Cognitive Effects of Perioperative Pregabalin: Secondary Exploratory Analysis of a Randomized Placebo-controlled Study

Pregabalin is increasingly used for the treatment of early postoperative pain as part of a multimodal pain regimen.¹  Its predecessor, gabapentin, was developed for the treatment of epilepsy,²  but later gabapentin and pregabalin were also approved for the treatment of neuropathic pain conditions, generalized anxiety disorders, and fibromyalgia.³  The justification for pregabalin perioperatively in multimodal pain treatment is determined based on analgesic efficacy and total burden of adverse effects.

Recently, there has been increased focus on adverse effects associated with the use of pregabalin.^4,5  Side effects, including somnolence, dizziness, visual disturbances, and ataxia, are frequently reported,^6,7  and recently, respiratory depressive effects have also been described.^8,9  Despite well-known central nervous system adverse effects, the cognitive effects of perioperatively administered pregabalin have not been investigated. Several factors, such as age, educational level, and preoperative mental health, are recognized as factors that contribute to early postoperative cognitive impairment.^10,11 

Postoperative cognitive dysfunction is not clearly defined in the literature but differs from postoperative delirium because it is more subtle and longer lasting.^12,13  Therefore, postoperative cognitive dysfunction may only be identified through specific and sensitive neuropsychological tests¹⁴  that measure processes like attention, working memory, decision-making, inhibition, psychomotor speed, and mental flexibility.¹⁵  The umbrella term “executive function” encompasses these processes and is regarded as essential for goal-directed behavior in which attentional functions like alerting, orienting, and executive control and their interactions play important roles.¹⁶  Moreover, executive function is essential for appropriate decision-making and impulse control,¹⁷  making it a crucial factor in risk of reinjury and impulsive actions after surgery.

We previously performed a randomized, parallel group investigation in 80 donor nephrectomy patients that evaluated the analgesic, opioid-sparing, and antihyperalgesic effects of pregabalin.¹⁸  In the this article, we present a secondary exploratory analysis of previously unpublished data of cognitive test performance in the same patients.

The aim of this study was to explore the cognitive effects of 150 mg of pregabalin—administered twice daily, on the day of surgery and on the subsequent day—using neuropsychological tests suitable for repeated assessments measuring visual memory, working memory, executive function, and attention. We hypothesized that pregabalin, relative to placebo, administered as part of a multimodal pain regimen consisting of opioids, steroids, local anesthetics, and acetaminophen, would affect executive functioning, including inhibition, strategic thinking, mental flexibility, working memory, and processing speed.

Approval for this study was obtained from the Regional Committee for Medical Research Ethics for Eastern Norway, Oslo, Norway, and the Norwegian Medicines Agency, Oslo, Norway, and was also registered at EudraCT (2009-015316-17) and www.ClinicalTrials.gov (NCT01059331, January 28, 2010). The study is reported in accordance with the Consolidated Standards of Reporting Trials Statement¹⁹  for randomized trials and conducted in compliance with the Declaration of Helsinki. Eighty patients with American Society of Anesthesiologists classification I or II scheduled for elective laparoscopic donor nephrectomy were included. After a written consent, the patients were randomized to one of two treatment groups: 150 mg of pregabalin or placebo was administered two times daily over two consecutive days (totaling 600 mg), starting in the morning before surgery. Standard postoperative analgesic treatment was administered comprising local anesthetic wound infiltration, a single dose of dexamethasone, acetaminophen regularly, and opioids (ketobemidone) as required. The analgesic, opioid-sparing, and antihyperalgesic effects of pregabalin in the same study population have been extensively reported in the primary publication.¹⁸  Details on inclusion and exclusion criteria, randomization, blinding, drug preparation, anesthesia, and perioperative treatment are also previously published.¹⁸ 

Cognitive function was measured using tests selected from the Cambridge Neuropsychological Test Automated Battery, a computer-based test battery assessing different domains of cognitive functioning, including executive function, attention, and psychomotor speed.²⁰  The tests were performed using a computer tablet and a press pad. During the test period, the patients were seated in an upright comfortable position with the computer tablet arranged on a table at a 50-cm distance, with the press pad placed on the table in front of the computer, allowing the patient’s dominant hand and arm to rest comfortably beneath it. The test battery included six different tests selected a priori based on earlier studies on cognitive functioning in patients treated with pregabalin.⁸ 

The motor screening test screens for visual, movement, and comprehension difficulties. It is administered to introduce the patient to the computer, the touch screen, and the press pad and is a prerequisite for administering the test battery. This test was not used for subsequent statistical analysis.

The reaction time test measures response speed when a yellow spot appears in the center of the screen or in five different locations on the screen, and the subject must react to the stimulus by releasing the press pad button and touching the screen where the spot appeared. The relationship between cognitive functions and physical movements are then used to calculate psychomotor speed. We used the following outcome: reaction time, five-choice, which measures the psychomotor speed.

