A randomized clinical trial of different associations between dexamethasone and non-steroids anti- in ammatories in the lower third molar extractions

Gustavo Antonio Correa Momesso Sao Paulo State University Gustavo Augusto Grossi Oliveira Sao Paulo State University William Phillip Pereira Silva Sao Paulo State University Renan Akira Sao Paulo State University Fernando Chiba Sao Paulo State University Tárik Ocon Braga Polo Sao Paulo State University Tiburtino José Lima Neto Sao Paulo State University Bárbara Ribeiro Rios Sao Paulo State University Ana Paula Farnezi Bassi Sao Paulo State University Doris Hissako Sumida Sao Paulo State University Michael Han University of Illinois at Chicago Michael Miloro University of Illinois at Chicago Leonardo Perez Faverani (  leonardo.faverani@unesp.br ) Sao Paulo State University


Introduction
Postoperative pain and edema remain signi cant concerns for patients who undergo third molar removal.
For this reason, the literature contains several studies about controlling these symptoms. [1] It is the responsibility of the surgeon to treat the postoperative symptoms of challenging third molar surgeries, which can be quite intense, limiting patients' daily routines, and causing considerable discomfort. [2,3] Thus, preoperative drug therapies have been suggested to prevent or relieve postoperative symptoms. [2,4] Preemptive analgesia consists of managing drugs prior to nociceptive stimulation and is de ned as an anti-nociceptive treatment that prevents the in ammatory response from stimulation of the afferent pathway to the central nervous system. This general mechanism is responsible for the ampli cation of the postoperative pain response. [5] Steroidal anti-in ammatory drugs (e.g. corticosteroids) are used extensively during the preoperative period prior to third molar surgery. Several authors have demonstrated the e cacy of these drugs, mainly in the control of postoperative edema and trismus (MMO). [6][7][8] This effect is achieved because corticosteroids inhibit the conversion of phospholipids into arachidonic acid by phospholipase A2. This mechanism is responsible for preventing the production of leukotrienes, prostacyclins, prostaglandins, and thromboxane A2, which mediate pain and in ammation. The use of corticosteroids is believed to inhibit the initial step in this process. [9] Nonsteroidal anti-in ammatory drugs (NSAIDs) are mostly used during the postoperative period of third molar surgeries due to their considerable anti-nociceptive effects. [10][11][12] However, recent studies have demonstrated that the preemptive combination of NSAIDs and corticosteroids seems to improve the relief of postoperative symptoms after third molar surgeries, mainly in the acute postoperative period. [4,13,14] A possible explanation for the e cacy of this combination is that when administered before surgical trauma and the onset of pain, the drugs will be present in the bloodstream at the appropriate levels, and at the appropriate time when pain symptoms begin, and the inhibition of cyclooxygenase decreases the production of prostaglandins and other in ammatory cytokines at the surgical site. [15] Since there is a lack of consensus and few studies regarding the preemptive combination of steroids and NSAIDs aimed at improving the postoperative pain symptoms after third molar surgery, the purpose of this study is to compare the preemptive use of dexamethasone alone, and the combination of dexamethasone with different NSAIDs, in third molar surgeries.

Ethical Considerations
This prospective, randomized, triple-blind clinical trial was approved by the Ethical Committee for Human Experimentation from the São Paulo State University -School of Dentistry, Araçatuba -São Paulo, Brazil (number # 88903518.4.0000.5420). The trial was registered in a Brazilian Clinical Trials Registry (REBECdate of registration: 14/02/2019 -number of registration: #RBR-8rfwnq) and was written in accordance with the CONSORT guidelines (www.consort-statement.org) and the principles for medical research involving human subjects in the 1963 Declaration of Helsinki [16]. All subjects received the research information through the informed consent process. The patients were initially informed that the research aimed to compare many different drugs usually used in relieving the in ammatory symptoms and pain, and they were not aware about what the drugs were prescribed for. Besides that, all the risks about the surgical procedure were informed such as local and systemic infections, alveolar inferior nerve and lingual paresthesia, facial nerve paralysis due to anesthetic technique, mandibular fracture, bleeding, and other minor risks. After this, the patients were instructed that they could quit the research at any moment. These terms explained the study objectives and justi cations, the bene ts and risks to which the subjects were exposed, and the other items described in the National Health Council's Guidelines (Resolution CNS 196/96).

