The impact of intraoperative N-acetylcysteine on opioid consumption following spine surgery: a randomized pilot trial

Abstract

Aim:

N-acetylcysteine (NAC) decreases inflammation and could augment perioperative analgesia.

Materials & methods:

This prospective pilot trial examined postoperative opioid consumption at 12 h following intraoperative NAC. In phase I, 20 adults scheduled for posterior spine surgery were randomized to NAC (0, 50, 100 and 150 mg/kg) to determine the optimal dose. In phase II, 30 patients were randomized to placebo or NAC (150 mg/kg). Opioid consumption, pain ratings and time to opioid rescue were recorded.

Results:

Postoperative opioid consumption was reduced in the NAC group 19.3% at 12 h and 20% at 18 and 36 h. Opioid consumption was reduced 22–24% in the NAC group at all times after adjusting for intraoperative opioid administration. NAC subjects reported lower pain scores relative to placebo.

Conclusion:

Subjects randomized to NAC consumed less postoperative opioids and reported less pain versus placebo. Larger randomized controlled trials are needed to further evaluate NAC for analgesia.

Clinical Trial Registration: NCT04562597 (ClinicalTrials.gov)

Plain language summary

N-acetylcysteine (NAC) is a powerful anti-inflammatory drug used to treat some types of poisoning. It could help pain for patients after surgery. This study looked at how much pain medicine patients needed after back surgery when they received NAC or no drug (placebo). In the first 20 patients, people randomly received placebo or a small, medium or large dose of NAC (0, 50, 100, and 150 mg/kg) with five patients in each group. Since there were only a small number of patients, it was difficult to see any definite differences, and the next 30 patient patients randomly received placebo or the large dose of NAC (150 mg/kg). Patients that were given NAC received 16–22% less opioids in the first 2 days after surgery compared with those that were given placebo. NAC patients also took longer to request pain medications after surgery and reported lower pain scores in the first 2 days after surgery relative to placebo.

Tweetable abstract

N-acetylcysteine (NAC) decreases inflammation and could augment analgesia. Patients undergoing spine surgery received NAC (150 mg/kg) or placebo. NAC reduced postoperative opioid consumption by 16–22% in the first 6–48 h versus placebo.

Postoperative pain remains an acute problem associated with increased postoperative complications [1]. As the opioid epidemic has gained notoriety, opioid sparing strategies have garnered more interest to provide analgesia without addictive potential, opioid related side effects and elevated costs. Anesthesiologists and surgeons must minimize risk of overexposure to opioids while still appropriately controlling the discomfort associated with invasive procedures.

N-acetylcysteine (NAC), a precursor to the amino acid L-cysteine and the antioxidant glutathione, is a glutamate modulator and promotes homeostatic glutamate regulation by astrocytes [2]. The sulfhydryl group within the NAC molecule directly scavenges reactive oxygen species [3], modulates the redox state of the NMDA and AMPA receptors [3] and inhibits the nuclear factor kappa-light-chain-enhancer of activated B cells to modulate cytokine synthesis creating an anti-inflammatory and anti-nociceptive effect [4].

NAC is best known as the treatment for hepatotoxicity from acetaminophen toxicity [5], but it may also have analgesic effects. NAC improved analgesia in rodent models of painful diabetic neuropathy [6]. Similarly, NAC treatment reduced pain ratings to laser stimuli in humans (1.2 g oral) and mice (100 mg/kg) [7]. NAC also reduced postoperative morphine consumption in patients after knee ligamentoplasty [8]. Mechanistically, preclinical literature suggests NAC improves analgesia through activation of type 2 neuronal metabotropic glutamate receptors [6,9] and matrix metalloproteinase inhibition [10]. Thus, these data indicate that NAC may be a novel and safe analgesic adjunct in the perioperative period.

We posited that patients randomized to intravenous (IV) NAC would consume less opioids after posterior spine surgery, involving 2–4 vertebrae, compared with those randomized to placebo. The primary aim of this randomized exploratory trial was to estimate the difference in opioid consumption between placebo and the selected optimal NAC dose.

Materials & methods

Study design

The trial was conducted in accordance with the original protocol, and written informed consent was obtained from all subjects. This manuscript adheres to applicable Consolidated Standards of Reporting Trials guidelines. This single-center, prospective, double blinded clinical trial was approved by the institutional review board (IRB; Pro00099062) and registered (www.clinicaltrials.gov/; NCT04562597; 24 September 2020; PI: Sylvia H. Wilson) prior to patient enrollment. The IRB determined that an investigational new drug application was not required.

