Now Reading
The effects of steroids in preventing facial oedema, pain, and neurosensory disturbances after bilateral sagittal split osteotomy: a randomized controlled trial

The effects of steroids in preventing facial oedema, pain, and neurosensory disturbances after bilateral sagittal split osteotomy: a randomized controlled trial



International Journal of Oral & Maxillofacial Surgery, 2015-02-01, Volume 44, Issue 2, Pages 252-258, Copyright © 2014 International Association of Oral and Maxillofacial Surgeons


Abstract

A randomized, prospective, controlled trial was conducted to determine the efficacy of single and repeated betamethasone doses on facial oedema, pain, and neurosensory disturbances after bilateral sagittal split osteotomy. Thirty-seven patients (mean age 23.62 years, range 17–62 years) with either mandibular prognathism or retrognathism were enrolled consecutively into the study and divided into three groups: control ( n = 12), repeated dose 4 + 8 + 4 mg betamethasone ( n = 14), single dose 16 mg betamethasone ( n = 11). The intake of diclofenac and paracetamol was assessed individually. Measurements of facial oedema, pain, and sensitivity in the lower lip/chin were obtained 1 day, 7 days, 2 months, and 6 months postoperatively. Furthermore, we investigated the possible influences of gender, age, total operating time, amount of bleeding, postoperative hospitalization, and advancement versus setback of the mandible. A significant difference ( P = 0.017) was observed in percentage change between the two test groups and the control group regarding facial oedema (1 day postoperatively). Less bleeding was associated with improved pain recovery over time ( P = 0.043). Patients who required higher postoperative dosages of analgesics due to pain had significantly delayed recovery of the inferior alveolar nerve at 6 months postoperatively ( P < 0.001). Betamethasone did not reduce neurosensory disturbances over time.

Apart from the immediate postoperative discomfort of pain and oedema, the most common complication of bilateral sagittal split osteotomy (BSSO) is injury to the inferior alveolar nerve (IAN). Mandibular osteotomies are performed with cutters, saws, chisels, and separators in close proximity to the mandibular canal and the neurovascular bundle, with a high risk of temporary or permanent damage to the IAN, particularly if the nerve is exposed through the cancellous part of the bone.

Postoperative nerve function is influenced by multiple factors, such as the patient's gender and age, preoperative information, extension of the medial segment, position of the mandibular canal, morphology of the mandibular angle, the surgeon's technique and skills, compression of the nerve trunk, fixation methods, operating time, and complications such as ‘bad split’. Regardless of whether the nerve damage occurs at the mandibular foramen, along the mandibular canal, or at the mental foramen, the symptoms of nerve lesions usually consist of varying degrees of numbness in the lower lip and chin, corresponding to the distribution area of the mental nerve. This complication continues to constitute a major drawback, with patient discomfort occasionally lasting for several months postoperatively. The vast majority of operated patients experience altered sensations and sensory impairment during the immediate postoperative period, but occasionally patients also report long-term permanent subjective sensory disturbances. In the literature, the extent and course of nerve recovery have varied greatly because of the lack of a uniform testing methodology.

Glucocorticoid administration in orthognathic surgery is recommended preoperatively to reduce postoperative pain, swelling, trismus, nausea, and vomiting and to promote nerve healing. However, there is a need for further clinical studies to support these statements. The most common glucocorticoids are betamethasone, dexamethasone, and methylprednisolone. The aim of the present study was to test two betamethasone regimens against a control group and to evaluate possible effects on postoperative facial oedema, pain, and neurosensory disturbances.

Materials and methods

Subjects

Thirty-seven patients requiring treatment for mandibular prognathism or retrognathism with a BSSO were enrolled consecutively into this study between February 2006 and March 2011. Informed consent was obtained from all subjects. The Central Ethics Review Board in Gothenburg approved the investigation.

For inclusion, patients had to be healthy without any regular medication. Contraceptive agents were allowed. Patients requiring an additional genioplasty or maxillary osteotomy were excluded, as well as patients with contraindications to steroids. All patients received either penicillin G (3 g 3×) intravenously (IV), or clindamycin (600 mg 3×) IV in the case of allergy, immediately preoperatively and postoperatively. Penicillin V (1 g 3×, oral) or clindamycin (300 mg 2×, oral) was administered for the first postoperative week.