Patients and Subjects
This study involved 60 subjects presented as healthy subjects t with the inclusion criteria (ASA I-II), never-smokers, between the ages of 18 and 35 years, with an indication of removal of the mandibular third molars with at least 2/3 of the root formed (Class I or II; Position A/B from Pell and Gregory). All surgical procedures required bony removal and tooth sectioning in order to extract the impacted third molars.
Patients evaluated might present with the inclusion criteria established by the authors, which were: 1. Healthy patients with no local and systemic disorders.
3. Need of removal lower third molars with at least 2/3 of root formation, evaluated by radiographic exam.
Besides that, patients which did not present the necessary characteristics to participate in the study were excluded. The criteria were: of loxoprofen (Daiichi Sankyo Brasil, Barueri, Brazil); DEX + KET, which represented subjects who received 8 mg of dexamethasone plus 10 mg of ketorolac (União Química, São Paulo, Brazil); and DEX + IBU, which represented subjects who received 8 mg of dexamethasone plus 600 mg of ibuprofen (Prati Donaduzzi, Toledo, Brazil). The drugs were administered 1 hour before surgery and administered orally. The surgical interventions were timed from the beginning of the incision until the end of the last suture.

Randomization
Subject randomization was performed by a researcher (L.P.F.), who selected groups using the envelope system. The envelope contained papers on which the ve groups' names were written, and a random selection was performed. If the subject presented with two lower third molars, which was compatible with the inclusion criteria, the subject underwent a double draw and could be allocated to two groups. The drugs were given to the subjects by the same researcher. All the drugs were manipulated, the capsules were equally the same and were given in a surgical envelope to avoid the recognize of the drug by the patient and create a bias. The surgeon (G.A.C.M.) and the researcher who conducted the analysis (T.O.B.P.) did not know the groups to groups to which the subjects were assigned, characterizing a tripleblind clinical trial.
When a subject had bilateral third molars, the choice to initially operate on the right or left side was also determined by the researcher (L.P.F.) using the envelope system. In this situation, the surgeries were conducted unilaterally with a minimum interval of 21 days between the extraction procedures.