Participants & treatment groups

Eligible subjects were 18 years and above and scheduled for elective posterior spine surgery involving 2–4 vertebrae of the thoracic, lumbar or sacral spine. Exclusion criteria included: weight under 40 kg; unable to provide written, informed consent; history of an adverse or anaphylactoid reaction to acetylcysteine; pregnancy; intraoperative neuromonitoring (necessitating remifentanil); blood clotting deficiency; or active asthma, wheezing, or use of inhaled bronchodilators.

After providing written informed consent and meeting inclusion criteria, subjects were assigned an enrollment number and randomized using computer generated simple randomization, created prior to study initiation by a computer, to one of four IV NAC groups in phase I (0 mg/kg [placebo; 0.45% NaCl], 50, 100, or 150 mg/kg) or to placebo or NAC in phase II. Randomization information was kept in the Investigational Drug Services pharmacy, which prepared all study medications in identical 250 ml infusion bags. All patients, research staff, and care providers were blinded to the allocation.

IV NAC or placebo infusion was initiated after general anesthetic induction and infused over 60 min. Patients were monitored intraoperatively while receiving NAC and in the postoperative anesthesia care unit (PACU) for any adverse events from study medications (e.g., flushing, dyspnea).

Anesthetic care was standardized and based on our institutional protocol. Unless contraindicated, patients were prescribed preoperative oral acetaminophen (1000 mg). No other preoperative analgesics were given. General anesthesia included ketamine (0.5 mg/kg), lidocaine (1 mg/kg), dexamethasone (10 mg/kg) and ondansetron (4 mg). Dexmedetomidine (max 0.25 g/kg) and fentanyl (max 2 g/kg) were given as needed by the intraoperative anesthesia team. Ketorolac, hydromorphone, morphine and remifentanil were not used intraoperatively. PACU orders (hydromorphone 0.2 mg IV every 8 min for pain and haloperidol 1 mg IV for nausea) and postoperative floor orders (Supplementary 1) were standardized.

Outcomes

The primary outcome was reduction in opioid consumption at 12 h postoperative between placebo and NAC. Opioids were converted to IV morphine milligram equivalents (MME) [11]. The recorded anesthesia stop marked time zero for postoperative data collection.

Phase I

The goal of phase I was to identify the optimal dose, defined as the dose of NAC for which opioid consumption was reduced >20%. 20 patients were randomized to one of four doses of NAC (0, 50, 100, and 150 mg/kg; Supplementary 2). Mean cumulative opioid consumption was estimated within each treatment group to determine the optimal dose for phase II of the trial. If 20% reduction was not achieved, the optimal dose would be NAC 150 mg/kg.

Phase II

In phase II, an additional 30 patients were randomized to placebo (n = 15) or the optimal NAC dose (n = 15) to estimate the difference in opioid consumption. Additional data included demographics, preoperative opioid consumption pain ratings using the visual analog scale (VAS) and numeric rating scale (NRS), time to first opioid rescue dose, procedure duration, PACU duration, time to hospital discharge and postoperative nausea and itching. Demographic data (age, sex, self-reported race), weight (kg), body mass index (BMI; kg/m²) and preoperative opioid consumption (yes or no) were collected after completion of informed consent. VAS pain scores were recorded preoperatively and immediately postoperative in the PACU by having the patient mark their current pain level on a 100 mm line (0 = no pain, 100 = worst pain imaginable). NRS pain scores were collected per nursing routine in PACU and the floor for up to 48 h postoperatively. NRS scores are assessed routinely in PACU approximately every 15 min until stable and then increased to floor intervals of every 4–6 h unless pain medication requested. Postoperative antiemetic and antipruritic use were captured as markers of postoperative nausea and itching, respectively.

No changes were made to the study protocol or outcomes after study commencement.

Statistical analysis

The goal of phase II of this pilot trial was to estimate efficacy of NAC for primary and secondary outcomes; thus, no formal hypothesis testing was conducted. An a priori sample size calculation was conducted to determine the precision of the estimated mean difference in opioid consumption at 12 h between treatment groups in phase II for a fixed sample size. A sample size of 20 subjects per group (5 from phase I and 15 from phase II) was needed to estimate a 95% confidence interval for the mean difference in opioid consumption with a width of +0.64 standard deviations from the mean.

