Assignment: Riku Case Study
Assignment: Riku Case Study
Riku is a 19-year-old college student. One morning, after a long night of studying, Riku woke up and made himself a hot cup of coffee and toast. Much to his surprise, when he brought the cup to his mouth to drink, the coffee spilt onto the table. Riku went to the bathroom mirror and noticed the left side of his face seemed to droop. He quickly got dressed and ran to the medical clinic on the college campus. As he ran, his left eye began to feel scratchy and dry, but he could not blink in response. The physician at the clinic listened to Riku’s story and then did a careful cranial nerve examination. She concluded that Riku had Bell palsy, an inflammatory condition of the facial nerve most likely caused by a virus.
Student Name:
What are an afferent neuron and efferent neuron? What are efferent components of the facial nerve and their actions?
Under certain circumstances, axons in the peripheral nervous system can regenerate after sustaining damage. Why is axonal regeneration in the central nervous system much less likely?
KINDLY ORDER NOW FOR A CUSTOM-WRITTEN AND PLAGIARISM-FREE PAPER
At a healthy myoneural junction, acetylcholine is responsible for stimulating muscle activity. What mechanisms are in place to prevent the continuous stimulation of a muscle fiber after the neurotransmitter is released from the presynaptic membrane?
Grossman, S. & Porth, C.M. (2013). Porth’s pathophysiology: Concepts of altered health states (9th ed.). Philadelphia, PA: Wolters Kluwer Health/Lippincott Williams & Wilkins. ISBN: 978-1451146004
Use this source and 2 others regarding the case study topic.
Cranial Nerve Injury After Carotid Endarterectomy: Incidence, Risk Factors, and Time Trends
Objective/Background
To review the incidence of post-carotid endarterectomy (CEA) cranial nerve injury (CNI), and to evaluate the risk factors associated with increased CNI risk.
Methods
The study was a meta-analysis. Pooled rates with 95% confidence intervals (CIs) were calculated for CNIs after primary CEA. Odds ratios (ORs) were calculated for potential risk factors. A fixed-effects model or a random effects model (Mantel–Haenszel method) was used for non-heterogeneous and heterogeneous data, respectively. Meta-regression analysis was performed to examine the influence of publication year upon CNI rate.
Results
Twenty-six articles, published between 1970 and 2015, were included in the meta-analysis, corresponding to 20,860 CEAs. Meta-analysis revealed that the vagus nerve was the most frequently injured cranial nerve (pooled injury rate 3.99%, 95% CI 2.56–5.70), followed by the hypoglossal nerve (3.79%, 95% CI 2.73–4.99). Fewer than one seventh of these injuries are permanent (vagus nerve: 0.57% [95% CI 0.19–1.10]; hypoglossal nerve: 0.15% [95% CI 0.01–0.39]). A statistically significant influence of publication year on the vagus and hypoglossal nerve injury rate was found, with the injury rate having decreased from about 8% to 2% and 1%, respectively, over the last 35 years. Urgent procedures (OR 1.59, 95% CI 1.21–2.10; p = .001), as well as return to the operating room for a neurological event or bleeding (OR 2.21, 95% CI 1.35–3.61; p = .002) were associated with an increased risk of CNI, whereas no statistically significant association was found between CNIs and the type of anaesthesia, the use of a patch, redo operation, and the use of a shunt.
Conclusion
The vagus nerve appears to be the most frequently injured cranial nerve after CEA, followed by the hypoglossal nerve, with only a small proportion of these injuries being permanent. The CNI rate has significantly decreased over the past 35 years to a point indicating that CNIs should not be considered a major influencing factor in the decision making process between CEA and stenting.
Keywords
What this paper adds
This meta-analysis summarises the available data on cranial nerve injuries from four randomised, eight prospective, and 14 retrospective studies, corresponding to 20,860 carotid endarterectomies. The results of the meta-analysis indicate that the incidence of cranial nerve injuries has significantly decreased from about 8% to <2% over the past 35 years. Fewer than one seventh of these injuries are permanent. These findings put the problem of post-carotid endarterectomy cranial nerve injury into a new perspective. Cranial nerve injuries should not be considered as a major influencing factor in the decision making process between carotid endarterectomy and stenting.
Introduction
Carotid endarterectomy (CEA) is associated with a number of serious complications, with stroke, death, and myocardial infarction being the main endpoints of all relevant studies. Cranial nerve injuries (CNIs) have received considerably less attention, despite the fact that they are quite frequent and potentially serious; they may even be life threatening when they are bilateral. Although it has been more than 40 years since CNIs after CEA were first described, several questions remain unanswered regarding the incidence, predictors, and the management of such injuries.1, 2 The reported incidence of CNIs after CEA ranges widely from 2% to >50%, depending mainly on the different investigative methods used for the evaluation of cranial nerve function.3, 4 Most of these nerve injuries are transient, being due to neuropraxia caused by excessive retraction. Thus, prevention is better than cure, especially as there is no specific and effective therapy.
