 2 hours and 8 milligrams at 24 hours compared to placebo. To the astute observer, I've made no mention at all about clinical significance, only statistical significance.

Now the trade-off here in terms of risks is that pregabalin was associated with higher incidences of sedation dizziness and visual disturbances relative to placebo. And as with most systematic reviews and meta-analyses, we have our limitations relating to heterogeneity. As you might expect, the studies were conducted across different types of surgery with different pregabalin doses, different anesthetic types, and ultimately quite a few different dosing regimens. Okay, so the second systematic review and meta-analysis we're going to discuss was published in 2016 entitled Gabapentin for Postoperative Pain Management, a systematic review with meta -analyses and trial sequential analyses published by Fabritius et al. This was published in 2016 and looked at the impact of GABAPentin on postoperative opioid consumption, pain intensity, and adverse effects in specifically adult surgical patients.

So any RCTs that looked at perioperative gabapentin versus placebo were considered. Prospective, observational, or quasi -randomized trials were included only for evaluation of harm. and detection of rare serious adverse events. Now before we go any further, I do want to spend some time talking about the concept of trial sequential analysis. Which is something you'll notice is in the title of the systematic review of meta-analysis that we're looking at. As we've probably come to appreciate over the course of these depth of anesthesia episodes Systematic reviews and meta-analyses are high on the hierarchy of evidence, and they definitely have an important role in guiding clinical practice.

But Uh, it's not uncommon for many of these analyses to have weak or insufficient evidence, largely due to a lack of statistical power. In other words, there might not be enough participants to adequately power the assessment of whatever intervention you're looking at. So when new clinical evidence such as trials emerge, there it can be a tendency to repeat the meta-analysis to look for signal. The problem this causes is that repetition of statistical testing increases the likelihood of error in the form of either false positive or false negative results. And so trial sequential analysis addresses this very specific statistical issue.

So conceptually, it's kind of similar to the interim analyses you might see in RCTs, where trial results are tested at regular intervals. So at the end of the day, what trial sequential analysis gives us is an adjusted confidence interval, which better controls for both type 1 and type 2 errors. Trial sequential analysis or TSA also gives us an estimate of the required information size on outcomes. So this is basically the number of participants or events necessary to detect or reject an a priori assumed intervention effect. If the amount of information you accumulate fails to achieve that required number, then the uncertainty of the estimate of the intervention effect increases.

Put another way, the higher the fraction of the required information size the meta-analysis obtains, the lower the overall uncertainty. So for this study specifically, the authors mentioned that TSA was used to adjust the confidence intervals to preserve the risk of type 1 and type 2 errors within 5 and 90%, respectively. because of the issue of sparse data and repeat testing in cumulative meta-analyses. You might need a quick rewind and maybe a little bit of time away from the podcast to review this concept, but I'll briefly go over it again before the next study, which also happens to use this technique. Alright, so the study ultimately included a total of 132 trials with about 9,500 patients for evaluation of benefit. and an additional three non randomized trials for evaluation of harm.

In terms of dose, the gabapentin ranges were from a hundred to twelve hundred milligrams with single dose therapy, and from nine hundred to twenty four hundred milligrams per day. with multiple dose therapy. And for timing, gabapentin was started from 30 minutes to 48 hours pre-operatively. Another interesting aspect of this study was that the authors decided a priori to make conclusions based primarily on results with low risk of bias, as determined according to the Cochrane Handbook As an example, for the co-primary outcome of post-operative opioid consumption, they looked only at the 13 trials which were assessed to have low risk of bias. These trials ultimately demonstrated a reduction in 24-hour postoperative morphine consumption of about 3 mg with a confidence interval of 0.5 to 5.6 The TSA adjusted confidence interval, however, was negative 0.2 to 6.3, which crosses zero, with an accrued percentage of required information size of 71%.

The other co primary outcome was serious adverse effects. This was defined as either life threatening, resulting in death, disability, or significant loss of function, and causing hospital admission or prolonged hospitalization. The relative risk in trials with low risk of bias of serious adverse effects in patients treated with gabapentin versus placebo was 1.61 with a confidence interval from 0.91 to 2.86. The TSA adjusted confidence interval was 0.57 to 4.57 with an accrued percentage of required information size of 42%. In terms of pain at six hours, the trials with low risk of bias demonstrated that pain at rest was not significantly reduced.

