Evaluation of the Effect of The Mouth Gag Application on Optic Nerve Sheath Diameter of Pediatric Patients Undergoing Tonsillectomy or Adenotonsillectomy: Observational study

Background A mouth gag is usually used during tonsillectomy and adenotonsillectomy surgeries, cleft palate repair, obstructive sleep apnea surgery and intraoral tumor excision. The placement of the gag causes hemodynamic changes similar to laryngoscopy. The aim of this study is to evaluate the effect of the mouth gag placement on optic nerve sheath diameter (ONSD) of pediatric patients. The secondary aim is to assess the relationship between neck extension for positioning of head and increase in ONSD measurement. Methods The trial is registered to Australian New Zealand Clinical Trials Registry with Trial ID: ACTRN12618000551291. This prospective, observational study was performed in a tertiary university hospital operating theatre between 01.05.2018-01.07.2018. Thirty-five children aged < 18 years, with ASA I status, and scheduled for tonsillectomy and adenotonsillectomy surgeries were prospectively included in the study. Measurements of ONSD were performed (T0) after induction of anesthesia, (T1) after endotracheal intubation, (T2) after placement of the mouth gag and (T3) 20 minutes after the placement of the mouth gag. The degree of neck extension was assessed by the angle between the Frankfort plane and horizontal plane of the operation The of a mouth gag significant gas analysis. Anaesthesia was induced with intravenous propofol 1-2 mg/kg, fentanyl rocuronium bromide 0.6 mg/kg. When BIS score decreased under 60, the patients were intubated anaesthesiologist at the first attempt. Anaesthesia was maintained with sevoflurane in 40% O 2 and 60% air mixture, and the inspired concentration of sevoflurane was targeted to maintain a BIS score between 40-60. Peak inspiratory pressure was strictly maintained between 11 and 13 cmH 2 O not to effect ICP.


Background
Introduction of oral antibiotics in the 1960s dramatically decreased the rate of tonsillectomy (T) and adenotonsillectomy (AT) surgeries in time; however, T and AT remain as one of the most common surgeries performed in children worldwide. Traditionally, the head of the patient is positioned in extension and a mouth gag is placed for these surgeries (1). The Crowe-Davis mouth gag was initially designed for mouth opening and intraoperative anaesthetic agent delivery, then Boyle modified the original device to use the mouth gag with endotracheal tubes. The mouth gag has three parts; the blade has a central groove for the positioning of the endotracheal tube, a gag helps mouth opening and lastly the suspension system of the gag maintains the position (2). Although it provides advantage for access to the intraoral cavity, placement of a mouth gag results in hemodynamic changes similar to laryngoscopy which causes significant increases in the intracranial pressure (ICP) and intraorbital pressures (IOP) (3). Moreover, excessive mouth opening causes tonic contractions at muscles of mastication and postoperative pain in the temporomandibular joint (4).
Several previous studies evaluated the ultrasonographic measurement of optic nerve sheath diameter (ONSD) as a non-invasive, simple and rapid way to detect the pressure changes of intracranial compartment (5,6). The sheath around the optic nerve is an anatomical extension of the dura mater, and within the sheath, intracranial subarachnoid space extends through the optic nerve. Therefore, a rise in ICP is directly transmitted to the distensible subarachnoid space around the optic nerve. The transbulbar sonography technique for the estimation of ICP by measuring ONSD of children was first described by Helmke et al (7). So far, several studies evaluated the reliability of ONSD measurements by concurrent magnetic resonance imaging and invasive methods (8)(9)(10). Steinborn et al (11) observed 99 healthy children and adolescents in order to determine the normal values of ONSD. They reported that the mean value for ultrasonographic ONSD measurements was 5.75 ± 0.52 mm. One year later, The authors observed 56 children with normal ICP and 25 children with elevated ICP to determine a cutoff value for normal ONSD (8). In this study, the diagnosis of elevated ICP (ICP ³ 15 mmHg) was based on different invasive measurement methods such as intracranial devices or lumbar puncture, concurrent imaging studies, and ophthalmologic findings. The researchers reported that mean ONSD in patients with normal ICP was 5.77 ± 0.48 mm while it was 6.85 ± 0.81 mm in children with elevated ICP. They calculated the optimal cutoff value of ONSD as 6.0 mm for estimation of elevated ICP with a sensitivity of 82% and specificity of 74%.
Although the measurement of ONSD has been used in different clinical scenarios in the current literature, there is no study evaluating the effect of the mouth gag placement on ONSD measurements. Thence, the primary aim of this study is to evaluate the effect of the mouth gag placement on ONSD of pediatric patients. The secondary aim is to assess the relationship between the degree of neck extension related to head positioning and increase in ONSD measurement.