CNIs have gained renewed interest over the last 15 years as they have become a point of comparison between CEA and carotid stenting (CAS) and have been defined as a secondary outcome in most recent CEA versus CAS trials.5, 6, 7, 8 Moreover, there has been a continuing claim by interventionalists supporting CAS that CNIs should be included in the composite endpoint of trials comparing CEA with CAS, as their clinical impact is similar to a minor stroke. This claim, however, has been challenged, as the incidence of a permanent or disabling CNI is very low and should not detract from the significant benefit conferred by CEA regarding stroke prevention.9
The aim of this study was to review the incidence of post-CEA CNI, to evaluate the risk factors associated with increased CNI risk, and to examine whether the incidence of CNI has changed over the past few decades.
Materials and Methods
Data collection
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used in the current study.10 The following medical literature databases were searched systematically: MEDLINE, Scopus, Embase, Google Scholar, Ovid, and the Cochrane Library. After retrieving the relevant articles, a snowball process in the reference list of eligible articles was followed to retrieve additional articles, which were, thereafter, included if they satisfied the inclusion criteria.
Search methodology, inclusion and exclusion criteria, data extraction
The following Medical Subject Headings (MeSH) terms were used: (“cranial” [All Fields] OR “brain” [All Fields]) AND (“endarterectomy” [All Fields]) AND (“nerve” [All Fields]) AND (“carotid” [All Fields]). All scientific papers between January 1970 and December 2015 were searched for, without sex or language restriction. Studies focusing on CNI after primary CEA were investigated. Studies reporting on CNIs after repeat CEA were excluded. Studies that did not report numbers of CNI in patients after CEA were also excluded. Data were independently extracted and analysed by two authors (J.D.K., C.N.A.) and the final decision was reached by consensus. Data extracted from eligible studies included the first author’s name, study year, country in which the study was conducted, total number of patients, total number of CEAs, number of CNIs, study period, type of study, male sex (%), mean age (years), description of otorhinolaryngologist’s examination, definition of neurological assessment, type of CEA technique, shunt use (%), and risk factors for CNI, as reported by the authors. Numbers for CNIs of facial (VII), glossopharyngeal (IX), vagus (X), spinal accessory (XI), hypoglossal (XII), and important nerve branches, namely marginal mandibular and great auricular, were extracted separately. The total number of CNIs (either transient or permanent) for each nerve, as well as the number of permanent CNIs, was also extracted separately.
Statistical analyses
Data synthesis and treatment effects
The CNI rates in patients after CEA were estimated for each cranial nerve studied and reported as the proportion of corresponding nerve injuries among all CEAs. Values of the concomitant injuries were subsequently appropriately calculated, expressed as proportions and 95% confidence intervals (CIs) and thereafter transformed into quantities according to the Freeman–Tukey variant of the arcsine square root transformed proportion. The pooled effect estimates were calculated as the back transformation of the weighted mean of the transformed proportions, using DerSimonian–Laird weights of random effects model and expressed as proportions (%).11
A second meta-analysis was performed, aiming to investigate the potential role of clinical risk factors found in the eligible studies, namely urgent CEA, local anaesthesia, use of patch, redo operation, return to the operating room for a neurological event or bleeding, and use of shunt on CNI during CEA, and was based on the available data extracted of the total number of patients and patients with and without the risk factor from each study. Thereafter, odds ratios (ORs) with corresponding CIs were appropriately calculated from 2 × 2 tables or directly extracted from the results of multivariate analyses reported in the eligible studies. A pooled estimate of ORs, together with the corresponding 95% CIs, was then calculated. A fixed effects or a random effects model (Mantel–Haenszel method) was used for non-heterogeneous or heterogeneous data, respectively. An OR > 1 favoured the positive effect of each risk factor on the risk for CNI.
Heterogeneity, publication bias, and meta-regression analysis
A formal statistical test for heterogeneity using the I2 test was performed.12 Publication bias was assessed using the Egger’s test for small study effects, as well as visual inspection of funnel plots.13, 14 Random effects meta-regression analysis was performed using aggregate level data to examine the influence of publication year on the incidence of nerve injuries. Analyses were conducted using STATA statistical software version 14 (Stata Corp., College Station, TX, USA) and Comprehensive Meta-Analysis (version 3.0, Biostat, Englewood, NJ, USA).
Validity assessment
Risk of bias in the eligible studies was assessed by each reviewer using the Newcastle–Ottawa Quality Assessment Scale for cohort studies.15 The Newcastle–Ottawa Scale consists of eight items grouped into three sections that are relevant to the quality of cohort studies. A study can be awarded a maximum of one star for each numbered item within the “Selection” and “Outcome” categories. A maximum of two stars can be given for “Comparability”. Validity scores were evaluated as follows: ≤ 5, low quality; 6–7, medium quality; 8–9, high quality. The Cohen kappa coefficient was used to assess agreement between the two investigators with respect to the outcomes of interest.
Results
Study characteristics
After the literature search, 322 potentially eligible studies were identified. Review of the abstracts showed that 63 were not relevant. One article was also removed because it referred to CNIs after repeat CEA.16 A total of 259 articles were further evaluated. Among these, five were excluded as case reports (<5 cases).1, 2, 17, 18, 19 Finally, 26 articles were eventually deemed relevant to be included in the meta-analysis (Fig. 1), corresponding to a total of 20,860 CEAs.3, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44
Figure 1. Flowchart presenting the selection of eligible studies.
