Influence of the anesthetic modality on the development of neurological injury after lower third molar extraction: A systematic review of the literature

The aim of this study was to determine if the risk of neurological injury to the inferior alveolar nerve (IAN) and the lingual nerve (LN) following the extraction of lower third molars is affected by the anesthetic modality (local anesthesia (LA) vs. general anesthesia (GA)). A systematic search was performed through the PubMed, Scopus, Cochrane Library, and Web of Science databases; furthermore, a manual search was performed by analyzing the references of full-text articles. From a total of 309 studies (collected after the removal of duplicates), 6 studies were selected. Of these, 4 reported a correlation between GA and nerve damage, while the other 2 did not show an obvious association. The level of bias in the studies was also calculated. Only 2 studies showed a medium risk of bias, while 4 studies showed a high risk of bias; no study showed a low risk of bias. Four of the 6 studies highlighted a higher incidence of IAN and LN injury following the extractions performed under GA. Although no scientific evidence is yet available, due to the scarcity and the limited quality of the studies in the literature, considering the risk–benefit ratio, LA should be the first choice in lower third molar surgery.


Introduction
The lower third molar is situated between the second lower molar and the mandibular ramus.
The primary reasons for the impaction of the wisdom tooth are to be searched in the lack of space, malposition, unfavorable eruption angulation, or physical impediments along the pathway of eruption. 1The etiology of tooth impaction may be associated with abnormalities in tooth development, and it is related to inherent genetic compo nents and specific environmental conditions. 2 The impaction of mandibular third molars is a condition associated with a different degree of difficulty during surgery and a higher risk of complications.The mesioangular im paction of third molars is the most frequent situation. 3Most common complications occur when the surgical removal is performed with the altered position of the wisdom tooth, which is generally more difficult, and also in elderly patients. 3,4omplications are usually local, like a hematoma or infections, and temporary, like iatrogenic inferior alveo lar nerve (IAN) or lingual nerve (LN) injury. 5,6For this reason, the 1979 conference of the American National Institutes of Health (NIH) suggested that third molars should be removed when there is evidence of pathological changes or irreversible pathology. 7n ethical terms, it is not advisable to perform a surgi cal procedure that carries a morbidity risk without valid indications.There are many indications for the extraction of the impacted lower third molar that are derived from the clinical symptomatology with the distinction between 'symptomatic' and 'asymptomatic' teeth.However, the term 'asymptomatic' is ambiguous, since the lack of symp toms should not be confused with the lack of pathology.Some diseases could remain asymptomatic before being diagnosed despite the presence of pathological (clinical or radiographic) signs.It is also important to consider clini cal situations where no pathology has been developed yet, but the predisposing factors are present, such as plaque accumulation, common in the case of partially erupted molars or teeth affected by dysodontiasis.
Intraoperative pain control is an intrinsic part of a surgi cal procedure.In most cases, third molar surgery can be performed under local anesthesia (LA), although in par ticular cases, general anesthesia (GA) is suggested.General anesthesia should be limited to those patients and clinical situations in whom/which LA cannot be used: unco operative patients; dental phobia; allergy to local anesthe tics; acute and extended infections; and extensive dental or maxillofacial surgery.In the literature, there are still no pre cise indications as to the choice of the anesthetic modality with regard to third molar surgery.On the other hand, con sidering that many types of risk, e.g., rare and unpredict able death, are associated with GA, this anesthetic choice should be taken into account only if necessary. 8herefore, the aim of this systematic review was to as sess the influence of various modalities of anesthesia, LA vs. GA, used for the extraction of the lower third molar on neurological injury to IAN and/or LN.

Material and methods
The present systematic review followed the Preferred Reporting Items for Systematic reviews and MetaAnalyses (PRISMA) statement 9 and used a previous systematic review as a template. 10he following PICOS (Population, Intervention, Com parison, Outcome, and Study design) criteria were set: -Population: patients undergoing lower third molar surgery; -Intervention: lower third molar extraction under LA; -Comparison: patients whose surgery had been performed under GA; -Outcome: neurologic injury to IAN and/or LN; and -Study design: systematic review.