Methods
This observational study was approved by Muğla Sıtkı Koçman University Clinical Research Ethic Committee (approval number: XII, 26.04.2018) and registered at anzctr.org.au (Trial ID: ACTRN12618000551291) and conducted in accordance with the current Declaration of Helsinki. Signed informed consents were obtained both from parents of all children and children themselves who were over 6 years of age. Patients between 3-18 years old with American Society of Anesthesiologists (ASA) physical status I-II and scheduled for a T or AT surgery were prospectively included in the study. The exclusion criteria were patients with known acute or chronic ophthalmic diseases, history of previous ophthalmic surgery, increased ICP, receiving ß blocker, calcium canal blocker, statin or nitrate treatment, more than one attempt for endotracheal intubation and duration of mouth gag application < 20 minutes.
All children received preoperative medication with midazolam 0.5 mg kg -1 orally (maximum dose of 15 mg) approximately 15-20 minutes prior to the induction of anaesthesia. A standard monitoring was employed to all children with electrocardiography, non-invasive arterial blood pressure, pulse oximetry, bi-spectral index (BIS) (Datex-Ohmeda S/5 monitor M-BIS module, Helsinki, Finland), nasopharyngeal temperature, end-tidal CO 2 (EtCO 2 ) measurement and gas analysis. Anaesthesia was induced with intravenous propofol 1-2 mg/kg, fentanyl 1 mcg/kg and rocuronium bromide 0.6 mg/kg. When BIS score decreased under 60, the patients were intubated by an experienced anaesthesiologist at the first attempt. Anaesthesia was maintained with sevoflurane in 40% O 2 and 60% air mixture, and the inspired concentration of sevoflurane was targeted to maintain a BIS score between 40-60. Peak inspiratory pressure was strictly maintained between 11 and 13 cmH 2 O not to effect ICP.
Following endotracheal intubation, the ear-nose-throat (ENT) specialist placed the Boyle-Davis mouth gag. The extension of mouth opening and head position of patients were adjusted by the same ENT specialist to enhance the exposure of adenoid and tonsillar tissue. When the placement of the mouth gag was completed, operating room (OR) anaesthesiologists took a photograph of the neck extension in the lateral view. The degree of neck extension was assessed by the angle between the Frankfort plane and horizontal plane of the operation table in natural position (Frankfort plane angle). The angle was calculated by using an application (Angles in Photos, 2015 kublaidos) (Fig. 1). The anesthesiologist recorded the average of 2 Frankfort plane angle measurements. A Frankfort plane was officially described in the anthropologic conference in Frankfort in 1884. It is an imaginary line passing from left orbitale to left porion point and it has been used as a reference plane for cephalometric studies. Recently, Frankfort plane angle was used for the assessment of neck flexion-extension in the study of Kobayashi et al (12). The horizontal plane was created by drawing an imaginary line touching the porion and passing parallel to the operation table while the table was in neutral position. A very recent study used a similar method to calculate the degree next extension to evaluate its effect on ONSD of children undergoing palatoplasty surgery (13). ONSD was measured by two investigators who had experience in over 50 cases. A linear 6-13 Hz probe (Fujifilm Sonosite, Bothwell, USA) was used for the sonographic measurements at four different time-points. A thick layer of water-soluble ultrasound-transmission jelly was applied over the left upper eyelid of each patient. Then the probe was gently placed over the eyelid without exerting excessive pressure. The probe was moved with careful attention to find the best image of optic nerve entering into the globe. The ONSD was measured 3 mm posterior to the globe (Fig. 2). The investigators measured ONSD 3 times from the same eye and recorded the average of these measurements at four different time-points: (T0) after induction of anaesthesia, (T1) after endotracheal intubation, (T2) after the placement of the mouth gag and (T3) 20 minutes after the placement of the mouth gag. At each time-point, heart rate (HR), mean arterial pressure (MAP), EtCO2 and nasopharyngeal temperature (temp) were also recorded. The haemodynamic parameters, temp and EtCO2 which are known to be associated with intracranial pressure (14), were maintained in normal ranges in order to minimize their effects on ONSD.
The primary out-come of the study was the change in ONSD measurements between T3 and T4 and the secondary out-come was the effect of degree of next extension on mean ONSD changes between T2 and T3 time-points.

Sample size
Power estimation analysis conducted a priori concluded a sample size of 30 with 80% power with alpha error of 0.05, then we decided to include 35 patients with assuming possible drop-outs. The mean ONSD measurement in healthy pediatric population is 3.08±0.36 mm. An increase ³ 0.3 mm in mean ONSD measurement (10% of mean ONSD value in healthy pediatric population) was considered clinically significant (15). Considering a 0.05 significance level for type 1 error and 0.20 significance level for type two error, the collected data was sufficient for the power of statistical tests that were used.