Meta-analysis
Table 1 presents the baseline study characteristics of the 26 eligible studies. The pooled nerve injury rate for the facial nerve (VII) was 1.97% (95% CI 1.37–2.66; Fig. 2), of which 11.68% were permanent (0.23%, 95% CI 0.05–0.50). When the marginal mandibular branch of the facial nerve was analysed separately, a pooled nerve injury rate of 1.58% (95% CI 0.82–2.54) was recorded. The glossopharyngeal nerve (IX) injury rate was 0.22% (95% CI 0.11–0.36). Of these injuries, 4.35% were permanent. A pooled injury rate of 3.99% (95% CI 2.56–5.70; Fig. 3) was recorded for the vagus nerve (X). A total of 14.30% of all vagus nerve injuries were permanent (0.57%, 95% CI 0.19–1.10). The pooled injury rate after CEA for the spinal accessory (XI) nerve was 0.21% (95% CI 0.08–0.39). The hypoglossal nerve (XII) injury rate was 3.79% (95% CI 2.73–4.99; Fig. 4). Of these, 3.96% were permanent (0.15%, 95% CI 0.01–0.39). For the great auricular nerve, there was a 12.71% (95% CI 1.56–31.42) injury rate. Table 2 presents a comprehensive summary of the results of the meta-analysis, with pooled CNI rates, 95% CIs, and results of heterogeneity and publication bias assessment.
Table 1. Baseline characteristics of eligible studies.
| Study | Country | CEAs (n) | Study period | Patients (n) | Type of study | Male sex (%) | Mean age (yr) | ENT examination | Neurological assessment | Endarte
rectomy technique |
Shunt use (%) | Risk factors for CNI reported by the authors |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bennett et al. (2015)20 | USA | NR | 2012 | 3762 | Retrospective | 61 | NR | NR | NR | Patch (78%) Eversion (6.5%) Re-operation (2.4%) |
33 | Independent predictors of CNI included age ≥ 80 y (reference group < 70 years; OR 1.74, 95% CI 1.00–3.03), presence of a pre-operative bleeding disorder (including patients in whom pre-operative non-aspirin anticoagulation therapy was not stopped before CEA (OR 1.66, 95% CI 1.03–2.68), duration of operation (OR 1.15 for each 30 min interval beyond an operative time of 90 min; 95% (CI 1.06–1.25), and need for re-operation (OR 2.65; 95% CI 1.03–6.80). |
| Hye et al. (2015)21 | Canada | NR | December 2000–July 2008 | 1151 | RCT | 67 | 69 | NR | Adjudication of the CNIs was performed by two neurologists and a vascular surgeon (pre-operative and 1 and 6 mo post-procedure) | Patch (70%) | 70 | Unable to identify any pre-operative characteristics or intra-operative variables except from operation under general anaesthesia that predicted the occurrence of CNI in CREST |
| Thirumala et al. (2015)8 | USA | NR | 2007–2012 | 587 | Retrospective | 62 | 77 | NR | Neurological status of the cranial nerves before and after CEA and comorbid conditions were obtained from medical records | NR | IOM using EEG and SSEPs has been used as an aid for the use of shunting during CEA to evaluate cerebral perfusion | The data showed that nerve injury was more frequent in operations performed with a shunt, with patch closure, and by a junior surgeon |
| Doig et al. (2014)23 | USA, UK, Netherlands | NR | NR | 821 | RCT | NR | NR | NR | All patients were then reassessed at 1 mo after the procedure by a neurologist or investigator under their supervision | The type of arterial reconstruction to be carried out was not specified in the protocol | NR | Independent risk factors modifying the risk of CNP were cardiac failure (RR 2.66, 95% CI 1.11–6.40), female sex (RR 1.80, 95% CI 1.02–3.20), the degree of contralateral carotid stenosis, and time from randomisation to treatment > 14 days (RR 3.33, 95% CI 1.05–10.57) |
| Fokkema et al. (2014)22 | USA | NR | January 2003–December 2011 | 6878 | Prospective | 60 | NR | Objective tests such as laryngoscopy for vocal cord function were not used routinely and their use was not recorded | Persistent CNI was identified by the vascular surgeon and defined as a CNI at discharge that was not resolved at the time of the surgical follow-up visit | Standard (91%) Eversion (9%) | 47 | CNI is more likely in urgent procedures and after re-exploration in the operating room, or return to the operating room, while redo CEA and a history of prior cervical radiation were not associated with increased CNI rate |
| Regina et al. (2009)24 | Italy | NR | March 1999–April 2006 | 1126 | RCT | 55 | 62–76 | Three days after the operation a neurologist and an otorhinolaryngologist evaluated the presence of cranial nerve deficits by clinical examination and videolaryngoscopy | Eversion CEA | Selective shunting | Most CNIs are caused by traction, compression, electrocautery, inadvertent clamping, ligatures, or partial to total transections. Pre- and post-operative administration of dexamethasone is effective in decreasing the incidence of temporary post-CEA cranial nerve dysfunction | |