Descriptive statistics were calculated for patient and procedural characteristics. Mean differences between treatment groups in opioid consumption at different postoperative times, including the primary outcome of 12 h postoperative, were estimated from a linear mixed model including fixed effects for group, postoperative time and their interaction and a random patient effect to account for correlation between measures collected on patients over time. Model assumptions were checked graphically and transformations were considered if necessary.

Average NRS pain score over postoperative time within and between groups was estimated from a linear mixed model including fixed effects for treatment group, postoperative time at pain score measurement, and their interaction as well as a random subject effect to account for correlation between pain scores collected on the same patient over time. Model assumptions were checked graphically, and transformations were considered if necessary. Confidence intervals for the median time to first opioid rescue dose, PACU and hospital length of stay were estimated using the Brookmeyer-Crowley approach using the log-log complement transformation [12]. Confidence intervals for the percent of participants experiencing nausea and itching were calculated using Clopper-Pearson exact confidence intervals for binomial proportions. All analyses were conducted in SAS v. 9.4 (SAS Institute, NC, USA).

Results

Between 20 January 2021 and 20 May 2022, 21 patients consented to participate in phase I (one excluded for neurological monitoring) and 35 consented to participate in phase II (five excluded for neurological monitoring [n = 2], surgery cancellation [n = 1], anticoagulation concerns [n = 1] and shortage of 0.45% NaCl [n = 1]). All patients excluded were identified after randomization but before any medications were administered. As phase I did not provide clear opioid reduction for any NAC dose and noted a high degree of variability within each dose, 150 mg/kg was used as the optimal NAC dose for the remainder of the study and an additional 30 subjects were recruited and randomized for both placebo (n = 15) and NAC (n = 15). The study ceased after 20 total subjects were enrolled in both the placebo and NAC (150 mg/kg) arms (Figure 1). All subjects were analyzed in their assigned groups.

Figure 1. . Consolidated Standards of Reporting Trials (CONSORT) flow diagram for subjects randomized to placebo or N-acetylcysteine (NAC) 150 mg/kg.

Descriptive statistics for placebo and NAC (150 mg/kg) participants are shown in Table 1. One patient in the NAC group required extension of planned L2-4 surgery to also revise prior L5-S1 surgery. Participants in the NAC group (150 mg/kg NAC) were on average older, with a lower BMI and lower preoperative pain score relative to the placebo group.

Table 1. . Descriptive statistics for patient and procedural characteristics for placebo and N-acetylcysteine (150 mg/kg) participants.

Characteristic	Placebo (n = 20)	150 mg/kg NAC (n = 20)
Age (years), mean ± SD	58.6 ± 12.8	67.4 ± 8.82
Sex (male), n (%)	11 (55.0)	13 (65.0)
Self-reported race, n (%)
White	15 (75.0)	16 (80.0)
Black	4 (20.0)	4 (20.0)
Other	1 (5.00)	0 (0.00)
BMI (kg/m2), mean ± SD	33.9 ± 7.79	29.1 ± 5.10
Home opioid use (yes), n (%)	5 (25.0)	5 (25.0)
Preoperative VAS (mm), median (IQR)	59.5 (28.5)	28 (30)
Pre & intraoperative IV MME, mean ± SD	14.3 ± 6.81	16.2 ± 7.04
Procedure information
Number of vertebrae instrumented, n (%)     2     3     4     5	10 (50) 8 (40) 2 (10) 0 (0)	11 (55) 8 (40) 0 (0) 1 (5)
Decompression/Laminectomy, n (%)	9 (45)	9 (45)
Fusion, n (%)	13 (65)	11 (55)
Duration (min), mean ± SD	247.0 ± 86.4	269.8 ± 91.8

BMI: Body mass index; IQR: Interquartile range; IV MME: Intravenous morphine milligram equivalents; NAC: N-acetylcysteine; SD: Standard deviation; VAS: Visual analogue scale.

Opioid consumption

The primary outcome was postoperative opioid consumption at 12 h.

The mean difference (95% CI) in opioid consumption at 12 h between the placebo and 150 mg/kg NAC group was 3.75 (-8.39, 15.9) in IV MME. Table 2 presents the mean cumulative opioid consumption for both groups, the mean difference between groups, and the percent decrease in opioid consumption between groups in 6-hour postoperative intervals up to 48 h. In the NAC group, opioid consumption decreased 20% or more at 18 and 36 h postoperative. Opioid consumption at the other time points investigated were not greater than 20%, but a 19% reduction was observed at 12, 30, 42, and 48 h postoperative.