A randomized, double-blind protocol was used to assign patients prospectively and consecutively to one of three experimental groups. Support staff drew designations from sealed envelopes. The following groups and betamethasone regimens were used: control ( n = 12), repeated dose (4 mg betamethasone administered orally 1 day prior to surgery, 8 mg betamethasone IV administered immediately preoperatively, and 4 mg betamethasone administered orally 1 day postoperatively; n = 14), and single dose (16 mg betamethasone IV administered immediately preoperatively; n = 11).

Facial oedema, pain, and sensitivity were recorded preoperatively and at 1 day, 7 days, 2 months, and 6 months postoperatively. Facial oedema was measured objectively as the distance between the earlobes and below the chin. Pain was estimated subjectively using a visual analogue scale (VAS) ranging from 0 to 10, with 0 indicating no pain and 10 indicating maximum pain. Sensitivity was evaluated subjectively in the lower lip, right and left side, using a VAS ranging from 0 to 10, with 0 indicating no sensitivity and 10 indicating maximum sensitivity. All measurements were obtained at the bedside or at scheduled return visits to the clinic.

All participating clinicians performed calibrated measurements. Furthermore, we investigated the possible influences of gender, age at the time of surgery, total operating time, amount of bleeding, postoperative hospitalization, and advancement versus setback of the mandible.

The primary hypothesis was that postoperative facial oedema is associated with neurosensory disturbances. The secondary hypothesis was that a repeated or single betamethasone regimen protects the nerve from surgical trauma and postoperative oedema and promotes nerve recovery.

Surgery

All patients were carefully assessed clinically and with radiographs. Cephalometric analysis and diagnostic imaging were performed using Facad software (Ilexes AB, Linköping, Sweden). Functional and aesthetic demands were discussed with the patient. Patients requiring an additional genioplasty or maxillary osteotomies were excluded from the study. The BSSO was performed with the Hunsuck modification of the basic Obwegeser–Dal Pont method (no attempt was made to dissect the IAN), and the position of the nerve was recorded, i.e. whether it was visible or hidden (embedded) in the cancellous bone. At the time of fixation, a straight miniplate (four holes and four screws) was placed monocortically on each side of the mandible (MatrixORTHOGNATHIC Plating System, DePuy Synthes, Zuchwil, Switzerland, or 2.0-mm mini-system, KLS Martin, Tuttlingen, Germany). Passive adaptation, with the monocortical approach, was chosen to minimize the compression effects of the IAN during rigid fixation of the proximal and caudal segments. After surgery, frontal and lateral radiographs were obtained ( Fig. 1 ). Bimaxillary postoperative elastics were positioned individually on the orthodontic appliances for approximately 2 months after surgery to correct occlusion and reduce muscular strain. Intraoperative complications (such as a ‘bad split’) and postoperative infections were noted. The first three authors of this study performed all surgeries.

Frontal postoperative radiograph.
Fig. 1
Frontal postoperative radiograph.

All patients were offered diclofenac (50 mg 3×) and paracetamol (1000 mg 4×) postoperatively, depending on their individual requirements. If pain relief was insufficient, opioids were administered on an individual basis.

Statistical analysis

VAS values for sensitivity from the patient's left and right sides were combined to obtain a mean value, regardless of whether the nerve was visible or hidden. This combining of values was also performed for cases in whom the nerves on both sides were visible or hidden during surgery. Pairwise comparisons of the VAS values showed that there were no significant differences between the hidden and visible nerves ( Table 1 ).