| Beasley et al. (2008)25 | UK | 236 | 16 year period | 236 | Retrospective | 70 | 43–85 | All suspected laryngeal nerve injuries were investigated using indirect laryngoscopy and all clinically detected cranial nerve injuries had targeted follow-up until resolved | NR | Eversion CEA with anteromedial and retro-jugular approach | Shunting was employed selectively according to the development of intra-operative neurological deficit | Most CNIs probably result from stretching, retraction, or clamping, but they may also arise from the injudicious use of diathermy or ligation for haemostasis—the retro-jugular technique does not appear to increase the risk of CNI during CEA |
| Assadian et al. (2004)26 | Austria | 180 | April 2002–September 2002 | 165 | Prospective | 93 | 71 | The vocal cords were assessed by fibreoptic laryngoscopy before and 4–5 days after CEA. In case of cranial nerve injury or hoarseness, follow-up continued until the symptoms had resolved | Pre- and post-operative neurological examination | NR | NR | Post-operative hoarseness is most frequently due to laryngeal haematoma |
| Maroulis et al. (2000)27 | UK | NR | Jan 1994–December 1997 | 269 | Retrospective | 62 | 63 | Not routinely performed | NR | Primary closure (71%) Patch (29%) |
68 | Most CNIs are transient and result from trauma during dissection, retraction, or carotid clamping |
| Ferguson et al. (1999)28 | Canada | NR | NR | 1415 | RCT | 71 | 65 | NR | External adjudication was conducted by a panel of neurologists and surgeons not otherwise involved in the trial | Primary closure (79%) Patch (10%) |
41 | NR |
| Ballotta et al. (1999)29 | Italy | 200 | 3 year period | 187 | Prospective | 73 | 70 | Pre- and post-operative before the operation and 1 wk after the operation; it was repeated 6 wk and 3 mo later on patients complaining of post-operative voice changes | Pre- and post-operative evaluations and repeated 1 and 6 wk after operation and, in the event of neurological deficits, the tests were repeated 3 mo later | Patch plasty (50%) Eversion CEA (50%) |
17 | Most peripheral nerve damage is caused by retraction, stretching, or clamping |
| Zannetti et al. (1998)30 | Italy | 190 | Jan 1994–April 1995 | 187 | Prospective | 78 | 68 | Pre- and post-operative cranial and cervical nerve assessments were carried out by a single otolaryngologist, blinded to the operative technique and findings | NR | Standard (32%) Patch (11%) Eversion (68%) | 15 | Vagus nerve lesions were significantly associated with long (>2 cm) carotid plaque (OR 3.5, 95% CI 1.09–12.37; p = .03) Cervical branch lesions were associated with the presence of neck haematoma (OR 1.9; 95% CI 0.7–4.7; p = .05) The incidence of single cranial nerve injuries was higher in patch (OR 2.7) and eversion (OR 1.9) procedures than in primary closure Multiple deficits (≥2) were most frequent in eversion CEA (OR 2.8) and in cases complicated by neck haematoma (OR 3.8) |
| Schauber et al. (1997)31 | USA | 183 | April 1990–February 1995 | 155 | Prospective | 98 | 66 | Pre- and post-operative cranial nerve assessment, including fibreoptic laryngoscopy, was performed | NR | NR | NR | Cranial and cervical nerve dysfunction after CEA is usually caused by direct trauma to the specific nerve by stretch, retraction, clamping, or transection |
| Forsell et al. (1995)32 | Sweden | 656 | 1982–1992 | 656 | Retrospective | 72 | 66 | Pre- and post-operative examinations were made the day before surgery and before discharge from the hospital or within the first post-operative month. The patients were examined with voice recordings and stroboscopic light examination of the vocal cords, as well as a clinical examination of the nerves | Standard CEA (97%) | 32 | Nerve injury was more frequent in operations performed with a shunt (p = .05), with patch closure (p = .01), and by a junior surgeon (p = .05) | |
| Maniglia and Han (1991)33 | USA | 336 | Jan 1984–December 1987 | 295 | Retrospective | 78 | 66 | Not routinely performed except for persistent post-operative cranial nerve abnormalities or after bilateral CEA | NR | NR | NR | Most injuries were due to either retraction or oedema of cranial nerves |
| Rogers and Root (1988)34 | USA | 433 | 1979–1985 | 355 | Retrospective | NR | NR | NR | NR | NR | A shunt was used in all but a few cases in which technical problems prevented its use | Complex or prolonged procedures |
| Aldoori and Baird (1988)35 | UK | 52 | NR | 51 | Prospective | 80.4 | 61 | Indirect laryngoscope pre- and post-operatively when indicated during the first 2 weeks | Pre-operatively/post-operatively/6 weeks/3 monthly intervals up to 42 months | Classical/ longitudinal incision in 71% Transverse incision in 29% |