Search strategy
The review was realized with the use of scientific databases: PubMed; Scopus; Cochrane Library; and the Web of Science, from the inception to the latest research in August 2019.A manual search was also performed among the references of all fulltext articles.No language restriction was applied.
The search of PubMed was performed using the follow ing algorithm: (local OR general) AND (anaesthesia OR an esthesia) AND (third OR 3 rd OR lower OR mandibular OR wisdom OR impacted) AND (molar OR molars OR tooth OR teeth) AND (extraction OR extractions OR removal OR removals)) AND (injury OR injuries OR damage OR lesion OR lesions OR disturbance OR disturbances) AND (lingual nerve OR mandibular nerve OR inferior alveolar nerve).The search strategy applied in Scopus involved the same Boolean string, including the article title, abstract and keywords.In the case of the Cochrane Library search, the "Advanced search" tool was used, choosing all content types.For the Web of Sciences, the "Advanced search" was used, selecting all languages and all document types.

Study selection
The selected studies had to meet the predefined eligi bility criteria.They had to be randomized clinical trials (RCTs), controlled clinical trials (CCTs), prospective co hort studies (PCSs), or retrospective studies (RSs), with or without a control group.
Studies that considered 2 groups of patients undergoing lower third molar extraction were included, and studies in which 3 groups were considered because of the addition of the LA plus sedation group were also taken into account.
Studies that investigated the risk of developing nerve injury (of IAN and/or LN) in a group of patients treated under GA, in a group of patients treated under LA and in a group of patients treated under LA plus sedation (if the 3 rd type of anesthetic modality was taken in considera tion) were collected.
Studies that distinguished temporary and permanent injury after the period of followup, with at least 6 months of followup were included.
Case report, case series, studies enrolling less than 10 subjects, comments, expert opinions, letters to the edi tor, reviews, and studies that analyzed the same sample as a preexistent study were excluded.
Studies that did not evaluate the anesthetic modality as a parameter in the development of neurologic injury and studies that evaluated a single type of anesthetic modality (only a group of patients treated under LA or GA) with regard to nerve injury after lower third molar extraction were excluded.
Studies that did not distinguish temporary and perma nent injury after the period of followup were excluded.
Redundant studies were excluded.

Data items
The following data was collected: study design; an esthetic modality (LA, GA and LA plus sedation); sample size; gender and age; number of teeth removed; operators' experience; surgical difficulty rated according to different classifications (Winter's classification, Wharfe's score, and Pell and Gregory's classification); use of a lingual retractor (if used, specifying the type); followup; nerve injury classified as temporary or permanent (persisting after 6 months); and clinical implications according to the authors.

Assessment of the risk of bias in individual studies
The risk of bias in individual studies was evaluated accord ing to the modified Downs and Black tool. 11The original Downs and Black tool 12 consists in the calculation performed by evaluating each study across 5 domains, including: -Reporting (10 items); -External validity (3 items); -Internal validity -bias (7 items) -Internal validity -confounding (6 items); and -Power (1 item).
The maximum possible score is 32.In the current re view, certain modifications were introduced to adhere to the studies dealing with the topic "nerve injury after third molar removal under local or general anesthesia".
These were as follows (Table 1): -Item 19 (Was compliance with the intervention/s reli able?) was not considered, as in the present research, the compliance of the patients was not required to evalu ate the results; Table 1.Modified Downs and Black tool used for the analysis of the risk of bias of non-randomized clinical trials (judgments and scores for each item as follows: no/not applicable (0); yes (1))