Pain during mobilization was decreased by 9 millimeters on a 100mm scale. with a confidence interval between four and thirteen, and with TSA adjustment four to eighteen. By twenty-four hours, however, pain at rest and during mobilization were not significantly different in the GABAPentin group. So, in terms of takeaways, based on trials with low risk of bias, we have a 24-hour post-operative morphine consumption decrease of about 3 milligrams with a TSA adjusted confidence interval that crosses zero Reductions in pain were only seen at six hours with mobilization, and that was a decrease of about 10 millimeters on a 100mm scale. So this leads us to the most recent systematic review and meta-analysis published in Anesthesiology in August of 2020 by Michael Varet and colleagues.

This systematic review was performed to look at the analgesic effect and adverse events of perioperative gabapentinoids in adult surgical patients. Now, contrary to prior studies, this review actually pooled pregavin and gabapentin together and included any RCTs that compared gabapentinoids as a class to placebo, any other analgesic regimen, or usual care. Gabapentinoids had to have been initiated between one week before and up to 12 hours after surgery. The co-primary outcomes were postoperative acute pain at 6, 12, 24, 48, and 72 hours after surgery, measured by any quantitative pain scale. Some secondary outcomes included post-operative pain between weeks four and twelve, incidence of post-operative chronic pain defined as pain lasting for three months or more.

Cumulative opioids administered within 24, 48, and 72 hours after surgery, and of course the incidence of adverse events. So a total of 322 trials were included in the systematic review and 281 reported quantitative data that was subsequently included for the meta-analysis. The most common trials were in patients undergoing orthopedic or spine surgeries. That was about 27%. And of the eligible trials, 52% looked at gabapentin, 43% at pregabline, and 5% at both Gabipentinoids were typically given as a single dose.

That was in 68% of trials, and most were given before surgery. For the primary outcome, a slightly lower postoperative pain intensity was observed at 6, 12, 24, and 48 hours with gabapenoid administration, but not at 72 hours. This effect was not clinically significant because it ranged below the minimally important difference of ten points out of a hundred for each time point and this was consistent across all the trials at either low or high or unclear risk of bias. Before we get any further, I do want to briefly touch on the concept of minimally important difference. This is something the authors are keen to emphasize and I do think it distinguishes their manuscript from prior publications and I alluded to it a little bit before with the idea of clinical versus statistical significance.

So uh minimally important difference or minimally clinically important difference is nicely defined by a gentleman named Jaysk. uh J A E S C H K E uh in nineteen eighty nine He writes that this concept is the smallest difference in score in the domain of interest, which participants perceive as beneficial and which would mandate in the absence of troublesome side effects and cost a change in the patient's management. So the authors here establish this minimal and clinically important difference to be 10 points on a 100-point scale. And they also characterize a difference of 20 to 30 points to be a quote appreciable effect and a difference of 50 to be a quote substantial effect. So using this idea, the authors looked at the probability of an experiencing an effect Greater than this minimally important difference of ten out of a hundred in the gabapentinoids group compared with the control group And at 6, 12, and 24 hours, gabapentinoids were associated with a slightly greater probability of an analgesic effect more than 10 points, but this difference was lost at 48 and 72 hours.

The proportion of patients achieving an effect greater than 20 out of 100 was small and limited to the first 12 hours and absent for an effect greater than 30. The amount of opioids administered in terms of IV morphine equivalents at 24 hours was slightly lower, a mean difference of about 8 mg with the use of gabapentinoids. Slightly less opioid use was also seen at 48 hours, at 10 milligrams less, this was uh across 24 trials, and at 72 hours at about 30 milligrams less, and this was across only four trials. So uh takeaways from the efficacy side of things, the study demonstrated statistically significant decreases in post-operative pain scores for up to 48 hours. But these scores did not meet the established minimally important differences of ten out of a hundred.

A slightly greater probability of analgesic benefit of at least ten out of a hundred was seen at six 12 and 24 hours, but that effect disappeared at 48 and 72. Few patients had an effect greater than 20, and of those that did, it was limited to 12 hours. and effects greater than 30 were absent. As with any therapy, we again need to weigh the potential benefits with the potential harms. Gabapentinoids here were associated with a greater incidence of dizziness with a risk ratio of 1.25 and visual disturbance with a risk ratio of 1.89. Gabipentenoids were not significantly associated with respiratory failure, ataxia, or falls, or delirium, and the risk of respiratory failure was not different when gabapentinoids were used with opioids.