Table 2. . Postoperative opioid consumption in intravenous morphine milligram equivalents.

Postoperative time (h)	Placebo (n = 20) mean (95% CI)	150 mg/kg NAC (n = 20) mean (95% CI)	Difference (Placebo – NAC) mean (95% CI)	% Decrease (Placebo-NAC) %
6	14.6 (5.97, 23.1)	12.0 (3.43, 20.6)	2.55 (-9.59, 14.7)	17.5
12	19.4 (10.8, 28.0)	15.6 (7.02, 24.2)	3.75 (-8.39, 15.9)	19.3
18	25.0 (16.4, 33.6)	19.9 (11.3, 28.5)	5.09 (-7.05, 17.2)	20.4
24	29.6 (21.0, 38.2)	24.7 (16.2, 33.3)	4.86 (-7.26, 17.0)	16.4
30	34.0 (25.4, 42.6)	27.2 (18.7, 35.8)	6.77 (-5.37, 18.9)	19.9
36	37.9 (29.4, 46.5)	29.7 (21.1, 38.3)	8.22 (-3.92, 20.4)	21.7
42	40.9 (32.3, 49.5)	33.1 (24.5, 41.7)	7.78 (-4.36, 19.9)	19.0
48	43.3 (34.7, 51.9)	34.9 (26.4, 43.6)	8.31 (-3.75, 20.4)	19.2

Data are reported as mean (95% CI) or mean difference (95% CI) estimated from a linear mixed model of opioid consumption over time including main effects for treatment group, postoperative time, the interaction between treatment group and postoperative time, and a random subject effect.

CI: Confidence intervals; NAC: N-acetylcysteine.

A sensitivity analysis was done that included intraoperative opioids in the model. The mean difference (95% CI) in opioid consumption at 12 h between the placebo and NAC groups adjusted for intraoperative opioid consumption was 4.91 (-6.36, 16.2) in IV MME, which was a 23.7% reduction in the NAC group (Supplementary 3). Postoperative opioid consumption decreased in NAC patients 21.9–24.6% at other examined times.

Additional outcomes

Table 3 presents additional studied outcomes. Participants in the NAC treatment arm had a 59% longer median time to first opioid rescue dose relative to those in the placebo arm (25.5 vs 16.0 min). Participants receiving NAC had similar PACU VAS pain scores, PACU length of stay, and hospital length of stay relative to those who received placebo. No adverse events or side effects were noted for any subjects.

Table 3. . Additional outcomes.

Outcome	Placebo (n = 20)	150 mg NAC (n = 20)	Difference (placebo-NAC)
Postoperative anesthesia care unit
Time to first opioid (min), median (95% CI)†	16.0 (5.0, 33.0)	25.5 (6.0, 84.0)	-7.0 (-35.0, 7.0)
VAS pain score (mm), mean (95% CI)	65.5 (53.8, 77.2)	64.8 (52.4, 77.2)	0.7 (-16.9, 18.2)
Length of stay (min), median (95% CI)†	76.5 (55.0, 91.0)	76.0 (65.0, 92.0)	-3.0 (-20.0, 14.0)
Hospital length of stay (days), median (95% CI)†	1.98 (1.19, 2.39)	2.04 (1.12, 2.26)	-0.06 (-0.96, 0.84)
Nausea, percent (95% CI)‡
0–24 h	20.0 (5.7, 43.7)	35.0 (15.4, 59.2)	-15.0 (-42.3, 12.3)
24 h to discharge	15.0 (3.2, 37.9)	25.0 (8.7, 49.1)	-10.0 (-34.6, 14.6)
Itching, percent (95% CI)‡
0–24 h	15.0 (3.2, 37.9)	0.0 (0.0, 16.8)	15.0 (-0.7, 30.7)
24 h to discharge	10.0 (1.2, 31.7)	0.0 (0.0, 16.8)	10.0 (-3.2, 23.2)

^†Confidence intervals estimated using the Brookmeyer-Crowley approach using the log-log complement transformation [13], and Hodges-Lehmann estimate of location shift was used to estimate the differences of medians.

^‡Confidence intervals estimated using the Clopper-Pearson exact confidence intervals for binomial proportions.

CI: Confidence intervals; NAC: N-acetylcysteine; VAS: Visual analogue scale.