Domain Question
Reporting 1.Is the objective of the study clearly described?2. Are the main outcomes to be measured clearly described in the Introduction or Methods section? 3. Are the characteristics of the patients included in the study clearly described?4. Are the interventions of interest clearly described?5. Are the distributions of principal confounders in each group of subjects to be compared clearly described?6. Are the main findings of the study clearly described?7. Does the study provide the estimates of random variability in the data for the main outcomes?8. Have all important adverse events that may be a consequence of the intervention been reported?9. Have the characteristics of the patients lost to follow-up been described?10.Have actual probability values been reported (e.g., 0.035 rather than <0.05) for the main outcomes except where the probability value is less than 0.001?External validity 11.Were the subjects asked to participate in the study representative of the entire population from which they were recruited?12. Were those subjects who were prepared to participate representative of the entire population from which they were recruited?13.Were the staff, places and facilities where the patients were treated representative of the treatment the majority of patients receive?
Internal validity -bias 14.Was an attempt made to blind the study subjects to the intervention they received?15.Was an attempt made to blind those measuring the main outcomes of the intervention?16.If any of the results of the study were based on 'data dredging' , was this made clear?17.In trials and cohort studies, were the analyses adjusted for different lengths of the follow-up of the patients, or in case-control studies, was the time period between the intervention and the outcome the same for cases and controls?18. Were the statistical tests used to assess the main outcomes appropriate?19.Were the main outcome measures used accurate (valid and reliable)?
Internal validity -confounding 20.Were the patients in different intervention groups (trials and cohort studies) or were cases and controls (case-control studies) recruited from the same population?21.Were the patients in different intervention groups (trials and cohort studies) or were cases and controls (case-control studies) recruited over the same period of time?22. Was there adequate adjustment for confounding in the analyses from which the main findings were drawn?23.Was the loss of patients to follow-up taken into account?Power 24.Sample size calculation -Item 23 (Were the study subjects randomized to in tervention groups?) and Item 24 (Was the randomized intervention assignment concealed from both patients and healthcare staff until recruitment was complete and irrevocable?) were not considered, as not all of the in cluded studies were randomized; and -Item 27 (Power) was simplified to "Sample size calcula tion".Thus, the domains of the modified tool were as follows: reporting (10 items); external validity (3 items); internal validity -bias (6 items); internal validity -confounding (4 items); and power (1 item), with a maximum score of 25.The overall risk of bias was defined as follows: -high: total score ≤16; -medium: total score >16 and <22; and -low: total score ≥22.

Study search
The results of the automatic and manual search are summarized in Fig 1 .From a total of 309 articles re trieved, 7 of the 13 fulltext articles assessed for eligi bility were excluded for the reasons provided in Fig. 1 and Table 2.
Finally, 6 studies were judged eligible according to the inclusion/exclusion criteria (Table 3).

