Does the Coronectomy a Feasible and Safe Procedure to Avoid the Inferior Alveolar Nerve Injury during Third Molars Extractions? A Systematic Review

This systematic review aimed to review the literature on the coronectomy technique and evaluate the incidence of success and complications as a surgical approach for inferior third molars. Online databases were searched for data on the frequency of inferior alveolar nerve damage, lingual nerve damage, root migration, pain, infection, dry socket, and extraction of the remaining root, and data on the necessity of reintervention were also extracted. Randomized clinical trials, controlled clinical trials, prospective cohort studies, and prospective and retrospective studies with or without the control group were intercepted. This systematic review was registered in PROSPERO (CDR 42020135485). Sixteen papers analyzed 2176 coronectomies in total, and only five of them were judged as appropriate according to methodological quality assessment. The incidence of inferior alveolar nerve injury was documented in 0.59% of the procedures, lingual nerve injury in 0.22%, infection 3.95%, dry socket 1.12%, extraction of the root 5.28%, and reintervention 1.13%. The pain was the most reported, in 22.04% of the population. This study provides an overview of the clinical success and complications of coronectomy, and their prevalence. A coronectomy may be considered a low-risk procedure and an option for treatment to avoid potential damage to nervous structures. However, patients should still undergo a full screening and evaluation of postsurgical procedures.


Introduction
The procedure of extraction of third molars accounts for from 35.9% to 58.7% of all oral surgical procedures, and is the most common procedure performed by oral maxillofacial surgeons [1][2][3]. The indications of the extraction of third molars according to the American Association of Oral and Maxillofacial Surgeons (AAOMS) are when the patient presents pain, ectopic position, tooth fracture, unrestorable caries, internal or external resorption of the tooth or adjacent teeth, tooth associated with cystic lesion or tumor, acute, chronic infection, or periodontal disease [4].
One of the possible complications, and the most common during the extraction of inferior third molars, is the IAN injury, with an incidence of from 0.5% to 8% [5][6][7][8]. The temporary paresthesia of the IAN ranges from 0.26% to 8.4%, with a permanent loss of more than 3.6%. Another injury that can occur is to the LN, which is less common, ranging from 0.1 to 22% [9]. The orthopantographic image can show the proximity of the third molar to the IAN. Rood and Shehab described seven signs that indicate the proximity of these structures: deviation or narrowing of the canal, periapical radiolucency, narrowing, darkening and curving of the roots, and loss of lamina dura over the wall of the mandibular canal all recorded a 30% incidence of injury [10]. When these radiographic markers are present, the indication for cone-beam-computed tomography (CBCT) is the gold standard, assisting in surgical procedure and avoiding damage to the IAN.
A new possibility of treatment for these cases of third molars with a high proximity to the IAN was first described in 1989 by Knutsson et al. [11]. The coronectomy consists of removing the crown, while leaving the roots intact, preventing injury to the IAN. Pogrel et al. described the sequence and basic rules needed to perform the technique in more detail: the crown and a sufficient part of the coronal portion of the roots must be removed until they are from 2 to 3 mm below the level of the alveolar crest [11][12][13]. There are few complications associated with the coronectomy, such as pain, postoperative infection, dry socket in the short-term, and healing delay and root migration in the long-term.
This study aims to review the literature on the coronectomy technique and evaluate the incidence of the success and complications of coronectomy as a surgical approach for inferior third molars. The evaluation of these complications was based on eight criteria: the frequency of inferior alveolar nerve damage, lingual nerve damage, root migration, pain, infection, dry socket, and extraction of the remaining root and necessity for reintervention. The selection of these criteria was chosen according to the frequency with which they are discussed in the literature.

Protocol and Registration
This SR followed the recommendations of the PRISMA guidelines [14]. The protocol for this SR was based on PRISMA-P [15] and registered in the PROSPERO database under the number CDR 42020135485. There were no deviations from the initial protocol.

Focused Question Based on PICOS Strategy
A systematic search of the literature was conducted to identify papers regarding the main question that was defined according to the population, intervention, comparison, outcome (PICO) format [16]. What are the clinical outcomes and complications (O) of third molar surgery coronectomy (I), with no comparison (C), in patients with the third molars suspect to be or in contact with the IAN, without any pathology associated (P)?

