Preoperatively assessed o ﬀ set analgesia predicts acute postoperative pain following orthognathic surgery

Objectives: High intensity and longer duration of acute postoperative pain are generally associated with a higher risk of developing chronic postoperative pain. Therefore, it is important to identify the preoperative predictors for acute postoperative pain.Preoperative evaluation of o ﬀ set analgesia (OA) and the PainCatastrophising Scale (PCS) may be potential predictors for acute postoperative pain. This study aimed to investigate the relationship between preoperative OA, PCS, and acute postoperative pain following orthognathic surgery. Methods: Thirty patients (19 females) scheduled to undergo orthognathic surgery were included in this study. OA and PCS wereevaluatedpreoperatively,andthepatientsreportedtheir postoperative pain intensity using the visual analogue scale [0 – 100 mm] until it reached zero (number of days with pain). OA was induced on the dominant forearm via three consecutive painful heat pulses delivered for 5 s (T1=46 ° C), 5 s (T2=47 ° C), and 20 s (T3=46 ° C). Subsequently, the associations between OA, PCS, and the number of days with pain were analysed. Results: The median duration of postoperative pain was 10.3 days. Multiple linear regression analysis showed a significant (p=0.0019) predictive value of OA (p=0.008) for the number of days with pain. The PCS-magni ﬁ cation component was positively correlated with the number of days with pain


Introduction
Severe acute postoperative pain (POP) has been associated with an increased risk of developing chronic POP following various surgeries, including total hip arthroplasty, breast cancer surgery, inguinal herniorrhaphy, caesarean section, and thoracic surgery [1][2][3][4]. Predicting acute POP is crucial as it may enable effective pre-and perioperative treatment to prevent the development of chronic POP.
The Pain Catastrophising Scale (PCS) is the most widely used instrument to assess pain catastrophising in chronic pain studies [4]. According to some studies [4][5][6][7][8], pain catastrophising is predictive of chronic POP; however, little is known about its ability to predict acute POP.
Quantitative Sensory Testing (QST) can be used to assess nerve function, and preoperative QST findings have been linked to the assessment of the development of chronic POP [9,10]. Offset analgesia (OA) is a recently developed QST method that refers to the phenomenon of a disproportional reduction in pain intensity caused by a slight decrease from one temperature to another [11][12][13]. OA has attracted attention in recent years as a method for evaluating the potential aspect of descending pain inhibitory capacity [11][12][13][14][15][16][17]. However, the involvement of peripheral and central mechanisms in OA is currently debated [11][12][13]. Pain-free volunteers show a larger OA response compared with that of those with chronic pain [17,18]. Unlike other QST assessment methods, OA has not been investigated as a predictor for acute POP.
The aim of this study was to investigate whether preoperative OA and PCS could predict acute POP (number of days with pain) following orthognathic surgery.

Methods Participants
This study was approved by the Ethics Committee of Meikai University (A1624, A2113) and was conducted in accordance with the Declaration of Helsinki developed by the World Medical Association. Written informed consent was obtained from all patients before inclusion. In addition, this study was registered with the University Hospital Medical Information Network Clinical Trials Registry (unique ID: UMIN000026719, UMIN000046957) before conducting the research and adheres to the disclosure requirements of the institutional registry.
The inclusion criteria were: (1) scheduled orthognathic surgery, (2) age ≥16 years, and (3) ability to provide informed consent. The exclusion criteria were: (1) presence of psychiatric diseases, (2) use of any pain medication 24 h prior to the investigation, and (3) inability to provide informed consent. The patients were recruited between October 2019 and January 2022.