Study design
The 6 selected studies included 1 retrospective obser vational study (ROS), 8 3 prospective studies (PSs), [13][14][15] 1 prospective longitudinal study (PLS), 16 and 1 RS. 17n half of the selected studies, the extractions were per formed under LA or GA 8,14,17 ; in the remaining half, the extractions were performed under LA, GA or LA plus sedation. 13,15,16The sample size in the selected stu dies ranged from 387 13 to 6,803 subjects. 17All the stu dies included both males and females.In 3 cases, 14,16,17 information on gender distribution was not available.The mean age of patients ranged between 24.92 ±4.67 years 13 and 41.3 ±17.8 years 8 ; in 2 cases, the mean age of patients was not considered. 14,17he number of teeth extracted under LA was in the range between 105 16 and 631, 15 under GA, it was be tween 194 8 and 535, 13 and the range for LA plus seda tion was between 15 15 and 41. 13 Only Nguyen et al. did not specify how many teeth were extracted under LA or GA. 17 In 4 cases, the extractions were performed by sur geons with different levels of experience [14][15][16][17] ; in the study by Brann et al., the operator's experience was not considered. 13Only in 1 study, the extractions were performed by a single surgeon with high experience. 8Surgical difficulty was rated according to differ ent classifications.Three studies recorded the degree of surgical difficulty by means of Wharfe's score. 13,14,16ill et al. found a score of 5.5 without differentiating the scores between LA and GA, 14 the other 2 studies found similar Wharfe's scores for the subjects treated under LA and the subjects treated under GA. 13,16Winter's clas sification was used by Gülicher and Gerlach to find out the prevalence of the positions of the extracted teeth (horizontal (53),vertical (350), mesioangular (490), disto angular (159), transversal (48), and other ( 6)) without specifying the anesthetic modality. 15Costantinides et al. assessed surgical difficulty by using Pell and Gregory's classification, starting from class IA (271 teeth in the LA group and 58 teeth in the GA group) up to class IIIC (2 teeth in the LA group and 15 teeth in the GA group). 8n the remaining study, the preliminary staging before the extractions was not clearly described. 17 lingual retractor was used in 3 cases. 13,14,16In 1 case, the Howarth retractor was used in all proce dures of extraction 14 ; the Howarth retractor was also used in the study by Brann et al. for all the teeth requir ing bone removal and the retraction of lingual tissues 13 ; in another case, the Howarth or Hovell retractors were used based on the choice of the surgeons (in 45 proce dures under LA and in 340 procedures under GA). 16he Prichard elevator was used in 1 study, when neces sary, protecting but not retracting the lingual flap. 8ülicher and Gerlach used a nonspecified periosteal elevator to protect the lingual nerve. 15Nguyen et al. did not specify the use of a lingual retractor or a peri osteal elevator. 15,17he period of followup was 6 months. 8,13-16Nguyen et al. followed the patients for a period of 12 months. 17

Clinical outcomes
Costantinides et al. did not observe any cases of LN injury (temporary or permanent) after LA. 8 Only 1 case (0.29%) of temporary IAN injury was observed after a week, which persisted after the period of followup and developed into permanent injury.In this single case, the extracted tooth showed a canal between the roots, in which IAN was entrapped.Therefore, the nerve le sion could be ascribed to the complications related to the anatomical conformation of the extracted tooth.Nine cases (4.64%) of temporary IAN injury and 4 cases (2.06%) of temporary LN injury were observed after GA.One case (0.52%) of permanent LN injury was noticed following GA. 8rann et al. found 5 cases (3.52%) of temporary nerve injury after LA, 90 cases (16.82%) after GA and 1 case (2.44%) for LA plus sedation, without distinguishing IAN and LN injury. 13Five cases (0.70%) of permanent nerve injury were observed without specifying the an esthetic modality and the nerve involved. 13 the study by Hill et al., 8 cases (3.98%) of temporary nerve injury to IAN and LN related to LA, and 16 cases (3.72%) of temporary nerve injury to IAN and LN related to GA were detected. 14No cases of permanent nerve in jury were found either for LA or for GA after the follow up period. 14ülicher and Gerlach found 8 cases of temporary in jury to LN under LA, 1 case for LA plus sedation and 14 cases under GA. 15Four cases (0.36%) of permanent LN injury and 10 cases (0.90%) of permanent IAN in jury were observed, without specifying the anesthetic modality. 15ehman et al. found 9 cases (8.57%) of postopera tive temporary nerve injury -5 (4.76%) to IAN and 4 (3.81%) to LN -following LA, while 37 cases (7.81%) of temporary nerve injury were observed in the GA group (15 (3.16%) to IAN and 22 (4.64%) to LN). 13 No cases of nerve injury were observed in the group of LA plus sedation.After the followup period, no cases of permanent injury to LN were noticed for any an esthetic modality; conversely, a total of 3 cases (0.49%) of nerve injury to IAN were observed, not specifying the anesthetic choice. 13n their study, performed on 11,599 cases of mandibular third molar removal (6,803 patients), Nguyen et al. found 1 case of temporary LN injury and no cases of permanent LN injury following LA; 4 cases of temporary IAN injury and 8 cases of permanent IAN injury were found for LA. 17 Fifty cases of temporary nerve injury were observed -8 to LN and 42 to IAN -after GA.Six cases of permanent LN injury and 15 cases of permanent IAN injury were de tected after GA. 17