Outcome Measures
The primary outcome variable was IAN damage. The secondary outcomes were LN damage, root migration, pain, infection, dry socket, extraction of the remaining root, and the necessity for reintervention.

Search Strategy
The search was conducted from inception until June 2020, without other restrictions on dates or language, on the following search platforms: PubMed/MEDLINE, Web of Science, Scopus, LILACS, and the Cochrane Central Register of Controlled Trials. A search of the grey literature was also carried out through the Literature Report database.
[  Cochrane Library third molar * OR wisdom tooth OR wisdom teeth OR tooth extraction * OR surgical removal in Title, Abstract, Keywords, and partial odontectom * OR coronectom * in Title, Abstract, Keywords and nerve injury OR alveolar nerve injury OR mandibular nerve * OR inferior alveolar nerve * OR paresthesia * OR dysesthesia * OR lingual nerve * OR lingual nerve injur * in Title, Abstract, Keywords in Trials

Eligibility Criteria and Study Selection Process
This SR included studies that analyzed the clinical outcomes of coronectomy procedures in patients with third molars suspect to be or in contact with the IAN, without any associated pathology, with a follow-up period over 3 months. Randomized clinical trials (RCT), prospective (PS), and retrospective (RS) studies, with or without a control group (third molar extraction), were included. Case reports, in vitro studies, letters to authors, and literature reviews were excluded from the process.
The process of searching and selecting studies was conducted in duplicate by two authors (R.C.S.P. and T.C.G). The titles and abstracts were carefully evaluated according to the eligibility criteria of this SR for relevance. A consensus was established to resolve any disagreement that occurred.

Data Synthesis
The data from the manuscripts were extracted by R.C.S.P., in cooperation with T.C.G., and arranged as follows: author name, study design, number of coronectomy procedures, occurrence and number of IAN damage, LN damage, root migration, pain, infection, dry socket, extraction of the remaining root and the necessity for reintervention.

Assessment of the Risk of Bias
All eligible studies were assessed for methodological quality by two reviewers (R.C.S.P. and T.C.G). For PS, RS, and case-control articles; the quality was rated based on the Newcastle-Ottawa Quality Assessment Scale (NOS) [17]. Using the NOS, each study is judged on eight items and categorized into three groups: the selection of the study groups, the comparability of the groups, and the ascertainment of either the exposure or outcome of interest for case-control or cohort studies, respectively. RCTs involve assigning a judgment of high, low, or unclear risk of material bias for each item considered, using the Cochrane Collaboration's tool [18] for guidance. In agreement with this, six predetermined criteria were used: sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other sources of bias. Material bias was then defined as a bias of sufficient magnitude to influence the results or conclusions of the trial, recognizing the subjectivity of any such judgment.

Statistical Analysis
The extracted data (e.g., IAN and LN damage, root migration, pain, infection, dry socket, extraction of the remaining root, and a necessity for reintervention) were pooled and analyzed qualitatively and quantitatively (variation, mean, and standard deviation) through descriptive statistics. For descriptive analysis, the StatPlus software (version 6, AnalystSoft, Walnut, CA, USA) was used. A meta-analysis pairwise cannot be performed due to the design and heterogeneity between the studies.

