S-ketamine in patient-controlled analgesia reduces opioid consumption in a dose-dependent manner after major lumbar fusion surgery: A randomized, double-blind, placebo-controlled clinical trial

Background Spinal fusion surgery causes severe pain. Strong opioids, commonly used as postoperative analgesics, may have unwanted side effects. S-ketamine may be an effective analgesic adjuvant in opioid patient-controlled analgesia (PCA). However, the optimal adjunct S-ketamine dose to reduce postoperative opioid consumption is still unknown. Methods We randomized 107 patients at two tertiary hospitals in a double-blinded, placebo-controlled clinical trial of adults undergoing major lumbar spinal fusion surgery. Patients were randomly allocated to four groups in order to compare the effects of three different doses of adjunct S-ketamine (0.25, 0.5, and 0.75 mg ml-1) or placebo on postoperative analgesia in oxycodone PCA. Study drugs were administered for 24 hours postoperative after which oxycodone-PCA was continued for further 48 hours. Our primary outcome was cumulative oxycodone consumption at 24 hours after surgery. Results Of the 100 patients analyzed, patients receiving 0.75 mg ml-1 S-ketamine in oxycodone PCA needed 25% less oxycodone at 24 h postoperatively (61.2 mg) compared with patients receiving 0.5 mg ml-1 (74.7 mg) or 0.25 mg ml-1 (74.1 mg) S-ketamine in oxycodone or oxycodone alone (81.9 mg) (mean difference: -20.6 mg; 95% confidence interval [CI]: -41 to -0.20; P = 0.048). A beneficial effect in mean change of pain intensity at rest was seen in the group receiving 0.75 mg ml-1 S-ketamine in oxycodone PCA compared with patients receiving lower ketamine doses or oxycodone alone (standardized effect size: 0.17, 95% CI: 0.013–0.32, P = 0.033). The occurrence of adverse events was similar among the groups. Conclusions Oxycodone PCA containing S-ketamine as an adjunct at a ratio of 1: 0.75 decreased cumulative oxycodone consumption at 24 h after major lumbar spinal fusion surgery without additional adverse effects.

Multimodal analgesia targets different pain signaling pathways by combining two or 67 more analgesic modalities, aiming at additive or even synergistic analgesic effect [3]. 68 Multimodal analgesia has proven feasible after major spinal fusion surgery in an effort to 69 optimize pain relief while minimizing opioid-related adverse effects [4,5].  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021. ; https://doi.org/10.1101/2021.01.22.21250352 doi: medRxiv preprint an adjunct analgesic has been documented in several clinical trials and meta-analyses 79 [12]. A recent review article concluded that combining ketamine with an opioid in IV- 80 PCA during the postoperative period has a beneficial effect on analgesia and opioid 81 consumption [13]. In orthopedic surgery, earlier trials found negative or unclear results 82 with opioid-ketamine PCA [14][15][16]. However, studies' vast heterogeneity and small 83 sample sizes have failed to establish a possible a possible dose-responsiveness. Thus, 84 the optimal ketamine to opioid ratio in intravenous PCA is yet to be elucidated. 85 We hypothesized that adjunct S-ketamine with oxycodone in an intravenous PCA is  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) A randomized, double-blind, controlled dose-response study design was used (Fig 1).

136
An independent statistician created a computer-generated randomization list. The list 137 was sent to the local hospital pharmacy, which took care of assignment. Coded PCA 138 reservoirs with no other markings were delivered to the operation room by the pharmacy 139 on the day of surgery to ensure double blinding. Patients, researchers, and clinical staff 140 were blinded to group allocation.

141
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)   after wound closure (levobupivacaine 2.5 mg kg -1 ; Chirocaine 2.5 mg ml .1 , AbbVie S.r.l., 163 Campoverde di Aprilia, Italy) as per hospital routines.

164
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  When NRS was 4 or lower, the PCA oxycodone dose was decreased to 1 mg (G1-G4).

171
The study PCA dosing continued for 24 h from the end of surgery, after which the PCA 172 cassettes were changed in all study groups and contained only 1 mg/ml oxycodone 173 thereafter in all study groups.

174
The total duration of PCA treatment was three days. Postoperative nausea and vomiting  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021. All clinical patient data were collected on individual case report forms. All data were 188 subsequently transferred to electronic format for exploratory data analysis. The authors approved the statistical analysis plan before the analyses began.

201
Explorative data analysis was conducted before statistical inference by plotting and 202 tabulating the data. Normality assumptions were tested before analysis, using probit 203 plots and the Shapiro-Wilk W-test. Levene's test was used to evaluate homogeneity of 204 variances.