The paired associated learning test assesses visual learning and recognition memory by displaying six to eight boxes on the screen that are opened in a randomized order. The patterns shown in the boxes are then displayed in the middle of the screen, and the subject has to remember in which box the pattern was originally located. The test has five stages with increasing memory loads, specifically, increasing the numbers of patterns from two to eight. We used the following outcome: paired associated learning, total errors (six shapes), adjusted for each stage not attempted because of previous failure.

The rapid visual processing test measures a subdomain of sustained attention, evaluating both continuous performance and visual sustained attention. In this test, a digit from 2 to 9 appears in a pseudorandom order at a rate of 100 digits/min. The patients are requested to detect three different target sequences of digits by pressing the response button. We used the following outcome: rapid visual processing A', measuring the sensitivity to the target regardless of response tendency.

The spatial working memory test assesses the ability to implement strategic thinking for problem-solving. The test has notable executive function demands and provides a measure of strategy as well as working memory errors. In this test, several boxes are displayed on the screen. The subject is required to search for one blue token in each of the boxes by the process of elimination. The number of boxes increases from three to eight. The existence of a predefined search strategy was examined.²¹  A high score indicates poor strategy, whereas a low score indicates an effective use of strategy. We used the following outcome measures: (1) spatial working memory within errors, four to eight boxes, defined as the number of times the patient revisits a box already found to be empty, and (2) spatial working memory strategy, defined as the number of times a new search begins with a different box.

The stop-signal task test is an executive function task measuring response inhibition and impulsivity. In this go/no-go task, a circle is shown on the screen. After a fixed delay, an arrow pointing to the left or right is displayed within the circle. The subjects are requested to press the left or right press pad button according to the direction of the arrow. In the second part of the test, the patients are asked to withhold their response when they hear an auditory signal. We used the following outcome: stop-signal task stop-go, measuring the number of direction errors on stop-and-go trials.

The cognitive test battery was administered three times during the hospital stay. The first assessment was conducted 1 to 3 days before surgery and served as a baseline. The second test was performed during the day 24 h after surgery, whereas the third assessment was completed 3 to 5 days postoperatively and at least 36 h after the final administration of the test drug. The primary outcomes of interest were the difference between the pregabalin group and the placebo group from baseline to 24 h after surgery for all six neurocognitive tests. Premorbid intelligence was estimated using a Norwegian version of the National Adult Reading Test.²² 

The sample size is based on a previous study of the same study population investigating the opioid sparing and analgesic effect of pregabalin.¹⁸  In this earlier study, 20% difference in opioid usage was considered clinically relevant. With mean opioid consumption of 50 mg and SD of 15 in the control group, a two-sided α = 0.05 and a statistical power of 0.8, 36 participants were required in each group. To compensate for dropouts, 80 participants were included. The effect of pregabalin on cognitive outcomes in the early postoperative period has not been well studied. Thus, this study must be considered exploratory with respect to cognitive outcome measures. Participant demographics and intraoperative data were summarized and stratified according to treatment groups as mean (± SD) and numbers, as appropriate and independent samples t tests and chi-square for independence tests were used to test for group differences. Mixed-effect negative binomial regression was used to fit overdispersed count data (spatial working memory within errors, stop- signal task stop-go, paired associated learning), and the linear mixed-effects model was used for normal distributed continuous data (spatial working memory strategy, rapid visual processing A', reaction time). In the models, treatment groups, time, and treatment-by-time interactions were defined as fixed effects, whereas subjects were treated as a random effect accounting for repeated measurements within patients. To test our hypothesis, we estimated the difference between the change from baseline to 24 h after surgery in the pregabalin group and the change from baseline to 24 h after surgery in the placebo group using the placebo as the reference group. The results were reported as rate ratios and estimated mean differences, with 95% CI. Although not of primary interest, we also explored differences between the groups from 24 h after surgery to 3 to 5 days after surgery in a secondary analysis. Data analyses were performed with IBM SPSS Statistics 25 (SPSS Inc., USA).

The study was performed at Oslo University Hospital, Rikshospitalet, Norway, between February 2010 and October 2012. In total, 80 patients, 55 females and 25 males, scheduled for elective donor nephrectomy were randomized. All 80 patients completed the study, with a limited number of protocol deviations (fig. 1). One patient in the pregabalin group received inhalation anesthesia with sevoflurane 90 min before switching to total intravenous anesthesia, whereas one patient received erroneously higher boluses of opioids from the patient-controlled analgesic device and was treated with naloxone. Seven patients in the placebo group and one patient in the pregabalin group refused testing 24 h after surgery because of nausea and pain. One patient in the placebo group refused the last test 3 to 5 days after surgery as well as further participation, because of complaints about discomfort and pain. The treatment groups were comparable in terms of age, sex, education, Norwegian National Adult Reading Test scores, and intraoperative data (table 1).