NR | Self retaining retractor, diathermy burn in the vicinity of nerve, inclusion of the nerve in the arterial clamp |
| Knight et al. (1987)38 | USA | 129 | 1974–84 | 112 | Retrospective | 70 | 61 | None of the patients underwent routine pre- or post-operative laryngoscopy unless they were symptomatic or bilateral operations were planned | NR | NR | 24 | Most injuries are transient and result not from transection but from trauma during dissection, retraction, and clamping of the vessels |
| Weiss et al. (1987)36 | Czech Republic | 536 | 1973–83 | 484 | Retrospective | NR | NR | Post-operative laryngoscopy was performed in the event of post-operative phonation disorder | Thorough neurological evaluation before surgery, post-operatively, before discharge and every 6 mo | NR | NR | Intra-operative trauma, improper surgical technique |
| Tucker et al. (1987)37 | USA | 850 | 1978–86 | NR | Retrospective | NR | NR | NR | NR | NR | Shunts were used on the last two patients only | The authors proposed surgical scar formation as a mechanism of accessory nerve palsy after CEA |
| Theodotou and Mahaley (1985)39 | USA | 192 | 1977–83 | 162 | Retrospective | 65 | 63 | NR | NR | NR | 49 | Poor knowledge of local anatomy and improper surgical technique |
| Massey et al. (1984)40 | USA | nr | 1974–78 | 158 | Prospective | NR | 61 | Computerised data bank of patients with symptoms of transient neurological ischaemia | NR | NR | Most of the deficits were transient, which indicates that they probably arise from a reversible injury following stretch, retraction, or clamping of nerves rather than permanent division of these structures | |
| Schmidt et al. (1983)41 | Germany | 109 | 1968–82 | 102 | Retrospective | 65 | 34–76 | Post-operative laryngoscopy was performed only when hoarseness was noted | Hypoglossal and facial nerve dysfunction was ascertained from the pre- and post-operative examinations by a consultant neurologist | NR | NR | The nerves are injured by retraction to clear the operative field or by post-operative haematoma. Risk factors include the hypoglossal nerve crossing close to the carotid bifurcation or procedures requiring long arteriotomy or skeletonisation of the internal carotid artery |
| Dehn and Taylor (1983)42 | UK | 43 | NR | 40 | Prospective | 78 | 53 | Vocal cord movements were assessed by indirect laryngoscopy. This latter examination was undertaken on all patients 1 wk after surgery and was repeated at 6 wk and 6 mo on those patients who had a documented cord palsy and on those who complained of post-operative voice changes | Pre-operative, 1 and 6 wk post-operatively and, in the event of persistent neurological deficit, at 6 mo after surgery | NR | Use of an intraluminal shunt is left to the individual surgeon | The use of a transverse skin incision may increase the likelihood of nerve injury—relevant when bilateral CEA is being considered |
| Evans et al. (1982)43 | USA | 128 | NR | 116 | Prospective | 46 | 34–87 | A speech pathologist made a subjective judgment regarding the presence or absence of pathology | NR | NR | NR | NR |
| Hertzer et al. (1980)44 | USA | 240 | April 1978–July 1979 | 240 | Prospective | NR | NR | 123 patients were examined pre-operatively and all 240 post-operatively by ORL | NR | Classical with transverse incision | Routinely used | Intra-operative retraction |
Note. NR = not reported; RCT = randomised controlled trial; CNI = cranial nerve injury; OR = odds ratio; CI = confidence interval; CEA = carotid endarterectomy; CREST = Carotid Revascularisation Endarterectomy versus Stenting Trial; IOM = intra-operative neuro-physiological monitoring; EEG = electroencephalography; SSEPs = somatosensory evoked potentials; CNP = cranial nerve palsy; RR = relative risk; ENT = ear, nose, throat; ORL = otorhinolaryngologist. Assignment: Riku Case Study
Figure 2. Forest plot presenting the meta-analysis and pooled nerve injury rate with 95% confidence intervals (CIs) for the facial nerve (VII). Proportions in the individual studies are presented as squares, with 95% CIs presented as extending lines. The pooled proportion with its 95% CI is depicted as a diamond.
Figure 3. Forest plot presenting the meta-analysis and pooled nerve injury rate with 95% confidence intervals (CIs) for the vagus nerve (X). Proportions in the individual studies are presented as squares, with 95% CIs presented as extending lines. The pooled proportion with its 95% CI is depicted as a diamond.
Figure 4. Forest plot presenting the meta-analysis and pooled nerve injury rate with 95% confidence intervals (CIs) for the hypoglossal nerve (XII). Proportions in the individual studies are presented as squares with 95% CIs presented as extending lines. The pooled proportion with its 95% CI is depicted as a diamond.