Table 1
Sensitivity of the IAN according to the VAS. All sites had full sensitivity preoperatively (VAS 10).
Visible ( n = 8) Visible and hidden ( n = 13) Hidden ( n = 16)
n Mean ± SD n Mean ± SD n Mean ± SD
After 1 day, (right + left)/2 7 2.07 ± 1.54 12 5.08 ± 2.43 15 5.30 ± 2.85
After 7 days, (right + left)/2 7 4.07 ± 2.28 13 5.58 ± 2.76 16 4.97 ± 2.40
After 2 months, (right + left)/2 6 7.50 ± 1.00 12 7.83 ± 2.06 15 7.50 ± 2.55
After 6 months, (right + left)/2 5 8.90 ± 1.02 11 9.14 ± 1.58 14 8.96 ± 1.31
After 1 day (visible minus hidden) 12 −0.50 ± 2.43
After 6 months (visible minus hidden) 11 −0.45 ± 1.51
IAN, inferior alveolar nerve; VAS, visual analogue scale; SD, standard deviation.

For each patient and each variable of the results (oedema, pain, and sensitivity), a linear regression coefficient was calculated to describe the trend over time from 1 day to 6 months postoperatively ( Table 2 ). Fisher's permutation test was used to test the between-group differences for all variables. Correlations were evaluated with Pitman's test. Mantel's test was used to test the between-group differences when adjusted for baseline variables. All tests were non-parametric. Two-sided P -values were used, with the significance level set at 5% ( P < 0.05).

Table 2
Speed of recovery represented by the regression coefficient for 1 day, 7 days, 2 months, and 6 months postoperative. a
Control ( n = 12) 4 + 8 + 4 mg ( n = 13) 16 mg ( n = 11) Two-sided P -value
Placebo vs. corticosteroid 4 + 8 + 4 vs. 16 mg
Oedema (mm), per day −0.27 ± 0.14 −0.72 ± 1.28 −0.73 ± 2.08 >0.30 >0.30
Sensitivity (VAS 0–10), per 10 days 0.52 ± 1.16 0.85 ± 2.25 0.27 ± 0.21 >0.30 >0.30
Pain (VAS 0–10), per 100 days −0.77 ± 0.71 −1.08 ± 1.15 −0.58 ± 0.54 >0.30 0.20
VAS, visual analogue scale; SD, standard deviation.

a Data are expressed as the mean ± SD.

Results

Subjects

The patients ranged in age from 17 to 62 years (mean age 23.62 years); seven were male and 30 were female. No statistically significant differences were found among the three groups with regard to age or gender. Details of the study population according to the randomized betamethasone dosages are presented in Table 3 .

Table 3
Demographic details and characteristics of the study population by randomized corticosteroid doses. a
Control ( n = 12) 4 + 8 + 4 mg ( n = 14) 16 mg ( n = 11) Two-sided P -value
Placebo vs. corticosteroid 4 + 8 + 4 vs. 16 mg
Subjects
Age, years, mean ± SD 26.4 ± 15.3 20.5 ± 4.2 24.6 ± 9.2 >0.30 0.26
Age, years, median (range) 20 (18–62) 19 (17–34) 20 (17–43)
Male/female ( n / n ) 1/11 3/11 3/8 >0.30 >0.30
Surgery
Duration of surgery (min) 104 ± 13 108 ± 22 117 ± 39 >0.30 >0.30
Bleeding (ml) 94 ± 47 111 ± 66 168 ± 129 0.24 0.23
Movement (mm) 6.9 ± 2.1 ( n = 8) 4.5 ± 2.5 ( n = 12) 3.8 ± 2.6 ( n = 10) 0.011 >0.30
Advancement/setback ( n / n ) 7/5 5/9 8/3 >0.30 0.15
Hidden nerves ( n )
0 1 6 1 0.19 >0.30
1 4 3 6
2 7 5 4
Duration of hospitalization (days) 2.3 ± 0.9 2.0 ± 0.4 2.6 ± 0.7 >0.30 0.049
Medication
Paracetamol (mg) 4200 ± 2400 3300 ± 1700 5000 ± 2800 >0.30 0.097
Diclofenac (mg) 171 ± 96 114 ± 72 209 ± 122 >0.30 0.041
Preoperative
Oedema (mm) 224 ± 19 230 ± 17 232 ± 24 >0.30 >0.30
Sensitivity (VAS 0–10) 10.0 ± 0.0 10.0 ± 0.0 10.0 ± 0.0 >0.30 >0.30
Pain (VAS 0–10) 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 >0.30 >0.30
Day 1 postoperative
Oedema (mm) 271 ± 25 259 ± 22 ( n = 13) 268 ± 27 ( n = 9) >0.30 >0.30
Oedema (% change from preoperative) 21.0 ± 8.3 13.6 ± 6.4 ( n = 13) 13.3 ± 10.4 ( n = 9) 0.017 >0.30
Sensitivity (VAS 0–10) 5.0 ± 2.8 4.2 ± 3.2 ( n = 13) 4.4 ± 2.0 ( n = 9) >0.30 >0.30
Pain (VAS 0–10) 3.5 ± 2.0 3.5 ± 1.7 ( n = 13) 4.3 ± 2.2 ( n = 9) >0.30 >0.30
Difference between 6 months postoperative and preoperative
Oedema (mm) −1.5 ± 11.4 ( n = 10) −9.0 ± 13.7 ( n = 10) −1.4 ± 15.2 ( n = 10) >0.30 >0.30
Oedema (% change from preoperative) −0.6 ± 5.0 ( n = 10) −3.8 ± 6.0 ( n = 10) −0.1 ± 6.7 ( n = 10) >0.30 0.21
Sensitivity (VAS 0–10) −1.1 ± 1.3 ( n = 10) −0.6 ± 0.8 ( n = 10) −1.4 ± 1.7 ( n = 10) >0.30 0.20
Pain (VAS 0–10) 0.60 ± 1.35 ( n = 10) 0.00 ± 0.00 ( n = 10) 0.50 ± 0.85 ( n = 10) >0.30 0.17
SD, standard deviation; VAS, visual analogue scale.