Cognitive functioning was evaluated using six different neurocognitive tests. Descriptive data presented as median and 25th and 75th percentiles at baseline, 24 h after surgery, and 3 to 5 days after surgery are shown in table 2. In the spatial working memory within errors test, the number of errors increased significantly from baseline to 24 h after surgery in the pregabalin group compared with the placebo group (rate ratio [95% CI], 3.20 [1.55 to 6.62]; P = 0.002; table 3). Similarly, in the stop-signal task stop-go test, the number of errors increased significantly from baseline to 24 h after surgery in the pregabalin group compared with placebo (rate ratio, 2.14 [1.13 to 4.07]; P = 0.020; table 3). In the time period from 24 h after surgery to 3 to 5 days after surgery, the difference between the pregabalin and placebo groups in the spatial working memory within error test (rate ratio, 0.43 [0.21 to 0.88]; P = 0.022), as well as the stop-signal task stop-go trial (rate ratio, 0.28 [0.16 to 0.50]; P < 0.001), reduced significantly. In the paired associated learning, spatial working memory strategy, reaction time, and rapid visual processing A' tests, the differences between the pregabalin group and the placebo group were not significant (table 3).

In this randomized controlled trial, the cognitive effects of perioperative pregabalin were assessed using neurocognitive tests measuring different aspects of memory, attention, and executive functions. We found that in tests of inhibition and working memory, performance was significantly impaired by the perioperative administration of pregabalin when compared to placebo.

More specifically, the results showed a significantly impaired performance in the pregabalin group in the spatial working memory within error test. This test assesses visuospatial working memory and strategic choices,²³  which are important for performance of higher cognitive functions.²⁴ 

Impulsivity and inhibitory functions were measured using the stop-signal task stop-go test.²⁵  The results in this study showed a significant increase in the number of errors in the pregabalin group. Poor inhibition has been linked to decreased motor control and increased risk of falling in elderly patients and those with mild dementia.^26,27 

In the reaction time five-choice test, a test of processing speed and attention, both groups showed significantly reduced performance from baseline to 24 h after surgery, although the difference between the two groups was not significant. Earlier studies have demonstrated that opioids reduce processing speed compared to placebo.^28,29  Patients in this study required less opioids when treated with pregabalin. Nevertheless, both groups showed similar worse performance on the first postoperative day, indicating that pregabalin may also contribute to reduced psychomotor speed.

The cognitive effects of pregabalin have been investigated in animal models³⁰  and in human preclinical models. In a crossover study in healthy volunteers by Hindmarch et al.,³¹  the cognitive and psychomotor effects of 150 mg of pregabalin (three times a day) for 3 days were investigated, using alprazolam and placebo as a positive comparator and control, respectively. Pregabalin showed limited negative effects in tests assessing arousal (critical flicker fusion) and attention (compensatory tracking task), but no significant effects in psychometrics compared with placebo were observed. In another study,³²  pregabalin was titrated more than 8 weeks to 600 mg/day and then maintained at this dose for an additional 4 weeks. Pregabalin impaired outcomes in half of the cognitive tests (digit symbol, Stroop, controlled oral word association), which was supported by subjective cognitive complaints in the patients treated with pregabalin. Furthermore, in a study by Ciesielski et al.,³³  the neuropsychological and psychiatric impact of 300 mg of pregabalin versus levetiracetam was investigated, and significantly reduced verbal and visual episodic memory was observed in patients treated with pregabalin compared to those treated with levetiracetam. Finally, in a crossover study³⁴  in young healthy volunteers, the psychomotor effect of 75 mg of pregabalin in combination with 10 mg of oxycodone orally was investigated. The results, however, did not reveal any significant psychomotor effect in performance on the test used. Thus, our results are in line with the findings in these studies suggesting that pregabalin may impair executive functions rather than psychomotor speed. Furthermore, control measurements without the test medication 3 to 5 days after surgery were similar between the treatment groups, implying that the cognitive impairments were transient.

Attentional functions such as being alert, orienting, and executive control are all important prerequisites for performing daily tasks and managing everyday life.¹⁶  Furthermore, cognitive function plays an important role in the control of gait and balance.³⁵  In a systematic review and meta-analysis of the adverse events associated with pregabalin, disturbances related to coordination and cognition were the most common, indicating involvement of both vestibulocerebellar and brainstem structures, as well as higher cortical functions.³⁶  There was a clear dose–response relationship in the onset of adverse events, and most adverse events were observed at pregabalin doses of 300 mg/day. The authors concluded that the use of pregabalin may lead to balance disorders, lack of coordination, and increased risk of falling trauma.