Figure 2 presents postoperative pain scores collected throughout the postoperative period. Postoperative patient reported NRS pain scores decreased in both groups with increasing time (mean decrease of 0.41 units per 12 h, 95% CI: 0.24–0.59). Patient reported postoperative pain was on average 1.1 units lower in the NAC treated group compared with placebo (mean difference in pain scores for placebo versus NAC (95% CI): -1.07 (-2.2, 0.07).

Figure 2. . Mean postoperative reported pain over time by treatment group.

Lines represent the estimated mean numeric pain scores for placebo (blue) and NAC (red) treatment groups over time estimated from the linear mixed model of postoperative pain over time. Shaded regions represent the 95% Wald confidence limits.

NAC: N-acetylcysteine; NRS: Numeric rating scale.

Discussion

In this study to identify the optimal dose of intraoperative IV NAC for perioperative analgesia, opioid reduction did not meet the desired 20% reduction compared with placebo at 12 h postoperative. However, a 20% opioid reduction occurred at 18 and 36 h and approached the 20% reduction goal at nearly every examined time point. Concurrently, NRS pain scores in NAC patients remained 1.1 units lower on average compared with placebo patients.

One prior publication has examined the impact of NAC (150 mg/kg IV) on opioid consumption and pain scores following elective surgery. In a randomized trial of 60 patients undergoing inguinal hernia repair under general anesthesia, pain scores did not differ over the 84 h study period [14]. Our study population differs from this prior study as 25% of our placebo (n = 5) and 25% of the NAC group (n = 5) identified as having chronic pain and chronic home opioid use. While our heterogenous study population suggests our findings may be generalizable, our data suggests that NAC may have an analgesic impact on postoperative opioid consumption and potentially improve pain scores later in the postoperative period.

The anti-inflammatory properties of NAC, noted by prior pre-clinical [15,16] and clinical [17,18] studies, may partially explain the analgesic potential of NAC. In 150 patients scheduled for coronary artery bypass graft surgery, subjects randomized to NAC (50 mg/kg intraoperative and postoperative day 1 and 2) had a reduction in postoperative atrial fibrillation and a decrease in the inflammatory marker high sensitivity C-reactive protein (p < 0.001) compared with placebo [17]. However, NAC did not decrease the incidence of atrial fibrillation or inflammatory markers in patients undergoing non-cardiothoracic surgery receiving amiodarone infusions and randomized to postoperative NAC (50 mg/kg bolus and 100 mg/kg infusion over 48 h) or placebo [19]. In a separate study, 60 patients requiring preoperative parenteral nutrition and postoperative intensive care admission for colon surgery were randomized to receive NAC (100 mg/kg preoperative and 50 mg/kg daily for 48 h) or placebo [18]. Although NAC did not decrease the incidence of systemic inflammatory response syndrome, a significant reduction in alanine aminotransferase, a neuroinflammatory marker of hepatocyte injury, was observed in the NAC group. Markedly, both publications [17,18] administered NAC through postoperative day 2 since postoperative inflammation peaks on postoperative day 2–3 [20,21]. While we did not measure laboratory inflammatory markers, the decrease in pain scores observed in our NAC group after 24 h may indicate that the anti-inflammatory impact of NAC may not have a measurable impact until inflammation peaks days later. This suggests a potential analgesic benefit beyond the first 24 h postoperative.

As patients undergoing spinal surgery often have chronic pain and are at increased risk for acute and persistent pain [22], future NAC studies should investigate patients with preoperative opioid use or chronic pain by prolonging NAC administration and patient follow-up farther into the postoperative period. Clinically, NAC has demonstrated potential benefits in patients with chronic painful conditions. In a study of 113 patients with diabetes suffering from diabetic peripheral neuropathy and randomized to pregabalin with oral NAC or pregabalin with placebo over an 8-week period, mean pain scores and mean sleep interference scores were decreased in the NAC group (p < 0.001 for both) [23]. Further, oral NAC reduced the number of vaso-occlusive crisis days for patients with sickle cell anemia compared with patients randomized to placebo [13]. While our study only followed patients for 48 h, prior publications found that patients with chronic pain and randomized to intraoperative ketamine for spine surgery may have decreased opioid consumption and improved analgesia at six weeks [24] and one year [22] postoperative compared with patients randomized to placebo. Further, NAC treatment significantly suppressed heroin seeking in a rodent model of heroin addiction and relapse [25]. However, a meta-analysis of nine studies (n = 863) investigating oral NAC in patients with a variety of chronic painful conditions concluded that there was insufficient evidence and larger randomized clinical trials were needed [26]. Thus, future NAC studies should examine opioid consumption, pain scores, and the continued need for opioids at weeks to months into the postoperative period.