Surgical Procedures
The surgical procedures were performed by the same surgeon. All subjects received preoperative antibiotic therapy with 2g of amoxicillin 1 hour before the surgical procedure. Intrabuccal antisepsis was performed with vigorous mouth washing for 1 minute with an aqueous solution of 0.12% chlorhexidine digluconate, followed by extraoral antisepsis with 0.5% chlorhexidine alcohol solution. The anesthetic technique employed comprised a regional blockade of the inferior alveolar, buccal, and lingual nerves.
This was performed using a re ux syringe with a long 27-gauge gingival needle for injection of 2% mepivacaine hydrochloride with epinephrine 1:100,000 (Mepiadre®, DFL) at a maximum volume of 4.5 ml, equivalent to 2.5 tubes.
A triangular linear ap was made with a number 15 scalpel blade in the distal region of the lower second molar in association with a buccal releasing incision in the mesial aspect of the second molar. Following this, mucoperiosteal detachment was performed with periosteal relief and a retractor was used to expose the operative eld. Osteotomy was performed using a model 702 carbide-tipped drill bit mounted on a high-speed pen, in addition to abundant irrigation with sterile saline NaCl 0.9%. Dental removal was completed with curved and straight extractors of the Seldin type (n° 2, 1R, or 1L), followed by careful inspection for the removal of pericoronal follicles with curettes and Kelly curve tweezers.
The bone margins were trimmed to remove bone spicules using a bone le, with abundant irrigation using sterile saline NaCl 0.9%. Thereafter, sutures were performed with nylon 5.0 (Ethicon®, Johnson & Johnson, Brazil). Surgical interventions were performed in the morning and afternoon (between 08 am and 6 pm) in a temperature environment control.
After the surgical procedure, all subjects received 12 tablets of 750 mg paracetamol (not identi ed to the subject, in white tablet form) and were advised to take 1 tablet in case of pain, with a minimum interval of 6 hours between administrations of the analgesic. Subjects were instructed to recording a log of all relevant variables such as VAS score at each timepoint, and when analgesics were taken.
Immediately after the surgical procedures, the subjects were prescribed a cold, liquid, pasty, hyperproteinbased diet for the rst 48 hours postoperatively, as well as other aspects of general care, such as avoiding physical exertion and sun exposure. Subjects were advised not to use cold compresses during the postoperative period. Any subject who developed postoperative complications, such as bleeding and alveolar osteitis was treated and consequently excluded from the study. Follow-up visits were scheduled at 48 hours and 7 days postoperatively to evaluate for edema and MMO. Sutures were removed at the 7day postoperative mark. Primary outcomes For this study, the primary outcome variables were the assessment of pain (visual analog scale VAS, and number of rescue analgesics, NRA). Salivary PGE2 concentration, edema and MMO were secondary outcomes variables.
Pain was measured using a visual analog scale (VAS). During the postoperative period, all subjects received a VAS protocol containing a linear chart with a scale from 0 to 10. The subjects were instructed to point out the level of postoperative pain on the chart at 6, 12, 24, 48, and 72 hours and 7 days after the surgical procedure. A VAS score of 0 represented the absence of postoperative pain, whereas a score of 10 represented the worst pain the subject had experienced. [10,17] The weighted average corresponding to the values of every subject in each group was obtained.
Another outcome measure for pain involved the number of analgesics consumed by the subject during postoperative period. Subjects were instructed to record the time of rst analgesic consumption. The patients were no induced to consumed or not the drugs, only were informed that were usually drugs destined to relieve pain, edema and trismus and might be used only if were extremely necessary for a rescue way. The weighted average corresponding to the values of every subject in each group was obtained. Secondary outcomes The salivary collection was performed before subjects took the preoperative drug, prior to the surgical procedure, and 48 hours after the surgical procedure. Saliva was collected directly in a salivate tube (Salivette®; Genese Produtos Diagnósticos Ltda, São Paulo, SP, Brazil). The subjects were informed not to eat or brush their teeth in the 2 hours preceding the collection procedure. The tubes were properly capped and stored at -20°C to keep the samples stable. The salivates were centrifuged at 1000 g for one minute, resulting in a clean and uid saliva sample that was used to determine the PGE2 salivary concentration. Salivary PGE2 was measured by competitive enzyme-linked immunosorbent assay (ELISA) method using a commercial kit (Diametra DKO020, Milano, Italy).
Edema measurement was performed using a three-point analysis with a tape measure. Measurements were made of the mandibular angle to the lateral canthus, tragus to labial commissure, and tragus to pogonion at the preoperative, 48 hours postoperative, and 7 days postoperative periods (Fig. 1). [18] The values obtained were recorded in the subjects' individual les. The researcher (T.O.B.P.) who performed these measurements was blinded and also determined all the results of the analyses of this study. To test intra-observer variability and increase measurement precision, all measurements were repeated by the same operator 3 times to ensure reliability of the method and consistency of the operator [19] . To assess postoperative edema, a weighted average of the three-point measurements was calculated: 48-hour postoperative values were subtracted from preoperative values to represent the 48-hour postoperative edema. The representative 7-day postoperative edema was obtained by subtracting the postoperative 7day values from preoperative values.
Maximum mouth opening was evaluated using measurements in millimeters made with a digital caliper (Mitutoyo®, Sakado, Japan) between the incisal edge of the upper central incisors and the lower central incisors on the right side. The measurements were performed preoperatively, as well as 48 hours and 7 days after the surgical procedure. [10,17] For the calibration of the researcher, four patients were selected to MMO and edema measurements, before the beginning of the study. The analysis was performed by the same researcher twice in a 21days of interval. The Kappa test showed the concordance intra-examinator (K > 0.95)

Follow-up
The follow-up time of the patients after the analyses (7 days) varied from 2 to 3 months after surgery.