Table 2. Pooled cranial nerve injury (CNI) rates with corresponding results of heterogeneity and publication bias. Assignment: Riku Case Study
| Nerve | Pooled CNI rates, % (95% CIs) | Heterogeneity | Publication bias (Egger’s test) | ||
|---|---|---|---|---|---|
| I2 | p | Tau | p | ||
| Facial (VII) | 1.97 (1.37–2.66) | 82.4% | <.001 | 1.10 | .290 |
| Marginal mandibular branch of facial nerve | 1.58 (0.82–2.54) | 80.3% | <.001 | 0.81 | .430 |
| Glossopharyngeal (IX) | 0.22 (0.11–0.36) | 0% | .800 | 0.07 | .970 |
| Vagus (X) | 3.99 (2.56–5.70) | 93.9% | <.001 | 3.17 | .005 |
| Spinal accessory (XI) | 0.21 (0.08–0.39) | 0% | .990 | 0.88 | .420 |
| Hypoglossal (XII) | 3.79 (2.73–4.99) | 90.1% | <.001 | 1.91 | .070 |
| Great auricular | 12.71 (1.56–31.42) | 97.7% | <.001 | 3.35 | .040 |
Seven studies tried to identify predictors of CNI after CEA.3, 19, 20, 21, 22, 29, 31 Meta-analysis revealed that urgent procedures (OR 1.59, 95% CI 1.21–2.10; p = .001 [Fig. S1; see Supplementary Material]), as well as return to the operating room for a neurological event or bleeding (OR 2.21, 95% CI 1.35–3.61; p = .002 [Fig. 2; see Supplementary Material]) were associated with an increased risk of CNI. On the contrary, no statistically significant association was found between CNIs and the type of anaesthesia (OR 0.90, 95% CI 0.67–1.20; p = .460), the use of a patch (OR 1.37, 95% CI 0.84–2.23; p = .110), redo operation (OR 1.20, 95% CI 0.77–1.88; p = .41), or the use of a shunt (OR 0.99, 95% CI 0.84–1.18; p = .950).
Meta-regression analysis
There was a statistically significant influence of publication year on vagus nerve injury rate (coefficient −0.19%, 95% CI –0.35 to −0.03; t −2.45; p = .020 [Fig. 5]). Similarly, a significant result was recorded for the hypoglossal nerve (coefficient −0.18%, 95% CI –0.34 to −0.02; t −2.26; p = .030 [Fig. 6]).
Figure 5. Meta-regression line (coefficient = −0.19%, p = .02) of reported vagus (X) nerve injury rates versus publication year. The regression line (meta-regression slope) is presented as a straight line. Each study is represented as a circle and the circle’s size reflects the sample size. Assignment: Riku Case Study
Figure 6. Meta-regression line (coefficient = −0.18%, p = .03) of reported hypoglossal (XII) nerve injury rates versus publication year. The regression line (meta-regression slope) is presented as a straight line. Each study is represented as a circle and the circle’s size reflects the sample size.
Validity assessment
The Newcastle Ottawa Scale for assessing the quality of the eligible studies revealed that 15 studies were of high quality and that 10 studies were of medium quality (Table 3).
Table 3. Newcastle–Ottawa quality assessment scale for cohort studies included in this review.
| Category | Newcastle–Ottawa quality assessment scale for cohort studies | Total | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Selection | Comparability | Exposure | |||||||
| Items | Representativeness of the exposed cohort | Selection of the non-exposed cohort | Ascertainment of exposure | Demonstration that outcome of interest was not present at start of study | Comparability of cohorts on the basis of the design or analysis | Assessment of outcome | Was follow-up long enough for outcomes to occur? | Adequacy of follow-up of cohorts | |
| Options | (A) Truly representative of the average patient with carotid stenosis in the community* | (A) Drawn from the same community as the exposed cohort* | (A) Secure record (e.g., surgical records)* | (A) Yes* | (A) Study controls for non-cranial nerve injury use* | (A) Independent blind assessment* | (A) Yes* | (A) Complete follow-up; all subjects accounted for* | |
| (B) Somewhat representative of the average patient with carotid stenosis in the community* | (B) Drawn from a different source | (B) Structured interview* | (B) No | (B) Study controls for any additional factor* | (B) Record linkage* | (B) No | (B) Subjects lost to follow-up unlikely to introduce bias—small number lost* | ||
| (C) Selected group of users (e.g., nurses, volunteers) | (C) No description of the derivation of the non-exposed cohort | (C) Written self report | – | – | (C) Self report | – | (C) Inadequate follow-up rate and no description of those lost | ||
| (D) No description of the derivation of the cohort | – | (D) No description | – | – | (D) No description | – | (D) No statement | ||
| Score | |||||||||
| Study | |||||||||
| Hye et al. (2015)21 | * | * | * | * | ** | * | * | * | 9 |
| Thirumala et al. (2015)8 | * | * | * | * | * | * | * | – | 7 |
| Doig et al. (2014)23 | * | * | * | * | ** | * | * | * | 9 |
| Fokkemma et al. (2014)22 | * | * | * | * | * | * | * | * | 8 |
| Regina et al. (2009)24 | * | * | * | * | ** | * | * | * | 9 |
| Beasley et al. (2008)25 | * | * | * | * | * | * | * | 7 | |
| Assadian et al. (2004)26 | * | * | * | * | * | * | * | * | 8 |
| Maroulis et al. (2000)27 | * | * | * | * | * | * | * | – | 7 |
| Ferguson et al. (1999)28 | * | * | * | * | ** | * | * | * | 9 |
| Ballotta et al. (1999)29 | * | * | * | * | * | * | * | * | 8 |
| Items | Representativeness of the exposed cohort | Selection of the non-exposed cohort | Ascertainment of exposure | Demonstration that outcome of interest was not present at start of study | Comparability of cohorts on the basis of the design or analysis | Assessment of outcome | Was follow-up long enough for outcomes to occur? | Adequacy of follow-up of cohorts | |