a Data are expressed as the mean ± standard deviation (SD).

A significant difference was found between the test groups and the control group with regard to movement ( P = 0.011). However, no significant association with facial oedema, pain, or sensitivity was found ( Tables 3 and 4 ). At 1 day postoperative, there was significantly less swelling in the test groups compared to the control group ( P = 0.017), whilst no significant differences were noted concerning pain and sensitivity ( Tables 3 and 5 ). No differences were found between the test and the control groups for any of the parameters at 6 months postoperatively ( Figs 2–4 and Table 3 ).

Table 4
Correlations between VAS results and baseline variables postoperatively.
Sensitivity (VAS 0–10) Pain (VAS 0–10)
1 day 6 months Regression coefficient 1 day 6 months Regression coefficient
Age (years) >0.30 0.046 a 0.23 >0.30 >0.30 >0.30
Gender >0.30 >0.30 >0.30 >0.30 >0.30 >0.30
Bleeding (ml) >0.30 0.22 >0.30 >0.30 0.12 0.043 b
Number of nerves hidden 0.018 b >0.30 0.11 >0.30 >0.30 0.042 b
Movement (mm) >0.30 >0.30 >0.30 >0.30 >0.30 0.097
Advancement/setback >0.30 >0.30 >0.30 0.26 0.11 >0.30
Paracetamol (mg) >0.30 <0.001 a >0.30 >0.30 >0.30 >0.30
Diclofenac (mg) >0.30 <0.001 a >0.30 0.26 0.12 0.27
Operation (min) >0.30 >0.30 0.096 0.30 >0.30 0.26
VAS, visual analogue scale.

a Negative correlation.

b Positive correlation.

Table 5
Characteristics of oedema by randomized corticosteroid doses.
Oedema (mm) Control ( n = 12) 4 + 8 + 4 mg ( n = 14) 16 mg ( n = 11)
n Mean ± SD 95% CI n Mean ± SD 95% CI n Mean ± SD 95% CI
Preoperative 12 224 ± 19 212–236 14 230 ± 17 220–239 11 232 ± 24 216–248
1 day postoperative 12 271 ± 25 255–287 13 259 ± 22 246–273 9 268 ± 27 247–288
% change from preoperative to 1 day postoperative 12 21.0 ± 8.3 15.7–26.3 13 13.6 ± 10.4 9.7–17.5 9 13.3 ± 10.4 5.3–21.3
7 days postoperative 12 245 ± 24 230–260 13 244 ± 19 232–255 11 237 ± 27 219–255
2 months postoperative 11 226 ± 17 215–237 12 221 ± 13 212–224 10 234 ± 22 218–250
6 months postoperative 10 226 ± 18 213–239 10 217 ± 13 207–226 10 232 ± 17 219–244
SD, standard deviation; CI, confidence interval.