Statistical Analysis
All quantitative data underwent a statistical analysis and homoscedasticity statistical testing (Shapiro-Wilk test) for homogeneity or heterogeneity distribution. For VAS, edema, and MMO, a two-way ANOVA and Tukey posttests were applied, and results were considered signi cant when P < 0.05. For PEG2 concentration, a one-way ANOVA and Tukey posttests were performed, results were considered signi cant when P < 0.05. Only NRA failed in the homoscedasticity statistical test, and a Kruskal-Wallis test was applied. An intention to treat analysis was performed for all randomized patients who were included in this study.

Demographic data
A total of 125 subjects were screened for eligibility. Sixty-one subjects were excluded. Fifty subjects did not meet the inclusion criteria and 11 subjects declined to participate in the study. Thus, 64 subjects were randomized and divided into ve groups. All 64 subjects received the planned intervention, however, at the follow-up visit, four subjects were excluded due to postoperative dry socket alveolitis and the use of drugs/medicaments not related to the study as described in the exclusion criteria (Fig. 2). Demographic data regarding mean age, sex, and number of teeth removed are shown in Table 1.  (Fig. 3).

Number of rescue analgesics (NRA)
NRA data showed that subjects in the DEX + ETO group consumed the fewest analgesics during postoperative period. The average consumption was 1.2 analgesics during the entire postoperative recovery period (1.2 ± 0.36). After this group, subjects in the DEX (2 ± 0.43) and DEX + KET (2 ± 0.41) groups an average of 2 analgesics during postoperative recovery period. Subjects in the DEX + LOX group consumed an average of 1.9 analgesics (1.9 ± 0.44). Subjects in the DEX + IBU group consumed an average of 3 analgesics (3 ± 0.35) during the postoperative recovery period (Fig. 4).