205
The primary outcome measure, cumulative opioid consumption during the first 24 h is 206 presented as both median and interquartile range (IQR; Q1-Q3) and mean (SD).

207
Cumulative opioid consumption at 24, 48, and 72 h was analyzed by using a linear 208 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.    (Table 1).

250
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021. ;  Cumulative oxycodone consumption was highest in group G1 and lowest in G4 (Fig 2,   261 Table 2). The median total oxycodone consumption during the first 24 h after surgery     (Table 3) 289  There was a small yet statistically significant mean change in postoperative NRS 298 measured at rest over the first 24 h between groups G4 and G1 (standardized effect 299 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021. ; https://doi.org/10.1101/2021.01.22.21250352 doi: medRxiv preprint reported NRS was smaller in groups G2-G4 than group G1 at the end of PACU 315 treatment, the finding was not statistically significant (Fig 3B).  . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021   Opioid-related adverse effects 331 Opioid-related adverse events during the study are summarized in Table 4. Nausea and 332 vomiting were the most frequent adverse events. At 24 h after PCA start, 16% of 333 patients (n = 16) reported nausea. Increasing ketamine dose seemed to increase the 334 incidence of PONV, but the changes were not statistically significant. Similarly, no 335 differences were seen at the end of 48-and 72-h follow-up. We analyzed these adverse 336 effects further with logistic regression. Our results indicate that preoperative weak opioid 337 use increased the incidence of PONV significantly (odds ratio: 9.23, 95% CI: 1.4-75).

341
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021. ; https://doi.org/10.1101/2021.01.22.21250352 doi: medRxiv preprint At 24 h after PCA start, 16% of patients (n = 16) reported pruritus (Table 4). The PCA 342 treatment groups did not show statistically significant differences at any of the three time 343 points (24, 48, or 72 h after the start of PCA). Similarly, the incidence of pruritus did not 344 change after changing to normal PCA with oxycodone only.

345
Nine patients (9%) reported nightmares or unpleasant dreams (Table 4). The groups 346 showed no difference in this regard, and the incidence did not change after changing to 347 normal PCA with oxycodone only. Age, sex, weight, chronic pain, and prior use of 348 gabapentinoids had no effect on the results.

349
Other adverse effects 350 There were no severe adverse effects. None of the patients developed respiratory 351 insufficiency requiring invasive or non-invasive ventilation during the 72-h study period.

352
No rescue medications were required during the study.

353
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.   However, the opioid:ketamine ratio in IV-PCA in previous studies was heterogeneous, 414 ranging from 1:0.5 to 1:2.5, and these studies were also heterogeneous in regard to the 415 opioid used, anesthesia methodology, type of surgery, patient population, and use of 416 racemic-or S-ketamine [35,36].

417
Our study has limitations. Our study was designed to detect a difference in 24 h 418 cumulative oxycodone consumption, which was the primary outcome. The beneficial 419 effect of adjunct S-ketamine in an oxycodone IV-PCA in reducing the 24 h oxycodone 420 consumption did not correlate with a reduction in opioid-related side effects, but this 421 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) analgesic consumption has been analyzed with point estimates or determining area 431 under the curve, but both of these methods introduce considerable bias. Statisticians 432 recommend using a model-based approaches, which has been used here. Therefore, 433 both mean (SD) and median (IQR) have been reported. From a clinical perspective, it 434 may be unrealistic to achieve a totally pain-free state after major surgical trauma 435 following surgery such as instrumented lumbar spinal fusion. We think that IV-PCA 436 enables the patient to titrate the opioid to reach a certain individual, tolerable pain 437 intensity level. Therefore, we consider that changes in cumulative opioid consumption 438 could serve as an acceptable surrogate for changes in postoperative pain. However, we 439 also evaluated the outcome by using risk ratios for pain intensity NRS >3 as a surrogate 440 for perioperative pain experience, and adjunct S-ketamine seemed to decrease the 441 likelihood of pain exceeding NRS >3.

442
Our study adds new data considering the optimal dose of adjunct S-ketamine to an 443 oxycodone IV-PCA after lumbar spinal fusion surgery. Because the oxycodone:S-444 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted February 16, 2021. ; https://doi.org/10.1101/2021.01.22.21250352 doi: medRxiv preprint ketamine ratio of 1:0.75 was not associated with increased adverse events, we suggest 445 that future studies aimed at solving the optimal opioid:S-ketamine ratio evaluate that 446 dose and higher, possibly with pharmacokinetic testing to characterize the dose-447 concentration-effect relationship.

448
In conclusion, IV-PCA containing adjunct S-ketamine with oxycodone at a ratio of 1:0.75 449 after major lumbar spinal fusion surgery is effective in decreasing the total oxycodone  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.