It is notable that there were no serious adverse events. The half-life of NAC is reported to be 6.25 h, and clearance is both renal and nonrenal, with the most common reported side effects being nausea, vomiting and diarrhea; although, these are more commonly observed with oral administration [27]. However, in a meta-analysis of nine studies [26], only one trial noted the oral NAC group to report increased gastrointestinal concerns [13], while the remaining eight studies found no differences between NAC and controls. NAC IV administration was also chosen for this study due to the poor bioavailability of oral NAC (approximately 9.1%) compared with IV (100%) [28]. IV doses of NAC ranging from 50–150 mg/kg [14,17,18] are generally well tolerated, and may increase to 300 mg/kg for acetaminophen poisoning or 600 mg/kg for chronic obstructive pulmonary disorder treatment [29]. Our low rate of adverse events contrasts with the findings of Mulkens et al. who noted adverse events in over 50% of the NAC group including high rates of flushing (53.8%) without or with urticaria (23.1%) and dyspnea (15.4%) [14]. This is likely attributed to their 15 min preoperative administration, which contrasts our 60 min intraoperative administration time and lack of adverse events. In other trials utilizing 50–100 mg/kg IV NAC, minimal side effects were also reported, with one participant experiencing tachycardia and another hypotension [17,18]. Vomiting has been reported at doses of 150 mg/kg [27]. Finally, as NAC is used for acetaminophen toxicity, its impact on acetaminophen's analgesic effect is a consideration. However, in preclinical studies, NAC demonstrated dose-dependent improvements in the analgesic effect of acetaminophen [30].

Limitations

Our study is not without limitations. Opioid consumption was highly variable, likely a consequence of our pilot sample size. A larger study with a greater sample size may reduce this variability and elucidate if NAC has further value for perioperative analgesia. Despite randomization, our placebo group was younger, had a higher average BMI, and higher preoperative VAS pain scores at rest compared with the 150 mg/kg NAC group. This could lead to some potential confounding. While VAS scores were collected preoperative and in PACU, postoperative NRS scores were dependent on nurse chart times and varied in quantity of data points for each subject in the first 48 h. Further, data was only collected up to 48 h postoperative, and this limits our ability to comment more on the usefulness of NAC as an agent for patients with chronic pain.

Conclusion

In this exploratory, randomized, placebo-controlled pilot trial, a 20% reduction in opioid consumption at 12 h was not achieved with intraoperative NAC (150 mg/kg) versus placebo. However, postoperative opioid consumption in patients receiving intraoperative NAC was decreased between 16–20% at all observed time points, and postoperative pain scores in the NAC group were lower relative to the placebo group. After adjusting for intraoperative opioid administration, postoperative opioid consumption was further decreased in NAC patients (21.9–24.6%). These results suggest there is good evidence for conducting a larger randomized controlled trial to formally test whether NAC decreases opioid consumption and postoperative pain using the effect sizes observed in this study to design such a trial.

Summary points.

• N-acetylcysteine (NAC) decreases inflammation and could augment perioperative analgesia after spine surgery.

• In this randomized, placebo controlled, exploratory pilot trial, patients undergoing spine surgery involving 2–4 vertebral levels of the thoracic, lumbar or sacral spine were randomized to one dose of intraoperative NAC or placebo.

• Patient demographics, opioid consumption, pain ratings, time to opioid rescue, procedure duration and length of stay were recorded.

• Mean differences between groups in opioid consumption were estimated from a linear mixed model including fixed effects for group, postoperative time and their interaction.

• Postoperative opioid consumption in patients receiving intraoperative NAC was decreased 16%–20% at all observed time points compared with patients receiving placebo and decreased 22–24% after adjusting for intraoperative opioid administration.

• Mean postoperative pain scores were lower in the NAC group relative to the placebo group.

• Participants in the NAC treatment arm had a 59% longer median time to first opioid rescue dose relative to those in the placebo arm.

• NAC may have an analgesic impact on postoperative opioid consumption and potentially improve pain scores later in the postoperative period. However, larger randomized controlled trials are needed.