Literature Search
The primary search, post the removal of duplicates, obtained 167 articles: 91 from MEDLINE/PubMed, 15 from the Cochrane Central Register of Controlled Trials, 115 from Scopus, 130 from Web of Science, 2 from LILACS, and no article in the Grey Literature Report of the New York Academy of Medicine. After the initial search, the collected data were analyzed by two independent reviewers (R.C.S.P. and T.C.G.), and the manuscripts were systematically excluded after screening titles and abstracts for relevance. Of these, 137 articles were excluded according to the exclusion criteria. Articles for which the titles and abstracts were missing sufficient information to determine eligibility according to the inclusion and exclusion criteria underwent a full-text review.
Thirty full-text records were screened for eligibility. In total, sixteen articles were included for qualitative analysis. A PRISMA diagram of the record selection process is presented in Figure 1, in which it is possible to better illustrate the search and selection process as well as the reasons given for the eliminated articles.
Of the sixteen inserted records, all the studies presented a minimum of 20 coronectomies, a follow-up period of more than 3 months, and contain at least four of the eight evaluation criteria: IAN and LN damage, root migration, pain, infection, dry socket, extraction of the remain root and necessity for reintervention. A total of 2176 coronectomies were included, and the evaluation showed a low risk of complications after the coronectomy.
The most reported complication was pain in 22.04% of the procedures. The IAN injury was documented in 0.59% of the proceedings, LN injury in 0.22%, infection 3.95%, dry socket 1.12%, extraction of the root 5.28%, and reintervention necessity 1.13%.
Of the sixteen inserted records, all the studies presented a minimum of 20 coronectomies, a follow-up period of more than 3 months, and contain at least four of the eight evaluation criteria: IAN and LN damage, root migration, pain, infection, dry socket, extraction of the remain root and necessity for reintervention. A total of 2176 coronectomies were included, and the evaluation showed a low risk of complications after the coronectomy.
The most reported complication was pain in 22.04% of the procedures. The IAN injury was documented in 0.59% of the proceedings, LN injury in 0.22%, infection 3.95%, dry socket 1.12%, extraction of the root 5.28%, and reintervention necessity 1.13%.  [23] and Pedersen et al. [25]) reported the occurrence of IAN damage post coronectomy procedure. Mendes et al. [23] presented 2.85% (one case) and Pedersen et al. [25] 2.16% (five cases) differently to the average, which was 0.59%. All the other articles cited that no patient experienced numbness, swelling or a thickening sensation of the lower lips after surgery after the procedure.

LN Injury
Compared with the IAN damage (total of thirteen cases), the LN injury (four cases) was less common. Of the sixteen studies, only three (Mukherjee et al. [26], Pedersen et al. [25], and Pogrel et al. [13]) reported the occurrence of this neurological alteration.
All of the remaining articles assessed no occurrence of damage on the lingual nerve following coronectomy/complete extraction.

Root Migration
Eleven authors (Agbaje et al. [ [29]) recorded the occurrence of root migration.
The high incidence (more than 80% of their cases) of root migration was reported in Kang et al. [20]; Hatano et al. [28]; Kohara et al. [27]; Kouwenberg et al. [3]; Mendes et al. [23]. Coronal migration of the roots was reported as the most commonly reported long-term consequence of coronectomy, which is confirmed by our review.

Pain
The onset of pain was evaluated in eleven of the included articles (Agbaje et al. [ [29]).
The assessment of pain with a Visual Analog Scale (VAS) post coronectomy was present in Mendes et al. [23] and Mukherjee et al. [26]. Both authors analyzed the first week after the procedure, but Mukherjee et al. [26] cited a case of one patient that reported discomfort in mouth-opening for a period of two months, which was attributed to enamel lipping left behind intraoperatively, and subsided spontaneously. The other studies evaluated pain based on the patient's complaint.

Infection
Comparing the quantity of all the procedures performed by the authors that analyzed these criteria (Agbaje et al. [ [29]) the percentage of infection was 3.95% (73 cases), which was higher in Pedersen et al. [25], with 27 cases (11.68%), more than one-third of the total cases. The authors of this study justified this expressive rate as being similar to corresponding studies after both coronectomy and total removal of mandibular third molars. Another hypothesis is that the indication for the total removal of these teeth was adequate.

Dry Socket
The incidence of the early complication included in this study was 16 cases (1.12%) of all articles that appraised this criterion (1428 coronectomies). In Cilasiun et al. [6], Leung et al. [8,22], and Pederson et al. [25], there were no cases of dry sockets. In the randomized study of Renton et al. [29], the percentage of dry sockets was 12.06%, the highest of all, followed by Agbaje et al. [19], with 4.16%.

Extraction of the Remaining Root
The necessity of extraction of the remaining root occurred in all studies. This later complication occurred in 115 cases (5.28%) of all the 2176 coronectomies. Discriminating the quantity of extraction of the remaining roots, Kouwenberg et al. [3] had almost a triple percentage of this complication (11.26%), followed by Mukherjee et al. [26] with 10%, Agbaje et al. [19] with 9.37% and Leung et al. [22] with 9.35%.
The study conducted by Mukherjee et al. [26] had two patients with failed coronectomy, with the mobilized roots removed. Both cases were female patients with conical root morphology.