Main reported results and clinical implications
In their conclusions, 3 studies suggested that mandibu lar third molar surgery under GA seemed to increase the risk of nerve injury as compared to surgery under LA. 8,13,15guyen et al. in their conclusions observed an increased risk of permanent IAN injury under GA and concluded that no factor increased the risk of LN injury. 17n 2 studies, no significant relationships were found be tween procedures under LA and GA and nerve injury. 14,16

Risk of bias in individual studies
According to the risk of bias analysis, only 2 stu dies were judged to have a medium risk of bias, with an overall score between 16 and 22, 8,13 and a higher score of 19. 8 The remaining 4 studies were judged to have a high risk of bias, [14][15][16][17] with a lower score of 12. 17 The internal validity -bias items were related to low scores, with the exception of 2 studies. 8,14Full details are summa rized in Table 4.

Discussion
The present review examined the potential development of nerve injury after lower third molar extraction per formed under LA and the same intervention performed under GA.The number of retrieved studies (6 studies) is still limited, since this aspect has been poorly investigat ed.Moreover, the heterogeneous designs and recordings of the included studies did not allow any metaanalysis, while the direct comparisons of the obtained results are not fully applicable.
Four of the 6 studies showed an association between GA and nerve injury. 8,13,15,17Conversely, 2 of the included studies did not show a significant relationship between nerve damage and the anesthetic modality. 14,16Further more, the selected studies showed generally a high risk of bias (except 2 studies, 8,13 which showed a medium risk of bias), limiting the strength of evidence.
][16][17] Brann et al. did not give information about the operator's experience. 13n 2 studies, the operator's experience was the major factor influencing the frequency of IAN and LN injury during lower third molar extraction. 13,15The uncontrolled application of force, low ability in the management of sur gical instruments and the lack of experience could cause nerve injury. 18][15][16] Only 3 studies comparing procedures under LA and GA in terms of surgical difficulty showed no substantial differ ences between the study groups. 8,13,16One of these studies did not show a relationship between nerve damage and the anesthetic modality 16 ; the other 2 studies showed that GA seemed to increase the risk of developing IAN and LN lesions. 8,13ble 4. Risk of bias in the included studies according to the modified Down and Black tool