Statistical Analysis
Statistical analyses were conducted using SPSS version 25 (made by SPSS Incorporated, located in Chicago, Illinois, USA). All continuous variables including age, weight, globe size, ONSD, EtCO 2 , Temp, HR, MAP are presented as mean and standard deviation (±SD), and the categorical variables, gender and surgical type, are presented as both numbers and percentile (%). The relationship between Frankfort plane angle and ONSD changes analyzed by regression model. Linear mixed model was used to observe the variation of repeated ONSD measurements and the other parameters (EtCO 2 , Temp, HR, MAP) over time. Post hoc analyses were performed using Bonferroni correction for multiple comparisons, for time-level as pairwise comparisons, since the time wise differences were statistically significant in all parameters observed. Additionally, figure for parameters changed by time is plotted, and presented. A p value of < 0.05 was considered significant for the analyses conducted.

Results
A total of 35 children were included in the study. A description of enrollment is summarized in Fig 3; 22 of them were male and mean age was 7.3±2.75 years. The demographic data of the patients were listed in Table 1.
Mean ONSD measurement was 4.56±0.41 mm at T0, 5.25±0.58 mm at T1, 5.92±0.63 mm at T2 and 6.28±0.55 mm at T4. The maximum increase in ONSD values was detected after intubation (0.69±0.06 mm) and immediately after the mouth gag placement (0.67±0.07 mm). According to the Pairwise Comparisons (time to time), the mean difference between T0 and T1 was calculated -0.07 (with CI -0.09,-0.05), the difference between T1 and T2 was calculated -0.07 (with CI -0.09,-0.05) and the difference between T2 and T3 was calculated -0.04 (with CI -0.05,-0.02). At all comparisons, the difference was found significant (p<0.001). The comparisons of mean ONSD values between different time-points were listed in Table 2.
Mean Frankfort plane angle was calculated as 130.41±7.5° (minimum 114.23° and maximum 144.65°). We assessed the relation between Frankfort plane angle and mean ONSD changes between T2 and T3 time-points to evaluate the effect of neck extension on ONSD measurements. According to regression model, there was no relation between the degree of Frankfort plane angle and mean ONSD changes (b = 0.63, p=0.715).
We summarized the hemodynamic parameters and other variables potentially affecting ICP, and consequently ONSD measurements, in Table 3. According to Greenhouse-Geisser analysis, EtCO 2 , temp, HR and MAP values changed over time. The maximum mean HR (110±17 beats/min) and MAP values (91.2±11.3 mmHg) were recorded immediately after the mouth gag placement (T2). Although HR and MAP significantly decreased 20 minutes after the mouth gag application, mean ONSD measurement increased by 0.36±0.04 mm between T3 and T4 time-points (p<0.001).