Necessity for Reintervention
The necessity for reintervention excluded the extraction of the remaining root, with only the incomplete removal of the crown or enamel retention and the exposure of the root in the oral cavity. This late complication is related to root migration. Only six articles assessed these criteria.
Comparing the number of coronectomies that evaluated these criteria, the percentage of this complication was 1.13% (12 cases), lower than Frenkel et al. [2] and Monaco et al. [24], who each presented four cases, two-thirds of the cases analyzed. Cilasun et al. [6] and Renton et al. [29] described any case of reintervention during their period of follow-up, of seventeen and thirteen months, respectively.
Diversely, for randomized clinical trials, the Cochrane Collaboration's tool was used. Figure 2 showed that the two included studies (Leung et al. [8]; Renton et al. [29]). Leung et al. [8] were best ranked, exhibiting a low risk of bias in the majority of indicators.

Discussion
This SR evaluated the rate of complications to analyze the success of coronectomy. This procedure is indicated in inferior third molars with proximity to or real contact with IAN. However, the coronectomy has other benefits beyond IAN preservation, such as avoiding LN injury and a less morbid procedure. For this, migration of the root that minimizes damage to the IAN with the extraction of the remaining root is necessary.
The higher limitation of the study was due to the heterogeneity of the studies included (PS, case-control, and RCT), making it necessary to perform a descriptive analysis of the eight criteria. The evaluation of the PS and RS studies using the NOS assessment demonstrated that only three (Hatano et al. [28]; Kang et al. [20] and Cilasiun et al. [6]) of the fourteen articles achieved a good methodological quality assessment. Additionally, the randomized studies (Renton et al. [29] and Leung et al. [8]) presented good qualifications. Due to the design (low number of clinical trials) and heterogeneity of the studies, it was not possible to carry out the meta-analysis pairwise.
In this SR, the IAN injury results were minimal, almost 0.6%, endorsing the idea of this being a good indication of technique for the extraction of third molars in proximity to the IAN, compared with the extraction of 18.6%, as cited in the Renton et al. [29]. Furthermore, the literature advises that there is a low possibility of LN injury, and our findings indicated the same, with injury being temporary and self-resolving.
The comparison of the loss of sensitivity between full extraction and coronectomy of the lower third molar was evaluated by Hatano et al. [28]. The extraction group consisted of six patients (5%) with signs of IAN injury (three of them were permanent). In contrast, in the coronectomy group, one patient (1%) complained of nerve symptoms postoperatively, but became asymptomatic within 1 month. Leung et al. [8] showed a postoperative neurosensory deficit in one patient (0.65%) of the coronectomy group, whereas nine cases (5.10%) occurred after the complete extraction. The findings of these studies corroborate the idea that coronectomy is a valid procedure to minimize the possibility of IAN damage.
The incidence of IAN injury post coronectomy in Frenkel et al. [2] was 0.5% (1 of 185 patients). This injury was temporary and manifested as hypoesthesia of the lower lip. Frenkel et al. [2], Kohara et al. [27] reported one hypoesthesia subject (1%) on a postoperative day, associated with surgical difficulties, which improved after 2 months.
The pattern of root migration is well-described in Leung et al. [30], showing that this migration is common in the first two years after coronectomy, and decreases after this period.