Preoperative evaluation
Experimental design: OA and PCS evaluations were performed the day before the surgery ( Figure 1). OA evaluation was performed 15 min after PCS evaluation ( Figure 1). All experiments were performed at a constant room temperature (25°C). In addition, patients were interviewed to evaluate POP at 3 months, 6 months, and 1 year postoperatively.
PCS evaluation: The Japanese version of the PCS questionnaire [28] comprising 13 items was used in this study. The participants rated the frequency with which they experienced different pain-related thoughts or feelings on a 5-point Likert scale, where 0 represents 'not at all', and 4 represents 'all the time'. The scores of the three subscales of PCS (rumination, helplessness, and magnification) were also calculated in addition to the sum of all items as a total score.

General anaesthesia and surgical procedure
Orthognathic surgeries were performed under general anaesthesia with endotracheal intubation using balanced anaesthesia with propofol, remifentanil hydrochloride, and rocuronium bromide, and regional anaesthesia with 1 % lidocaine hydrochloride monohydrate containing 1/100,000 adrenaline. Acetaminophen (1,000-2,000 mg) was administered 30 min before the end of the surgery (Figure 1).
Postoperative protocol for pain management and evaluation of surgical pain POP was managed with acetaminophen; 1,000-2,000 mg was administered the day after the surgery. After postoperative day 1 (POD1), 3,000 mg/day (1,000 mg at 8 am, 12 pm, and 6 pm) was administered. Loxoprofen sodium hydrate, additional acetaminophen, or other pain medications were administered if the pain was not well-controlled. The analgesic dose was adjusted and reduced according to the POP level. Depending on the POP, trigeminal nerve block (mandibular and mental nerve block) or local infiltration anaesthesia was performed using levobupivacaine hydrochloride postoperatively.
The POP intensity was evaluated using VAS (VAS-POP). On the operative day, VAS-POP was assessed immediately, 1 h, and 2 h after returning to the ward and at 7 pm and 9 pm. After POD1, VAS-POP was assessed at 6 am, 1 pm, and 7 pm (Figure 1). In addition, VAS-POP was assessed at 8 am, 12 pm, and 6 pm during acetaminophen administration. The administration of acetaminophen was discontinued if VAS-POP was <30/100 or if the patient did not request an additional prescription; this decision was made during the morning rounds.
The duration of acetaminophen administration and the total dose of analgesics administered were recorded. The postoperative ('analgesic requirement period') was defined as 'the final analgesic administration time (day) − the end time of anaesthesia (day)'.
VAS-POP was assessed daily until discharge. If VAS-POP did not reach 0/100 at discharge, patients were requested to record it until it reached 0/100. Furthermore, patients were interviewed to evaluate POP at 3 months, 6 months, and 1 year postoperatively. The number of days until VAS-POP reached 0/100 was calculated as the 'number of days with pain.' Furthermore, the VAS-POP area under the curve (VASAUC) (mm × day) was calculated by summating the VAS-POP areas. The VASAUC for postoperative 24 h (VASAUC_24 h) (mm × day) was calculated by summating the VAS-POP areas for postoperative 24 h, and VAS-peak values for postoperative 24 h (VAS-peak_24 h) was checked from the recording. The analgesic requirement period, number of days with pain, VASAUC, VASAUC_24 h, and VAS-peak_24 h were calculated for all patients and the two groups according to the surgery type (Le Fort type I osteotomy (Le Fort 1) and bilateral sagittal split ramus osteotomy (SSRO) group [n=10]; bilateral SSRO group [n=15]).

Statistical analysis
Data regarding patient background, OA, PCS-total score, PCS-subscale scores, analgesic requirement period, number of days with pain, and VASAUC are presented as medians [interquartile range].
T-tests or Welch two-sample t-tests: T-tests or Welch two-sample t-tests were performed to determine the difference between the two groups according to the surgery type (Le Fort 1 and SSRO group [n=10]; SSRO group [n=15]) for the analgesic requirement period, number of days with pain, VASAUC, VASAUC_24 h, and VAS-peak_24 h.
Spearman's rank correlation: Spearman's rank correlation was used to evaluate the correlations between 1) OA, PCS-total score, and PCS-subscale scores vs. the analgesic requirement period; 2) OA, PCS-total score, PCS-subscale scores vs. number of days with pain; 3) OA, PCS-total score, PCS-subscale scores vs. VASAUC; and 4) VASAUC_24 h vs. number of days with pain.
Multiple linear regression analysis: Multiple linear regression analysis was performed for the "number of days with pain" and VASAUC as the outcome variable, and OA and PCS as the explanatory variables. The residuals of the multiple linear regression model were analysed by graphically plotting the residuals against the predicted values and plotting normal quantile-quantile (Q-Q) plots. Statistical analyses were performed using EZR software (Jichi Medical University, Tochigi, Japan) [29]. Statistical significance was set at p<0.05.