And finally, as you might recall, I mentioned that we would see the concept of trial sequential analysis again, and here we are The authors mentioned that to limit a potential type 1 error and to inform future research, a trial sequential analysis was performed on the primary outcome. And based on the results, the sample size of the review and meta-analysis was larger than the required information size, suggesting that further research is unlikely to provide new evidence. as it pertains to post operative pain at six, twelve, twenty four, and forty eight hours. Alright, so the systematic reviews we've covered today have all in some way or another looked at the risk of adverse effects of gabapentinoids. And so we'll wrap up this episode with one more interesting study that looks specifically at GABAPentoids and the risk of postoperative naloxone administration.

So this study was published in the British Journal of Anesthesia in 2018 and was designed as a case control retrospective study. Their aim was to look at the association between the loxone administration, which was used as a surrogate for post-operative oversedation or respiratory depression, and gabipetenoid use. The authors retrospectively look for adult patients uh who were extibated from general anesthesia and were given naloxone within forty-eight hours of surgery from a time interval uh between twenty eleven and twenty fifteen. Now for each case of a patient who was given naloxone, two control patients were matched. So in terms of the findings, during the study timeframe, a total of over 110,000 patients underwent general anesthesia. and were subsequently admitted to standard post op wards.

A total of 128 patients were given naloxin within 48 hours. fifty one percent within the first twelve hours, and eighty percent by the first twenty four hours. Most interestingly, continuation of chronic abipetnoids into the post op period was associated with an odds ratio of six point three, so It's more than a six-fold increased risk of naloxone administration with a ninety-five percent confidence interval of two point four to sixteen point seven. The postdoc analysis also found associations between both low-dose and high-dose regimens, low dose being an odds ratio of 5.1 and high dose being an odds ratio of 14.1. Low dose was daily gabapentin less than 1800 milligrams or pregaballin less than 300 in 24 hours, and high dose was greater than or equal to 1800 of pre gabapentin. or greater than or equal to 300 of pregavin in 24 hours.

There was no association seen between a single perioperative gavapendent dose and post-operative naloxone administration But the authors were fairly careful to point out that the control patients were a match to the exact type of procedure. So this means that if a patient uh underwent, for example, a procedure with an enhanced recovery protocol that included a preoperative gabapenoid, then the two match controls were also very likely to have had the same protocol, which you can imagine would limit the ability to look at this as a risk factor. The study probably also underestimated the true incidence of post-op respiratory depression because the loxone administration typically is reserved for fairly high severity symptoms. In terms of takeaways, we have a retrospective case control study that suggests that continuation of chronic abendoids into the post op period was associated with an increased rate of naloxin administration with an odds ratio of 6.3.

As we wrap up this episode, I want to take a quick second to re-emphasize the importance of continuing to critically evaluate your practices. And question what might be generally accepted practices that might have become dogma. As we see in this episode today, early published reports of gabapentenoid benefits were largely devoid of data on adverse effects and risks. Subsequent reviews were certainly promising in terms of benefit, but on closer inspection, the astute reader might start to question and consider the concepts of clinical significance versus statistical significance. This I think is all set on the backdrop of a national epidemic of oral opioid over-prescribing that has encouraged clinicians to look for alternatives.

And the fire has certainly been fueled further by professional society guidelines in recovery after surgery protocols. I think the takeaways from the last systematic review and meta-analysis we covered are a nice review of the overall state of the literature and what we've covered today. We've seen from a few large meta-analyses that there are statistically significant decreases in post-operative pain scores with gabipentenoids at various time intervals. But most do not meet established minimally important differences of ten out of a hundred. And we also have data from these reviews that suggest that gabapentinoids are not benign and associated with adverse events, including things like dizziness and visual disturbances.

Thank you all for joining me on the show today. I hope this episode and this podcast has made you a better clinician by encouraging you to reevaluate your practices and to explore the primary literature. We have a few more episodes already lined up, which I hope to get to you all soon. Please let your colleagues know about the work that we're doing and don't hesitate to contact me with questions or episode suggestions. As always, stay hungry and keep asking questions and I'll see you next time.

See you guys next time.