Item
Costantinides et al. 8 Brann et al. 13 Hill et al. 14 Gülicher and Gerlach 15 Rehman et al. 16 Nguyen et al. 17 1 yes (1) yes ( 1) yes ( 1) yes ( 1) yes ( 1 In the studies by Gülicher and Gerlach 15 and by Nguyen et al., 17 patients were assigned to surgery under GA based on many factors, including the expected high level of surgical difficulty.In 2 studies, there could be observed a trend to direct patients to surgery under GA rather than LA. 13,16In fact, in the UK, dental procedures are frequently performed under GA because of the preference of both surgeons and patients. 5 nonhomogeneous distribution of surgical difficulty inside experimental groups could be the reason of bias in the evaluation of the influence of LA and GA on the development of nerve injury.8 Costantinides et al. sup posed that the anesthetic regimen should not be chosen according to the surgeon's preference, but to the patient's choice or clinical characteristics (e.g., dental phobia).8 Such an attitude results in a better distribution of surgical variables, and consequently enables a more reliable com parison between groups, with a lower risk of bias.
The use of a lingual retractor is another parameter that can produce a bias.In particular, it regards the Howarth retractor, which is a narrow instrument used for lifting and retracting the surgical flap; it offers poor flap protec tion and is capable of exerting considerable force on LN, inducing injury, and the alteration or blockage of nerve conduction. 16,17The Howarth retractor was used in 3 studies. 13,14,169][20] For this reason, many authors suggest avoiding the preparation and retraction of the lingual flap.No cases of permanent LN injury was found when the Howarth retractor was used. 14,16,18,21ehman et al. observed a total of 3 cases of permanent IAN injury, without distinguishing the anesthetic mo dality, while no permanent LN injury was observed. 13ill et al. did not observe any permanent nerve injury in their study. 14Both authors concluded that there was no significant difference between lower third molar surgery performed under LA or GA. 13,14he number of cases of permanent injury after GA is greater in the studies by Gülicher and Gerlach 15 and by Nguyen et al. 17 Such data could be a consequence of the high number of extractions performed (1,106 and 11,599 teeth removed, respectively), which could increase the probability of adverse events. 15,17rann et al. elaborated some theories on a high risk of developing nerve injury after GA and hypothesized that procedures under GA could be complicated by the supine position or by the extent of mucoperiosteal strip ping and bone removal. 13The same author suggests that the degree of surgical force may be greater under GA and that a conscious patient provides a series of signals to the surgeon, who tends to limit tissue retraction and surgical force, which decreases the risk of nerve injury. 13ased on the retrieved data, it is not possible to deter mine a correlation between the anesthetic modality and the nerve injury because of different variables that are related to each procedure of extraction and a high risk of bias across the studies.
The only certain indication described in the litera ture is that the prevalence of complications after dental procedures under GA have induced a reduction in the number of procedures performed with this anesthetic modality. 13,22'Eramo reported a mortality rate of 1:1,733,055 and a frequency of adverse events of 1:26,698 in patients under going GA for oralmaxillofacial surgery. 23The most commonly observed adverse event is laryngospasm, present in 1 out of 833 patients (0.12%) treated under GA.The same author reported with a lower frequency the following: cardiac arrhythmias; bronchospasm; hypertension; hypotension; congestive heart failure; angina pectoris; myocardial infarction; the nerve and/or cervical lesions associated with changes in the patient's position during anesthesia; phlebitis; insulin shock; and diabetic ketoacidosis. 23our of the 6 analyzed studies showed a greater inci dence of neurological damage following the extractions performed under GA.
The quality and the number of studies on the topic dis cussed in the present review is low, and more investiga tions are necessary with betterquality studies.

Conclusions
In third molar surgery, LA should be preferred when possible because of the increased rate of complications under GA.However, GA remains an appropriate an esthetic modality in case of complex and long procedures, uncooperative patients, dental phobic patients, and pa tients with allergy to a local anesthetic.

Registration
The present systematic review was registered in the International Prospective Register of Systematic Reviews under No. PROSPERO CRD42021231823.

Table 2 .
Studies excluded after full-text consideration with the corresponding main reason for exclusion Fig. 1.Flow diagram of the search strategy

Table 3 .
Summarized data of the 6 studies included in the review M -mean; SD -standard deviation; T -temporary; P -permanent; ROS -retrospective observational study; PS -prospective study; PLS -prospective longitudinal study; RS -retrospective study; LA -local anesthesia; GA -general anesthesia; M -male; F -female; WC -Winter's classification; WS -Wharfe's score; P&G -Pell and Gregory's classification; H -horizontal; V -vertical; MA -mesioangular; DA -distoangular; Trans -transversal; O -other; PRE -Prichard elevator; HR -Howarth retractor; PE -periosteal elevator; HOR -Hovell retractor; IAN -inferior alveolar nerve; LN -lingual nerve; a -merging cases of IAN and LN injury; b -merging cases of IAN and LN injury in the LA, GA and LA + sedation groups; c -a total of 39 cases of IAN injury by merging the LA, GA and LA + sedation groups; d -a total of 10 cases of IAN injury by merging the LA, GA and LA + sedation groups; e -a total of 4 cases of LN injury by merging the LA, GA and LA + sedation groups; f -a total of 3 cases of IAN injury by merging the LA, GA and LA + sedation groups; NA -data not available.