Patients
In total, 32 out of 47 patients were recruited. Two patients, one who refused to receive analgesics and another patient with postoperative paralysis, were excluded from the analysis. Finally, 30 patients were analysed ( Figure 2). Table 1 shows the patient's demographic data and surgery type.

OA, PCS, and POP parameters
The descriptive statistics of OA, PCS, and POP parameters are listed in Table 2. A frequency plot of individual OA

Perioperative analgesic procedure
During the surgery, 2,000 mg of acetaminophen was administered for bilateral SSRO with Le Fort 1 (n=10) and bilateral SSRO with an anterior maxillary osteotomy (n=1), whereas 1,000 mg of acetaminophen was administered for the other surgery types (n=19).
For postoperative analgesic consumption, acetaminophen was used as the first-choice drug. The median administration period was 8.0 [6.3-9.0] days, and the median total dose was 19,000 [14,375] mg. In 19 patients, the single dose was reduced from 1,000 mg to 500-600 mg after the POP was reduced compared with that just after the operation. The median number of days for dose reduction was 6.0 Regarding other analgesics, among 30 patients, 28 received other analgesics in addition to acetaminophen. Among them, 27 required the administration of loxoprofen sodium hydrate. In particular, one patient was required to switch from acetaminophen to loxoprofen sodium hydrate because of liver dysfunction, and another discontinued acetaminophen and switched to loxoprofen sodium hydrate because of its bitter taste. A single dose of loxoprofen sodium hydrate was 60-120 mg. The administration period of loxoprofen sodium hydrate was 3.0 [2.0-5.0] days. The total loxoprofen sodium hydrate dose was 390 [180-480] mg.
On the day of surgery, intravenous flurbiprofen axetil was administered to 19 patients. The single dose was 50 mg; one patient received the drug three times, whereas all other patients received the drug once. Pentazocine was administered to one patient on the operative day and POD1. One patient required a single dose of pentazocine (15 mg) three times on the operative day and POD1.
Local anaesthesia (0.5 % levobupivacaine hydrochloride) was administered on the operative day for POP control in two patients. A mandibular nerve block to the right and left sides was performed in one patient, whereas a mandibular nerve block to the right and left sides along with a mental nerve block to the right side was performed    in another patient. The dose of 0.5 % levobupivacaine hydrochloride was 5 mL on each site.

POP following orthognathic surgery
The POP parameters for the surgery type (Le Fort 1 and SSRO group [n=10] and SSRO group [n=15]) are listed in Table 3. Student's t-test and Welch two-sample t-test showed that there were no significant differences between the two groups in the analgesic requirement period (p=0.394), number of days with pain (p=0.31), VASAUC (p=0.479), VASAUC_24 h (p=0.475), and VAS-peak_24 h (p=0.824).

Predicting postoperative pain
Normal Q-Q plots revealed that the VASAUC was normally distributed. The normal Q-Q plots revealed that the number of days with pain was not normally distributed; thus, a logarithmic translation was required. The log (number of days with pain) was normally distributed after logarithmic translation; however, the analgesic requirement period was not normally distributed before or after logarithmic translation.