| Options | (A) Truly representative of the average patient with carotid stenosis in the community* | (A) Drawn from the same community as the exposed cohort* | (A) Secure record (e.g., surgical records)* | (A) Yes* | (A) Study controls for non-cranial nerve injury use* | (A) Independent blind assessment* | (A) Yes* | (A) Complete follow-up all subjects accounted for* | |
| (B) Somewhat representative of the average patient with carotid stenosis in the community* | (B) Drawn from a different source | (B) Structured interview* | (B) No | (B) Study controls for any additional factor* | (B) Record linkage* | (B) No | (B) Subjects lost to follow-up unlikely to introduce bias—small number lost* | ||
| (C) Selected group of users (e.g., nurses, volunteers) | (C) No description of the derivation of the non-exposed cohort | (C) Written self report | – | – | (C) Self report | – | (C) Inadequate follow-up rate and no description of those lost | ||
| (D) No description of the derivation of the cohort | – | (D) No description | – | – | (D) No description | – | (D) No statement | ||
| Score | |||||||||
| Study | |||||||||
| Zannetti et al. (1998)30 | * | * | * | * | * | * | * | * | 8 |
| Schauber et al. (1997)31 | * | * | * | * | * | * | * | * | 8 |
| Forssell et al. (1995)32 | * | * | * | * | * | * | * | – | 7 |
| Maniglia and Han (1991)33 | * | * | * | * | * | * | * | – | 7 |
| Rogers and Root (1988)34 | * | * | * | * | * | * | * | * | 8 |
| Aldoori and Baird (1988)35 | * | * | * | * | * | * | * | * | 8 |
| Knight et al. (1987)38 | * | * | * | * | * | * | * | – | 7 |
| Weiss et al. (1987)36 | * | * | * | * | * | * | * | – | 7 |
| Tucker et al. (1987)37 | * | * | * | * | * | * | * | – | 7 |
| Theodotou and Mahaley (1985)39 | * | * | * | * | * | * | * | – | 7 |
| Massey et al. (1984)40 | * | * | * | * | * | * | * | * | 8 |
| Items | Representativeness of the exposed cohort | Selection of the non-exposed cohort | Ascertainment of exposure | Demonstration that outcome of interest was not present at start of study | Comparability of cohorts on the basis of the design or analysis | Assessment of outcome | Was follow-up long enough for outcomes to occur? | Adequacy of follow-up of cohorts | |
| Options | (A) Truly representative of the average patient with carotid stenosis in the community* | (A) Drawn from the same community as the exposed cohort* | (A) Secure record (e.g., surgical records)* | (A) Yes* | (A) Study controls for non-cranial nerve injury use* | (A) Independent blind assessment* | (A) Yes* | (A) Complete follow-up; all subjects accounted for* | |
| (B) Somewhat representative of the average patient with carotid stenosis in the community* | (B) Drawn from a different source | (B) Structured interview* | (B) No | (B) Study controls for any additional factor* | (B) Record linkage* | (b) No | (B) Subjects lost to follow-up unlikely to introduce bias—small number lost* | ||
| (C) Selected group of users (e.g., nurses, volunteers) | (C) No description of the derivation of the non-exposed cohort | (C) Written self report | – | – | (C) Self report | – | (C) Inadequate follow-up rate and no description of those lost | ||
| (D) No description of the derivation of the cohort | – | (D) No description | – | – | (D) No description | – | (D) No statement | ||
| Score | |||||||||
| Study | |||||||||
| Schmidt et al. (1983)41 | * | * | * | * | * | * | * | – | 7 |
| Dehn and Taylor (1983)42 | * | * | * | * | * | * | * | * | 8 |
| Evans et al. (1982)43 | * | * | * | * | * | * | * | * | 8 |
| Hertzer et al. (1980)44 | * | * | * | * | * | * | * | * | 8 |
Note. A study can be awarded a maximum of one star for each numbered item within the “Selection” and “Exposure” categories. A maximum of two stars can be given for “Comparability”.
Discussion
The vagus nerve appears to be the most frequently injured cranial nerve during CEA, with an incidence of 3.99% and a permanent injury rate of 0.57%. The hypoglossal nerve follows, with a total injury rate of 3.79% and a permanent injury rate of 0.15%. The marginal mandibular nerve is injured less frequently (1.58%), whereas injuries of the glossopharyngeal and the spinal accessory nerves are quite rare (0.22% and 0.21%, respectively). The current meta-analysis is focused mostly on transient CNIs because analysis of permanent CNIs is limited by the fact that, in most of the studies, there is no definition of “permanency”. In only three of the studies were permanent CNIs clearly defined as deficits lasting >12 months from the operation,24, 30, 36 whereas in another study permanent injuries were defined as impairments lasting >6 months.32 However, there have been reports of patients with CNIs regaining full nerve function after 3 or even 4 years.29, 31 Consequently, the exact time point at which nerve function recovery occurred or at least the follow-up period should be available if a valid meta-analysis of permanent CNIs is to be performed.Assignment: Riku Case Study
Given these limitations in the analysis of permanent CNIs, it is interesting that the rate of permanent nerve injury among the total number of CNIs (the likelihood that the nerve will not recover) varies widely: 4.35% for the glossopharyngeal nerve, 3.96% for the hypoglossal, 11.68% for the facial, and 14.3% for the vagus nerve. This wide difference could represent a difference in the mechanism of injury. In the vast majority of cases, CNIs are caused by blunt trauma, due to excessive retraction, whereas less common causes include injuries by forceps, electrocautery, or the application of arterial clamps. Though different mechanisms of nerve injury will inevitably have different rates of nerve recovery, no firm conclusion based on the current meta-analysis can be reached.
