Classification of postoperative edema based on the anatomic division with mandibular third molar extraction

Purpose Several investigations have been performed for a postoperative edema after extraction, but the results have been controversial due to low objectivity or poorly reproducible assessments of the edema. The aim of this study was to suggest a classification and patterns of postoperative edema according to the anatomical division associated with extraction of mandibular third molar as a qualitative evaluation method. Methods This study was conducted forty-four mandibular third molars extracted and MRI was taken within 48 h after extraction. The postoperative edema space was classified by MRI (one anatomic component—buccinator muscle—and four fascial spaces—supra-periosteum space, buccal space, parapharyngeal space, and lingual space), and evaluated independently by two examiners. The inter-examiner reliability was calculated using Kappa statistics. Results The evaluation of buccinator muscle edema showed good agreement and the fascial spaces showed constant high agreement. The incidence of postoperative edema was high in the following order: supra-periosteum space (75.00%), buccinator muscle (68.18%), parapharyngeal space (54.55%), buccal space (40.91%), and lingual space (25.00%). Conclusion Postoperative edema could be assessed clearly by each space, which showed a different tendency between the anatomic and fascial spaces.


Background
Edema is the swelling of a part of the body due to fluid buildup in the tissues and is one of the major discomforts for patients after the extraction of a third molar with pain and trismus [1]. Inflammatory mediators are released after the surgical extraction, and an increase in vascular dilatation and permeability results in postoperative edema. Several studies have been conducted to evaluate and reduce the postoperative edema using modified surgical techniques, postoperative medication, and physiotherapy [2][3][4]. The postoperative edema gradually reaches the maximum by 48 h [3] and regresses by the fourth day with resolution 7 days after extraction [4]. The most common methods for evaluating postoperative edema are the subjective scale and objective craniometrics, measuring the overall facial swelling, using a range of indicators, such as flexible tape, drawing, or silk [5].
The conventional objective evaluation methods, on the other hands, have limitations in that the assessment is only for overall edema with low reproducibility of the measurement [6]. Several methods were reported to overcome these limitations [5,7]; however, magnetic resonance imaging (MRI) is still the most objective assessment for the extent and area of swelling. Despite this, few studies have used MRI to measure swelling after a third molar extraction because of the cost and limited utility. Therefore, it is necessary to suggest new parameters for a postoperative swelling evaluation beyond the measurement of the swelling on MRI.
The military hospital has unique characteristics due to the nature of the lack of medical accessibility for military patients and can provide the medical service to the soldiers for free. With regard to the high risk of complications such as nerve damage after the extraction, the most accurate method, including MRI, can be used immediately after the extraction in the military hospital. Accordingly, we tried to develop an edema measurement method that can be used even in the absence of an MRI, using our MRI information from military hospitals.
An impacted mandibular third molar is adjacent to several anatomical structures that the postoperative swelling could spread. Two anatomical components (periosteum and buccinator muscle) and four fascial spaces (buccal, parapharyngeal, sublingual, and submandibular space) were considered to be clinically meaningful among the anatomic spaces around an impacted third molar. The aim of this study is to classify postoperative edema with MRI according to the anatomical division and to suggest patterns of the edema according to the anatomical spaces associated with the extraction of an impacted mandibular third molar.

Materials and methods
This retrospective study included adult patients, who visited for the mandibular third molar extraction on the Department of Oral and Maxillofacial Surgery of Armed Forces Capital Hospital from May 2018 to February 2019. The inclusion criteria of the patients were as follows: (1) healthy status without any underlying disease, (2) impacted mandibular third molar which is horizontal (80 to 100°) with less than half of the third molar crown above the CEJ of the adjacent second molar and contacted with the inferior mandibular canal on computed tomography ( Fig. 1) [8,9], and (3) facial MRI within 48 h after surgical extraction of their mandibular third molar. The exclusion criteria were as follows: (1) simultaneous extraction of impacted maxillary third molar with incision, ostectomy, or odontomy, (2) poor oral hygiene control, and (3) uncontrolled systemic diseases.
Surgical extraction of the mandibular third molar was performed with operculectomy. After operculectomy, the extraction was performed through odontomy on the exposed third molar with or without additional mesial incision [8]. Patients were instructed to take oral antibiotics (625 mg, amoxicillin, Ilsung Pharmaceutical, Korea) and NSAID (500 mg, dexibuprofen, Samil Pharmaceutical, Korea) thrice daily for 5 days and daily mouth rinse with a chlorhexidine solution.
Ethical approval was approved by the Institutional Review Board at Armed Forces Capital Hospital (No. AFCH-19-IRB-008) and followed the STROBE Guidelines with the Helsinki Declaration.
In MRI, the spaces around the mandibular third molar were divided into the buccinator muscle, supraperiosteum space, buccal space, parapharyngeal space, and lingual space (sublingual and submandibular space) [10,11]. Each space was defined as follows: 1. Buccinator muscle (Fig. 2a). An anatomical component organized from the origin (from the alveolar processes of the maxilla and mandible, buccinators crest, and temporomandibular joint) to insertion (in the fibers of the orbicularis oris muscle). 2. Supra-periosteum space (Fig. 2b). When enhancement is observed along the outer border of the mandible on MRI, it corresponds anatomically to the periosteum, but it is defined as the supra-periosteum space because the periosteum cannot be stretched enough to allow fluid collection and is attached firmly to the underlying bone. 3. Buccal space (Fig. 2c). The fascial space with a buccal fat pad consists of anterior (angle of the mouth), posterior (masseter muscle), superior (zygomatic process of the maxilla and zygomaticus muscles), inferior (depressor anguli oris muscle and attachment of the deep fascia to the mandible), medial (buccinator muscle), and lateral (platysma muscle, subcutaneous tissue, and skin) borders. 4. Parapharyngeal space (Fig. 2d)