In contrast to the studies mentioned above, the current study was a postoperative trial. The etiology of postoperative cognitive decline is not well elucidated, although multiple factors have been proposed, including surgery, use of analgesics, inflammation, sleep disturbances, comorbidity, and age.^37,38 

There is evidence that inflammation and pain interfere with cognitive function.³⁹  In our study population, pain intensities were generally low and similar between the treatment groups and did not significantly affect the cognitive test results. Preclinical and clinical studies suggest that the gabapentinoids alleviate pain by actions at both spinal and supraspinal sites.^40,41  Furthermore, supraspinal nociceptive processing and cognitive processing are assumed to be closely linked, with common neurotransmitters and anatomic structures.⁴²  The drugs used for multimodal pain treatment may therefore interact with both nociceptive and neurocognitive circuitry.⁴²  The evidence that opioids per se impair postoperative cognitive performance is scarce⁴³ ; however, studies show that opioids disrupt sleep architecture, which may indirectly lead to reduced cognitive performance.⁴⁴  Nevertheless, in a previous study in healthy volunteers investigating the cognitive effects of 10 mg of oxycodone orally, the results revealed a significant decline in simple and sustained attention and working and verbal memory compared to baseline.⁴⁵  In the present study, inhibition and working memory were significantly affected when pregabalin was combined with opioids compared with opioids alone, even though the opioid dose required was lower in the pregabalin group compared to the placebo group. Hence, there is reason to believe that the combination of pregabalin and opioids may cause additive or even supraadditive cognitive impairments compared with the use of each drug alone.

Age, educational level, and prehospital mental state are all independent factors linked to postoperative cognitive decline.⁴⁶  Consequently, particularly susceptible groups, such as older patients and patients with different neurologic and psychiatric disorders, may be more sensitive to drugs like pregabalin negatively affecting cognition and coordination. Moreover, reduced renal function in older adults may require reduced dosage, as more than 89% of the drug is eliminated renally. The participants in the current study were all young or middle-aged kidney donors, and educational status and Norwegian Adult Reading Test scores were comparable between the treatment groups. Nevertheless, the results indicate that even younger patients may be negatively affected by pregabalin.

There are some limitations to this study. First, this is an exploratory study in which we examined the effect of perioperative pregabalin on performance on six different cognitive tests without correcting for multiple testing.⁴⁷  When P values were corrected for multiple outcome variables using the Benjamini and Hochberg method⁴⁸  (software R, version 3.4.0⁴⁹ ), the difference in the stop-signal task stop-go test disappeared (P = 0.06), whereas the difference in spatial working memory within error test was still significant (P = 0.012). Thus, our results require replication. Furthermore, the clinical relevance of our findings remains to be determined. We did not measure patients’ self-reports of cognitive functioning. This could have supplemented our data.

In our study, only one dose of pregabalin was used; hence, the validity of our results is restricted to this dose and cannot be extrapolated to other doses. However, the dose of 300 mg of pregabalin daily used in this study is consistent with the recommended dose for the treatment of acute postoperative pain.^1,50  Moreover, pregabalin was only administered on the day of surgery and the day immediately after surgery; thus, the results are not valid for longer perioperative treatment periods. Some patients (seven in the placebo group and one in the pregabalin group) denied testing 24 h after surgery because of pain and nausea. Thus, our data are somewhat incomplete, which may have influenced the results. Finally, we did not obtain blood samples to measure pregabalin plasma concentrations; thus, potential pharmacokinetic interactions between pregabalin and ketobemidone cannot be completely eliminated. A pharmacokinetic drug–drug interaction is unlikely, however, because pregabalin is neither metabolized by nor bound to plasma proteins affecting the cytochrome P450 liver system.²  In conclusion, this exploratory study found that perioperatively administered pregabalin negatively affects executive functioning by impairing working memory and inhibition, measured as neuropsychological task performance.

The authors thank the many patients involved in this study; Leiv Arne Rosseland, M.D., Ph.D., Department of Anesthesia, Division of Emergencies and Critical Care, Oslo University Hospital, Rikshospitalet, Oslo, Norway, for kind assistance with randomization; and the surgical team, all affiliated with the Department of Transplantation Medicine, Oslo University Hospital, Rikshospitalet, Oslo Norway, for excellent support and collaboration – the surgeons for performing the operative procedures, and the nurses for drug handling and assistance with the registration of pain scores and adverse effects.

Supported by a scholarship from the University of Oslo, Norway, and funds from Inven2, a nonprofit research organization owned by Oslo University Hospital, Oslo, Norway (to Dr. Myhre).

The authors declare no competing interests.

Full protocol available at: marianne.myhre@medisin.uio.no. Raw data available at: marianne.myhre@medisin.uio.no.