An interesting finding of this meta-analysis is that the incidence of hypoglossal nerve injury has significantly decreased from about 8% to 2% over the last 35 years, with a mean reduction rate of 0.18% per year. Similarly, the incidence of vagus nerve injury has decreased from about 8% in 1980 to <1% nowadays, with a mean reduction rate of 0.19% per year. Although the exact reason for this decreased incidence cannot be revealed by statistical means in this meta-analysis, it is probably due to increased awareness of this kind of injury and the preventive measures taken against them.
Several studies have identified various predictors of CNI after CEA, including age ≥80 years,20 presence of a pre-operative bleeding disorder,20 duration of operation,20 need for re-operation,20, 22 general anaesthesia,21 urgent procedures,22 cardiac failure,23 female sex,23 the degree of contralateral carotid stenosis,23 long (>2 cm) carotid plaques,30 the use of a patch,30, 32 the use of a shunt,32 eversion endarterectomy,30 cases complicated by neck haematoma,30 operations performed by a junior surgeon,32 and intra-operative neurophysiological monitoring changes.3 Meta-analysis revealed that only urgent procedures and return to the operating room for a neurological event or bleeding are associated with a statistically significant increase in the risk of CNI after CEA, whereas the use of a shunt or a patch, the type of anaesthesia, and redo operations did not prove to be of prognostic significance.
Apart from the obvious strategies for CNI prevention focusing on good knowledge of the anatomy, sharp dissection close the arterial wall, and strict adherence to some general surgical rules, including careful use of forceps, retractors, cautery, and arterial clamps, peri-operative administration of dexamethasone has been found to reduce the incidence of temporary CNIs during CEA, without reducing the prevalence of permanent CNIs.24 However, there is a remarkable paucity of data concerning the treatment of CNIs, once they have happened.
Several limitations of the meta-analysis should be acknowledged. First, among the 26 studies included in the meta-analysis, there were only four randomised trials. In one, CNI was the primary endpoint,24 whereas in the other three studies, which were part of the CREST, the ICSS, and the NASCET trials, respectively, CNI was a pre-specified secondary outcome measure.21, 23, 28 Among the rest of the 22 studies, eight were prospective and 14 retrospective.Assignment: Riku Case Study
Second, and more importantly, there was an inconsistency in the investigative methods used for the evaluation of the cranial nerve function. Routine pre- and post-operative laryngoscopy was performed in only seven studies.24, 26, 29, 30, 31, 32, 42 In most studies, laryngoscopy was performed only in the presence of hoarseness. In the same context, routine pre- and post-operative assessment of cranial nerve function by a neurologist was reported in only 10 studies.21, 23, 24, 26, 29, 32, 35, 36, 41, 42 Consequently, the reported rate of CNIs may represent an underestimate of the actual incidence, as vascular surgeons may have overlooked some minor defects.
Third, the Egger’s test suggested evidence of publication bias for the vagus, recurrent laryngeal, and great auricular nerves. This might indicate that some studies reporting higher rates for the respective CNI injuries may be missing.
In conclusion, the vagus nerve appears to be the most frequently injured cranial nerve after CEA, followed by the hypoglossal nerve. Most of these injuries are transient, recovering within 6–12 months, with the recovery rate being highest in the glossopharyngeal nerve and lowest in the vagus nerve. The CNI rate has significantly decreased over the past 35 years, with an absolute risk reduction rate of about 0.2% per year. Urgent procedures and return to the operating room for a neurological event or bleeding are associated with an increased risk of CNI. The very low permanent CNI risk puts the problem of post-CEA CNI into a new perspective, indicating that CNIs should not be considered as a major influencing factor in the decision making process between CEA and stenting.
Conflict of Interest
None.
Funding
None.
Appendix A. Supplementary data
The following are the supplementary data related to this article:
Fig. S1. Forest plot presenting the meta-analysis based on odds ratios (ORs) for the association between urgent procedures and risk for cranial nerve injury. ORs in the individual studies are presented as squares with 95% confidence intervals (CIs) presented as extending lines. The pooled OR with its 95% CI is depicted as a diamond.
Fig. S2. Forest plot presenting the meta-analysis based on odds ratios (ORs) for the association between return to the operating room for a neurological event or bleeding and risk for cranial nerve injury. ORs in the individual studies are presented as squares with 95% confidence intervals (CIs) presented as extending lines. The pooled OR with its 95% CI is depicted as a diamond. Assignment: Riku Case Study