Statistical analysis
The inter-examiner reliability was calculated using Kappa statistics. Kappa values were rated as follows: < 0.200 was considered poor, 0.201-0.400 fair, 0.401-0.600 moderate, 0.601-0.800 good, and > 0.800 excellent [12]. Except for the inconsistent postoperative images of edema between the two examiners, the correlation of the postoperative incidence of edema among each space was analyzed using a Pearson Chi-square test. In addition, the differences between the patients who underwent extraction only mandibular third molar or with maxillary third molar, and under local or general anesthesia were analyzed using a Fisher exact test and Chi-square test. Two-sided P values of < 0.05 were considered significant. The analysis was performed using SPSS 25.0 for Windows (SPSS Inc., Chicago, IL, USA). The continuous variables were expressed as mean ± standard deviation (SD) and qualitative variables as absolute and percent frequencies.
Overall, high mean Kappa values were calculated, and the inter-examiner agreement was constant among the spaces. Except for the nine extractions with inconsistent MRI images, 44 extractions were analyzed for the postoperative edema space tendency according to the variables. Overall, the incidence of postoperative edema was 75.00% on the supra-periosteum space, 68.18% on the buccinator muscle, 40.91% on the buccal space, 54.55% on the parapharyngeal space, and 25.00% on the lingual space. Among the spaces, the incidence of edema in the buccinator muscle and supra-periosteum space showed a significant correlation (P = 0.009, Fig. 4).

Discussion
The authors hypothesized that postoperative edema related with the surgical extraction of impacted mandibular third molar could be separated by anatomic divisions. The periosteum, as a highly vascularized tissue, could play a role in edema drainage [13]. In this study, edema was observed frequently along the outer border of the mandible. Radiologically, this condition is barely regarded as edema of the periosteum because it is a thin Simultaneously extraction of maxillary third molar (only mandibulular third molar:with maxillary third molar) 9:35 structure and attaches firmly to the underlying bone. Nevertheless, there were also no other anatomical structures known to exist along the mandibular surface. The authors assumed that it was either a loose connective tissue space for muscular fiber insertion into the periosteum or an unknown fascial space, and it was defined as the supra-periosteum space. As a result, the five postoperative edema spaces (one anatomic structure-buccinator muscle-and four fascial spaces-supra-periosteum, buccal, parapharyngeal, and lingual space) were classified clearly with substantial reliability. The frequency of postoperative edema was observed in the order of the supra-periosteum space; buccinator muscle; and parapharyngeal, buccal, and lingual spaces. A significant correlation was observed between the incidence of edema on the supra-periosteum space and the buccinator muscle (P = 0.009). Therefore, most postoperative edema could anteriorly spread to the supraperiosteum space and buccinators muscle. In contrast to the muscle fibers, which are inserted into the bone and have natural circulation to the bone, the fascial spaces do not have connected anatomic structures with the bone [10]. Without the invasion of the anatomic structure during the operation, postoperative edema may have a difficulty to spread toward the fascial space. Considering the extraction process, a distal (ramus area)  incision, with the invasion of pterygomandibular raphe, might be related to edema in the parapharyngeal space. The damage to the lingual plate might be related to lingual space edema. With regard to reduced swelling after surgical extraction, the basis could be established by the pattern of postoperative edema in each space. The clinical swelling could be associated with edema of the buccinator muscle, buccal, and supra-periosteum spaces, which are located in the buccal side of the mandible. The parapharyngeal and sublingual spaces could be associated with neck and swallowing discomfort. Considering this edema pattern, the authors noted the possibility of anatomic discontinuity on the buccinator muscle or periosteum to allow edema to spread into the buccal space. The buccal space is one of the fascial spaces and its volume can be extended easily to cause facial swelling and drain slowly via the cutaneous sinus at the inferior of the space [14]. In accordance with some studies, this possibility was supported by their results of significant swelling in the flap design with a mesial vertical incision [15][16][17]. In addition, this possibility supported a clinical study of flap design for guided bone regeneration in that flap management without a vertical incision can avoid cutting the muscle fibers, leading to a decrease in postoperative swelling and wound dehiscence [18]. Therefore, an incision without invasion of the anatomic structures, such as the buccinator muscle and periosteum, can reduce the level of postoperative edema.
Many studies have been conducted on the contributing factors, including incision design and affected postoperative swelling [2]. However, previous research on postoperative swelling has been controversial because the clinical methods used to evaluate swelling (such as a questionnaire or distance measurement between facial reference points) were not objective and reproducible [5,6]. Therefore, the authors suggest region-specific edema assessment of mandibular third molar extraction as a qualitative method. The postoperative edema in each space could be classified independently, leading interpretation of various postoperative edema degrees and allowing the symptoms to be estimated according to each space. In the limitation of this retrospective study, the authors could not correlate the radiological (MRI) edema with surgical procedure and clinical symptoms. Further clinical studies should be necessary to reveal and reduce postoperative edema by considering each divided space.

Conclusion
An evaluating postoperative edema based on anatomic division showed high reliability according to the space classification, which was divided into five spaces: one anatomic structure (buccinator muscle) and four fascial spaces (supra-periosteum, buccal, parapharyngeal space, and lingual space). Most postoperative edema was spread though buccinators muscle and supra-periosteum space with a significant correlation. Further study should be conducted for the clinical validation of this classification according to the relevance with the swelling.