Pain VAS scores (scale 0–10) with 95% confidence intervals.
Fig. 2
Pain VAS scores (scale 0–10) with 95% confidence intervals.

Sensitivity VAS scores (scale 0–10) with 95% confidence intervals; mean value from the right and left sides of the mandible combined.
Fig. 3
Sensitivity VAS scores (scale 0–10) with 95% confidence intervals; mean value from the right and left sides of the mandible combined.

Facial oedema (measured between the earlobes under the chin) with 95% confidence intervals.
Fig. 4
Facial oedema (measured between the earlobes under the chin) with 95% confidence intervals.

Correlations showed that increased age hindered full recovery of nerve function at 6 months postoperatively ( P = 0.046) ( Table 4 ). More intraoperative bleeding resulted in delayed recovery from pain during the postoperative period ( P = 0.043) ( Table 4 ). A visualized nerve at surgery resulted in impaired sensitivity on the first day postoperatively ( P = 0.018), with minimal differences between the groups at 7 days, and with no differences at 2 months and 6 months postoperatively ( Tables 1 and 4 ). There was a significant difference between the two glucocorticoid regimes regarding the postoperative intake of diclofenac ( P = 0.041). Hence, to interpret the results, a sensitivity analysis adjusted for diclofenac intake was performed to compare the two groups. There were no changes in the results, nor were any significant between-group differences observed when comparing the regression coefficients representing changes over time ( Table 2 ).

Higher postoperative requirements for diclofenac and paracetamol were associated with a significantly delayed recovery of sensitivity in the lower lip ( P < 0.001) ( Table 4 ).

Surgery

No serious complications, such as a ‘bad split’, occurred during the operations. After surgery, no wound infections, bone instability, or non-union complications were noted.

Discussion

Facial oedema is difficult to quantify in a standardized manner. Previous reports using various investigations and methods have highlighted the problems involved in obtaining accurate measurements of three-dimensional (3D) facial changes over time. Hence, a simple method for measuring facial oedema caused by operative procedures was utilized. A VAS was used in order to estimate pre- and postoperative sensitivity and pain.

Glucocorticoids reduce oedema by decreasing the permeability of the capillary endothelium and, therefore, reducing the amount of fluid, protein, macrophages, and other inflammatory cells entering areas of tissue injury. Betamethasone is a potent glucocorticoid, with anti-inflammatory and immunosuppressive properties. Unlike other drugs with similar effects, betamethasone does not cause water retention. Betamethasone appears to be the most suitable drug because it has the highest anti-inflammatory activity, minimal mineralocorticoid activity, and a biological half-life of 36–54 h, which is the longest available. The use of glucocorticoids in orthognathic surgery has not yet reached full acceptance because of concerns over its side effects, particularly adrenal suppression, the risk of osteonecrosis, impaired wound healing, and concerns about its efficacy. The main indication for corticosteroid use is postoperative oedema, which in severe cases can compromise the airway, patient recovery, and surgical outcomes. Randomized trials have shown that low-dose and short-duration corticosteroid regimes are safe and effective for reducing postoperative pain.

Furthermore, strong grade A evidence supports corticosteroid use in multimodal analgesia protocols for facilitating postoperative recovery of the patient by minimizing opioid doses and, therefore, side effects. However, the optimal mode, dose, and timing of administration remain unclear. In addition, high-dose steroids have been used selectively in cases of severe spinal cord injury to prevent neurological loss. Three large and randomized clinical trials, the National Acute Spine Cord Injury Studies (NASCIS I, II, and III), have analyzed the effects of high-dose steroids on total motor and sensory scores. The efficacy of high-dose steroid regimens remains uncertain and requires further study. Treating traumatic optic neuropathy with steroids remains controversial, and the supporting evidence is weak. Early treatment, based on high-dose steroids, has been found to slightly accelerate spontaneous improvement in Bell's palsy patients.