Discussion
In the current study, we evaluated the effects of the mouth gag placement and degree of neck extension on ONSD measurements. We detected significant increases in ONSD immediately after the mouth gag placement and additionally ONSD values continued to rise 20 minutes after the mouth gag application. However, the degree of neck extension as assessed by Frankfort plane angle had no effect on ONSD measurements.
Previously, Padayachy et al (16) analyzed 174 children and they reported that the optimal cutoff value for detecting an ICP ≥ 15 mmHg in the children > 1 year old was 5.49 mm with a sensitivity of 93.7%, specificity of 74.4% and for ICP ≥ 20 mmHg, the cutoff value of ONSD was measured as 5.75 mm with a sensitivity of 85.9 %, a specificity of 70.4 % . In our study, the mean ONSD value was 5.25±0.58 mm after endotracheal intubation. However, it was measured as 5.92±0.63 mm immediately after the mouth gag placement and ultimately, it increased over 6.0 mm (6.28±0.55 mm) only 20 minutes after the mouth gag placement. According to the data provided by previous studies (8,16), the children in our study probably had an elevated ICP 20 min following the mouth gag application. Direct laryngoscopy is well known to cause marked stress response with hypertension, tachycardia, arrhythmias and a significant increase in ICP and IOP (17). These alterations may result in adverse events especially on cardiovascular and nervous systems (18). Previous studies evaluated different drugs (19) and intubation techniques (20,21) to attenuate this response. Previously, An et al (3) showed that the mouth gag placement for exposure of pharyngeal tonsils during T and AT surgeries caused significant increases in HR and MAP measurements of paediatric patients. The authors stated that these haemodynamic changes were similar to haemodynamic response caused by direct laryngoscopy. According to these data, we hypothesized that the mouth gag placement during surgeries would cause a significant increase in ONSD measurements and ONSD values would probably remain to increase as long as the mouth gag application continued. In our study, mean ONSD value increased by 0.36±0.04 mm and passed the pathologic cutoff values in 20 minutes after the mouth gag placement. The hemodynamic responses caused by laryngoscopy are believed to be induced by direct contact of the blade with the posterior third of the tongue, manipulation of the richly innervated epiglottis and insertion of endotracheal tube between the vocal cords (22). During the mouth gag placement, a similar blade contacts directly to the tongue and the suspension system causes contractions in the muscles of mastication. As the oropharynx is a sensory organ capable of initiating sympathetic reflexes (23), the catecholamine release due to the mouth gag application is not surprising. However, we found the effect of the mouth gag placement on ONSD measurements much more significant than the effect of direct laryngoscopy. Furthermore, mean ONSD values continued to rise as long as the gag remained in the mouth. Whereas, increased ICP is known to reduce cerebral perfusion pressure and regional oxygenation that may result in postoperative neurological complications (24). Although the mouth gag remains for a short time in usually healthy children during A and AT surgeries, it has to be kept in mouth for significantly longer durations during some other procedures such as cleft palate repair, obstructive sleep apnea surgery and intraoral tumor excision. Consequently, long term placement of the mouth gag may cause deteriorative results especially in patients with comorbidities during longer procedures and a limitation might be concerned for the duration of the mouth gag application.
Previously, Panjabi et al (25) reported that a rotation of upper cervical spines over 20° in the sagittal plane exceeds the normal ranges of physiological motion. The authors calculated the angle by three reference lines passing through C1, C2 and C3 cervical spines in the radiologic view. According to these reference lines, Erden et al (26) reported that endotracheal intubation with Macintosh blade caused a maximum of 19.4° movement in C1/C2 spines. This brings the idea that excessive extension of neck during the mouth gag application is likely to exceed the normal ranges of cervical motion. In addition, placement of the mouth gag during ENT surgeries was shown to be associated with temporomandibular joint dysfunction, pain and trismus in the postoperative period. These adverse effects were reported to be related with the duration of the gag application (27). Considering this information, we evaluated the effect of neck extension angle on ONSD measurements. However, we could not detect any relation between degree of neck extension and ONSD measurements of children according to regression model. Previous studies showed that some intraoperative factors such as haemodinamic parameters, EtCO2 and nasopharyngeal temp may effect ONSD by changing ICP. An increase in carbon dioxide causes an increase in ICR by vasodilating the blood vessels, while a decrease in carbon dioxide decreases ICP and probably ONSD (14). Similarly, hypothermia which is frequently seen under general anesthesia, is known to decrease ICP (28). In the current study, although time-wise differences were detected in EtCO2 and temp, the parameters were within normal ranges throughout the operations. The effect of EtCO2 and temp were not considered as clinically important. Haemodinamic parameters increased after endotracheal intubation and the mouth gag application. Although HR and MAP significantly decreased 20 min after the gag application, ONSD values continued to increase. According to these findings, intraoperative parameters were not thought as the main reason for increased ONSD values.
The main limitation of the study is that we couldn't have a chance to evaluate the postoperative impacts of increased ONSD. All participants were healthy children and the duration of the mouth gag application was relatively short than some other surgeries. Procedures with longer duration of the mouth gag placement, such as intraoral tumor excision, would be more helpful in determining postoperative cognitive deteriorations. However, postoperative cognitive deterioration was not an out-come of the current study. Secondly, we could not measure ONSD of the children prior to induction. Mean age of the children was 7 years, therefore, we could not assess ONSD while the patients were awake. As a result, we could not provide pre-operative data. In order to compensate this limitation, we defined the primary out-come as the changes in ONSD measurements between the mouth gag application and 20 min after the mouth-gag application. We did not consider the changes between the first (following induction of anaesthesia) and the last (20 min after the mouth gag application) measurements. Last of all, the assessment of neck extension was based on Frankfort plane angle which was calculated by a phone application. The lines passing through Frankfort plane and horizontal plane of the operation table in natural position were drawn manually. Although we measured the angle three times for each patient, miscalculation is still possible.

Conclusions
Placement of the mouth gag causes significant increases in ONSD measurements of children. Therefore, the duration of the mouth gag application might be limited during surgeries.

Ethics approval and consent to participate
Ethical approval for this study was granted by the Mugla Sitki Kocman University Training and Research Hospital Biomedical Research Ethics Committee (approval number: XII) on 26.04.2018. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consents were obtained from the parents/guardians of all individual participants included in this study

Competing interests
The authors declare no competing interests.

Funding
There was no study funding or competing interest.  Continues numbers presented as mean ± standard deviation p value obtained with greenhouse-geisser, within-level p value represents the time wise change in overall.  The ultrasonographic view of the axial axis of optic nerve. Optic nerve sheath diameter was measured between the A and B points at 3 mm posterior to the globe.

Figure 3
Flowchart of the study.