Discussion
This SR evaluated the rate of complications to analyze the success of coronectomy. This procedure is indicated in inferior third molars with proximity to or real contact with IAN. However, the coronectomy has other benefits beyond IAN preservation, such as avoiding LN injury and a less morbid procedure. For this, migration of the root that minimizes damage to the IAN with the extraction of the remaining root is necessary.
The higher limitation of the study was due to the heterogeneity of the studies included (PS, case-control, and RCT), making it necessary to perform a descriptive analysis of the eight criteria. The evaluation of the PS and RS studies using the NOS assessment demonstrated that only three (Hatano et al. [28]; Kang et al. [20] and Cilasiun et al. [6]) of the fourteen articles achieved a good methodological quality assessment. Additionally, the randomized studies (Renton et al. [29] and Leung et al. [8]) presented good qualifications. Due to the design (low number of clinical trials) and heterogeneity of the studies, it was not possible to carry out the meta-analysis pairwise.
In this SR, the IAN injury results were minimal, almost 0.6%, endorsing the idea of this being a good indication of technique for the extraction of third molars in proximity to the IAN, compared with the extraction of 18.6%, as cited in the Renton et al. [29]. Furthermore, the literature advises that there is a low possibility of LN injury, and our findings indicated the same, with injury being temporary and self-resolving.
The comparison of the loss of sensitivity between full extraction and coronectomy of the lower third molar was evaluated by Hatano et al. [28]. The extraction group consisted of six patients (5%) with signs of IAN injury (three of them were permanent). In contrast, in the coronectomy group, one patient (1%) complained of nerve symptoms postoperatively, but became asymptomatic within 1 month. Leung et al. [8] showed a postoperative neurosensory deficit in one patient (0.65%) of the coronectomy group, whereas nine cases (5.10%) occurred after the complete extraction. The findings of these studies corroborate the idea that coronectomy is a valid procedure to minimize the possibility of IAN damage.
The incidence of IAN injury post coronectomy in Frenkel et al. [2] was 0.5% (1 of 185 patients). This injury was temporary and manifested as hypoesthesia of the lower lip. Frenkel et al. [2], Kohara et al. [27] reported one hypoesthesia subject (1%) on a postoperative day, associated with surgical difficulties, which improved after 2 months.
The pattern of root migration is well-described in Leung et al. [30], showing that this migration is common in the first two years after coronectomy, and decreases after this period.
One of the studies [16] reported that 97% of the retained roots showed signs of migration, and 65% showed signs of rotation. Panoramic radiograph analysis showed that migration mainly occurred during the first year and that this migration led to a structural separation between the root complex and the mandibular canal in more than half of the cases. In addition, Kohara et al. [27] investigated the migration pattern of the retained roots postoperatively after 3 years. In the first year, a rate of 74.3% root migration was found. The study showed that root migration increased in the first 2 years after coronectomy, but stabilized between the second and the third year.
Our study evaluated whether this movement were present or not, and we found that half the cases had some degree of migration (50%). The most critical evaluation that those articles discussed was that the migration of the roots was moving away from the IAN, and some of them could expose of the root to the oral cavity, necessitating surgical reintervention for enamel edge trimming or extraction of the remaining root. All the articles included in this evaluation stated that this is a regular complication, and that it is essential to explain the possible necessity of a new procedure to the patient. The period of stabilization was similar to the one cited in Leung s article: approximately three years.
The principal, short-term complications that can be found after the coronectomy procedure are pain, infection, and dry socket. These complications may occur in the same proportion during tooth extraction [29].
The assessment of pain with the Visual Analog Scale (VAS) post coronectomy by Mukherjee et al. [26] showed that three (out of a total of 20) coronectomy sites developed pain after the first week of the procedure. One patient had pain on the soft tissue around the coronectomy site, with discomfort in mouth-opening for two months. This was attributed to enamel lipping left behind intraoperatively, and subsided spontaneously.
According to Leung et al. [8], among the coronectomy group, 41.9% of the teeth were reported to be in pain 1 week postoperatively. The corresponding proportion in the control group was 57.3%, which was statistically significant. However, there was no difference between the groups from 1 to 24 months after surgery.
Only one patient presented pain (1.13%) following the coronectomy, with no pain reported for the coronectomy group, according to Cilasun et al. [6]. These results cause a discrepancy with Cilasun et al. [6], Hatano et al. [28], who estimated a relevant difference in pain measures, with the incidence of postoperative pain being more significant in the coronectomy group. In the coronectomy group, 18.63% showed pain, compared to 6.78% in the extraction group; however, all pain decreased within 1 week. This information assists the professional in choosing the appropriate pain control medications. Renton et al. [29] considered that the selection of patients with a high proportion of difficult and deeply impacted teeth could be the reason for the higher rate of dry sockets. Agbaje et al. [19] were the only authors that described how their group managed the dry socket. The procedure consisted of wound debridement and irrigation, after which Alvogyl™ (Septodont; Saint-Maur-des-Fossés, France) paste-dressing material was placed to control pain: an effective treatment which led to appropriate healing.
Favorable factors to predict the failure of coronectomy, leading to a second surgical approach to the root extraction, may occur most often in women with conically rooted teeth that narrow within the nerve canal [29].

Conclusions
Coronectomy may be considered a low-risk procedure, and an option for treatment to avoid potentially more severe damage to nervous structures. The reintervention to remove the remaining roots or the reduction in the remaining roots may be considered part of the treatment because the possibility of this causing nerve injury reduces, due to the migration of the roots. This is the main reason to select coronectomy over the extraction of the third molar, close to or in contact with the IAN. However, patients should still undergo a full screening and evaluation of postsurgical procedures.