Spearman's rank correlation
Spearman's rank correlation revealed that OA was inversely correlated with the number of days with pain (R=−0.539, p=0.002), indicating that lower preoperative OA was associated with a longer period of POP. A function in conjunction with a polynomial approximation (R 2 =0.2951) is presented in Figure 4. PCS-magnification was positively correlated with the number of days with pain (R=0.369, p=0.045), indicating that higher preoperative PCS-magnification was associated with a longer period of POP. There were no significant correlations between the following: 1) OA, PCS-total score, and PCS-subscale scores vs. analgesic requirement period; 2) PCS-total score, PCS-rumination, and PCS-helplessness vs. the number of days with pain; and 3) OA, PCS-total score, PCS-subscale scores vs. VASAUC.
Similarly, there was no significant correlation between VASAUC_24 h and the number of days with pain (R=0.132, p=0.486). Table 4 presents the results of the multiple linear regression model of the factors that explain the number of days with pain and VASAUC. Since the analgesic requirement period was not normally distributed, multiple linear regression analysis for the dependent variable (of the analgesic requirement period) was not performed. PCS-total and subscale scores were not significant predictors of log (number of days with   pain) and VASAUC. In contrast, OA was a significant independent predictor of log (number of days with pain) but not of VASAUC (Table 4). As a result of multiple linear regression analysis, the derived prediction formula was as follows: log (number of days with pain)=(−0.01 × OA) + 2.53 (adjusted R 2 =0.3251, p=0.0019, OA; p=0.008).

Discussions
This study showed that a low preoperative OA value predicted the number of days with acute POP following orthognathic surgery. In addition, PCS-magnification was related to the number of days with POP.

OA and POP
OA was initially reported by Grill and Coghill in 2002 and is defined as a disproportionately large pain reduction after a small change (reduction) in temperature [11]. OA is assumed to be a filtering mechanism for nociceptive information, that is, an inhibitory temporal sharpening mechanism [11,14]. A systematic literature review including healthy volunteers and patients reported that OA activates brain regions such as the periaqueductal grey, dorsolateral prefrontal cortex, insula, medulla, pons, and cerebellum [25]. Zhang et al. examined OA in patients with heterogeneous chronic pain disorders using functional magnetic resonance imaging (fMRI) and demonstrated that OA attenuation was associated with suppressed activation of the descending pain modulatory and reward systems [13]. Niesters et al. observed reduced or absent OA responses in patients with chronic neuropathic pain [30].
Diffuse noxious inhibitory controls and conditioned pain modulation (CPM) have been proposed to evaluate different aspects of descending inhibitory mechanisms [23,25,[31][32][33]. CPM was suggested to be predictive of a lower risk of chronic post-thoracotomy pain [34] and pain after total knee arthroplasty [35]. In addition, patients with chronic pain after abdominal surgery showed lesser preoperative CPM and greater postoperative hyperalgesia [36]. Interestingly, CPM, but not OA, has been suggested to be predictive of the analgesic response to non-steroidal anti-inflammatory drugs, which indicates that the two assessments might represent different predictive components [24]. Acute POP is also predicted by perioperative CPM in patients undergoing orthognathic surgery [37]. Taken together, weakened endogenous pain modulation seems to be associated with more POP problems [34][35][36][37], indicating that descending modulatory pathways play a predominant role. Therefore, it may also be important to evaluate whether OA can contain similar synergistic information to predict, in the first place, the development of acute POP.
These findings suggest that the modulatory mechanisms involved in OA are centrally mediated [12] and can influence the duration of POP, which is consistent with the findings of a previous study [37]. The scatter plot for OA vs. the number of days with pain ( Figure 4) shows outliers. It is evident that there is much variability in most QST assessments [9,38]. Accumulating evidence suggests that this variability is associated with POP after the surgery [10], which indicates that variability (and outliers) might hold important clinical value.