Postoperative pain management relieves suffering and results in earlier mobilization, a shortened hospital stay, and reduced hospital costs; it also increases patient satisfaction. Acute postoperative pain can be considered a complex relationship among three components: afferent nociceptive stimulation; interpretation and modulation of these three signals by higher centres (involving memory and previous experiences); and an effective component (involving fear, anxiety, and depression). Therefore, the degree of postoperative pain experienced by patients can vary enormously, and pain control should not be standardized. Instead, pain control should be tailored to the needs of the individual patient. Current evidence suggests that a combination of paracetamol and non-steroidal anti-inflammatory drugs (NSAIDs) could offer superior analgesia compared to either drug alone.

The trigeminal somatosensory-evoked potential (TSEP) is a technique developed for the measurement of hypoesthesia in peripheral nerves, and was first described by Larsson and Prevec. The method makes it possible to classify IAN injuries caused by surgery into axonal or demyelinated injuries, with intraoperative continuous neurophysiological monitoring. We know that demyelinating lesions heal completely, normally within 2–4 months, along with remyelination, whilst axonal injuries often recover incompletely, slowly over months and years, also with a higher risk of developing postoperative sequelae such as pain and paresthaesia. When monitoring BSSO with the TSEP technique, Nakagawa et al. concluded that trigeminal hypoesthesia appeared to be induced by direct injury to the IAN during the bone split or by compression injury at fixation. Coincidentally, surgical trauma may occasionally result in more severe axonal injuries and therefore might explain our unexpected results that a higher postoperative need for diclofenac and paracetamol was associated with a prolonged recovery period for sensory function.

Current evidence supports steroid use preoperatively, but the timing of administration and postoperative use are contentious. The choice of defined steroid groups and regimes in the present study was selected in part from the study by Weber and Griffin, and the statement that steroids significantly reduce postoperative oedema was also confirmed in the present study. Steroids decrease postoperative oedema and pain significantly, with no higher risk of infection and with a minimum risk of other side effects. An evaluation of steroids in the postoperative period (prednisolone 30 mg for 7 days, 15 mg for 4 days, and 5 mg for 3 days) showed improved recovery of sensory impairment. Only a few orthognathic trials have evaluated the effects of steroids on facial oedema using pre- and postoperative steroid dosages. Furthermore, no studies have as yet attempted to prospectively evaluate single and repeated dosages with regard to the reduction of postoperative pain, facial oedema, and trigeminal neurosensory disturbances over time. In the future, perhaps repeated and low-dose steroids administered pre- and postoperatively over a longer period of time will be utilized.

A limitation of this study was that a minority of the patients were lost to postoperative follow-up and therefore not included in the entire analysis ( n = 8). These dropouts were compared to the patients who contributed to the whole analysis regarding variables at baseline and the treatment groups. Nevertheless, no significant differences could be found between the treatment groups. The study design did not answer the question about the optimal steroid dosage or regimen.

Facial oedema was measured as the distance between the earlobes and below the chin. These measurements might be affected by the mandibular movement resulting from advancement or setback. However, the significant difference identified between the control and the two test groups regarding mandibular movement was not confirmed with any differences between facial oedema, pain, or sensitivity.

The use of betamethasone resulted in a significant reduction of facial oedema in the short term and therefore we believe that steroids should be used as part of the multimodal analgesia protocol to facilitate postoperative recovery and mobilization of the patient. However, neurosensory disturbances were not necessarily associated with postoperative facial oedema and therefore steroids could not prevent neurosensory disturbances or promote healing of the nerve over time. Interestingly, we observed that less intraoperative bleeding was associated with significantly improved recovery from pain postoperatively.

Funding

This study was supported by grants from the NU Hospital Organization (93381) , Västra Götalandsregionen, Sweden. The NU Hospital Organization provides grants in order to promote high quality research and was not involved in the study design.

You're Reading a Preview

Become a DentistryKey membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here