Discussion
This study aimed to evaluate whether bene ts exist with a combination of NSAIDs plus DEX in decreasing postoperative pain, edema, and MMO following third molar surgery. Prior studies are unclear whether the combination of these drugs could provide an advantage in postoperative period, which justi es the clinical use of various drugs preoperatively. However, the combination of some NSAIDS plus DEX seems to improve symptoms in the rst postoperative hours of acute pain.
The preemptive use of steroids for third molar surgery, mainly DEX, is well established, and provides positive results regarding a comfortable postoperative period for the patients. [6,20,21] A 2008 clinical study by Laureano-Filho and colleagues [22] compared a preoperative dose of 4 mg and 8 mg of DEX in third molar surgeries and observed that there was a signi cant difference between the doses regarding postoperative MMO and edema, but not postoperative pain, as was observed in another study. [23] This lack of effect on pain may be explained by the ineffectiveness of DEX in reducing prostaglandins at the site of an injury. [24] Another possible explanation for this phenomenon is that cortisol naturally inhibits pituitary beta-endorphins, which are potent endogenous analgesic present in the peripheral circulation. [25] Some clinical trials have suggested that the combination of steroids and NSAIDS improves preemptive effects on MMO, edema, and pain compared to steroids alone. [4,13] This study found that the combination of DEX plus several NSAIDS (DEX + ETO, DEX + KET, DEX + IBU, and DEX + LOX) showed no signi cant difference in the edema and MMO parameters. The exception was the combination of DEX + KET, which showed a signi cantly higher value for 48-hour postoperative MMO compared with DEX alone. These results could be explained because, as noted above, the use of preoperative DEX alone is well established and provides positive results regarding the improvement of postoperative edema and MMO. [6,20,21] Although data were obtained about postoperative pain, the VAS analysis showed that patients who took preoperative DEX + ETO and DEX + KET exhibited signi cantly lower pain perception during the rst 6 postoperative hours compared with all the groups. At 12 hours postoperative, the DEX group showed similar values to the DEX + ETO and DEX + KET groups, whereas the DEX + LOX and DEX + IBU groups maintained higher pain values during all periods. After the rst day (24 hours), subjects who took DEX + KET showed an increase in pain perception compared with the DEX and DEX + ETO groups, which had lower VAS values during the postoperative period. There was no signi cant difference between these two groups.
NRA did not signi cantly differ among the groups; however, it seemed to be lower for the DEX + ETO (average of 1.2 postoperative analgesics) group compared to the other groups. In addition, the DEX + IBU group presented the highest consumption of all groups. However, it is important to highlight that even though NRA was higher for DEX + IBU group, subjects who received this drug preoperatively took an average of three analgesics during the seven postoperative days. This value is well below what is usually prescribed in most practices, which also often include three days of acetaminophen with opioids at an interval of 6 hours. [26,27] DEX alone was the only preoperative therapy that decreased the salivary expression of PGE2 48 hours after the surgical procedure. The other groups showed an increase in salivary PGE2. However, the DEX + ETO and DEX + KET groups showed a discreet increase, whereas the DEX + IBU and DEX + LOX groups showed a signi cant increase in PGE2. This data disagrees with the VAS, which showed that the DEX, DEX + ETO, and DEX + KET groups presented similar pain perception values during the entire postoperative period. However, PGE2 is not the only pain mediator released during an in ammation response. Other algogenic substances are generated from arachidonic acids, such as bradykinin, leukotrienes, tumoral necrose factors (TNF-α) and interleukins. [28] Although the PGE2 expression values were higher, VAS evaluated the subjects' pain perception. Thus, even though the drugs had no effect on PGE2 expression, they might have decreased the subjects' pain perception. Laureano-Filho and colleagues [22] showed that 8 mg of DEX 1 hour before surgery decreased the expression of PGE2 and thromboxane A2 at the surgical site, but this had a minimal effect on reported pain the day of the surgery.
Similar results were found in another trial with methylprednisolone 125 mg [1] .
There is no consensus about the preemptive combination of DEX and an NSAID. Although trials have shown great results for this therapy, this study observed that not all combinations produced good results. DEX + IBU and DEX + LOX presented poor results for all parameters, and DEX alone seems su cient to relieve postoperative pain. Currently, the medical eld seeks to decrease the number of drugs prescribed to patients. Furthermore, NSAIDs have several adverse effects, such as cardiovascular and gastrointestinal events. [29] Preemptive DEX alone performed well for postoperative pain, edema, MMO and, decreased salivary PGE2, but KET and ETO, when combined with DEX in a preemptive way, showed similar results to DEX alone in terms of VAS scores 7 days after the surgical procedure and signi cantly decreased patients' pain perception at 6 hours after surgery, compared to DEX alone, which resulted in the worst acute pain after the procedure. This demonstrates important bene ts for patients' postoperative comfort.
Indeed, ETO and KET have proven effective against pain when prescribed after third molar surgeries. [10,30,31] When combined with DEX, a single preoperative dose of KET and/or ETO seems to signi cantly reduce acute pain during the rst few hours, and there is no need to prescribe any postoperative NSAID except a rescue analgesic. This study presented some limitations. It did not account for other algogenic cytokine expressions did not evaluate other NSAIDs, so more trials should be designed to establish a clinical consensus about this eld. However, the use of speci c NSAIDs combined with 8 mg of DEX after third molar surgeries might provide patients with postoperative bene ts and should be considered in the dentistry routine.

Conclusion
It can conclude that the preemptive use of steroids with NSAIDS does not always provide more postoperative bene ts than steroids alone, but the combinations involving KET and ETO signi cantly  Flowchart of subjects screened according to CONSORT statement.

Figure 3
Data regarding VAS analysis. It was possible observed that at rst 6 hours postoperative, pain perception from DEX+ETO and DEX+KET groups were signi cantly lower than DEX and DEX+IBU (P < 0.05 -Tukey test). The other postoperative periods showed no signi cant difference between groups.

Figure 4
Chart regarding number of rescue analgesic consumption for 7 days postoperative.  Chart regarding postoperative edema (mm). Data showed no signi cant difference between groups in both periods.