PCS and POP
Pain catastrophising has been defined as an exaggerated negative mental response to actual or anticipated painful experiences [5]. Psychological factors are consistently associated with chronic POP, including anxiety, depression, pain catastrophising, and general psychological distress [2]. Pain catastrophising has been identified as one of the strongest psychologic predictors of pain and pain outcomes [6,39,40].
PCS represents a cognitive aspect of pain [28], and its different components, individual interactions, and modelling of catastrophising have been recently reviewed [39]. Catastrophising has been associated with heightened pain experience in clinical and experimental studies [41], and POP intensity has been estimated based on the levels of preoperative anxiety and catastrophising [5,8]; hence, PCS is predictive of chronic POP [4]. A previous study of breast surgery patients reported that PCS predicted POP in only patients with pre-existing chronic pain [7].
Nevertheless, the multiple linear regression analysis in the present study revealed that the PCS-total and PCS-subscale scores were not predictive factors for the number of days with pain.
The participants in the current study may not have experienced strong anxiety or catastrophising. Matsuoka et al. validated the Japanese version of the PCS and reported that the average values for healthy volunteers were as follows: PCS-total, 21.39 (9.92); PCS-rumination, 11.55 (4.49); PCS-helplessness, 6.16 (4.29); and PCS-magnification, 3.67 (2.73) [28]. The PCS scores obtained in the current study were similar to those of healthy Japanese volunteers [28].
The following reasons could have led to the similarity in PCS scores between the validated values for healthy volunteers from the study by Matsuoka and Sakano [28] and our findings: 1) the participants in the current study were healthy young patients without chronic pain; and 2) as the included patients were scheduled for an orthognathic surgery, which is a cosmetic surgery and not a surgery for cancer or chronic pain such as knee osteoarthritis, they might not have experienced strong anxiety or catastrophising.
The fact that the obtained PCS was similar to that in healthy volunteers might be a possible reason why PCS was not a predictive factor of acute POP in the current study.
Although PCS-total and PCS-subscale scores were not predictive factors for the number of days with pain, the present findings demonstrated that patients with higher PCS-magnification scores experienced a longer duration of POP, indicating that there is a relationship between the preoperative PCS-magnification score and acute POP as reported by previous studies [1][2][3][4][5].

Clinical implication
A higher POP intensity may have led to a longer duration of POP [1][2][3][4]. Nevertheless, our result revealed no significant correlation between acute pain intensity (VASAUC_24 h) and the number of days with pain, possibly due to intensive pain control with rescue analgesics and nerve blockade.
One of the most consistent predictors of chronic POP is the presence of severe acute pain in the first week after the surgery [4]. Poorly controlled acute POP has been suggested as a predictor of the development of chronic POP [1][2][3][4]. Therefore, preoperative prediction and the detection of high-risk groups may pave the way for developing individualised management regimes that may increase the chances of preventing the development of chronic POP.
OA is a novel and easy pain assessment test for evaluating peripheral and central pain modulatory processes [12]. In addition, OA evaluation is easy to perform, and it is possible to perform tests at the bedside in a short time. Therefore, preoperative OA evaluation could be an attractive bedside test tool for identifying patients at risk of acute POP. Intensive care with tailor-made care for high-risk patients will help avoid the onset of chronic POP.

Limitations
The study population comprised only 30 patients. In addition, different types of surgery, such as Le Fort 1 and SSRO, or SSRO were performed. Therefore, the findings from the current study should be interpreted with caution, and larger studies should be conducted to validate these findings.
Although our result showed no significant correlation between the POP intensity and the duration of POP, it is possible that the POP intensity may have affected the results of this study.
A recent systematic review identified that temporal summation of pain is often an independent predictor for chronic POP when the temporal summation of pain is assessed in combination with other QST assessments [10]. However, the current study did not assess the temporal summation of pain. Future studies should assess whether OA is an independent predictor of acute or chronic POP.

Conclusions
This study demonstrated that the preoperative evaluation of OA predicts the number of days with acute POP following orthognathic surgery, and PCS-magnification is related to the number of days with POP.