Diagram of great auricular nerve sensory distribution regions. Number 1 is the preauricular region; 2, mandible body region; 3, infra-auricular region; 4, postauricular region; 5, lobule region; 6, inferior helix region; 7, superior helix region; and 8, concha region.
A patient 3 months postoperatively (A), 6 months postoperatively (B), 9 months postoperatively (C), and 12 months postoperatively (D). Areas outlined in dark green represent areas of anesthesia (ie, the patient did not respond to light touch). Areas outlined in red represent areas of paresthesia (ie, the patient felt an electrical, burning, or tingling sensation with light touch). This patient had one of the best sensory outcomes by 12 months.
A patient 3 months postoperatively (A), 6 months postoperatively (B), 9 months postoperatively (C), and 12 months postoperatively (D). Areas outlined in dark green represent areas of anesthesia (ie, the patient did not respond to light touch). Areas outlined in red represent areas of paresthesia (ie, the patient felt an electrical, burning, or tingling sensation to light touch). The outcome in this patient is representative of the average sensory outcome.
Ryan WR, Fee WE. Great Auricular Nerve Morbidity After Nerve Sacrifice During Parotidectomy. Arch Otolaryngol Head Neck Surg. 2006;132(6):642-649. doi:10.1001/archotol.132.6.642
Copyright 2006 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2006
To clarify the extent, timing, and patient perspectives of great auricular nerve (GAN) morbidity and recovery after nerve sacrifice during parotidectomy during the first postoperative year.
Tertiary care academic medical center.
Twenty-seven consecutive patients who underwent parotidectomy with GAN sacrifice.
Main Outcome Measures
Preoperatively and at 3, 6, 9, and 12 months postoperatively, we performed light touch sensation tests on each patient to develop an ink map representing anesthesia and paresthesia in the GAN sensory territory; patients also completed an outcomes questionnaire.
Twenty-two (81%) of 27 patients completed follow-up. The prevalence and average area of anesthesia decreased continually during the first year according to sensory testing and patient scoring. Half of the patients had no anesthesia at 12 months. The prevalence and average area of paresthesia increased during the first year according to sensory testing; however, the contiguity and subjective scoring of paresthesia peaked at 6 months and decreased in subsequent follow-up points. Throughout the first year, patients had difficulty using the telephone, shaving, combing their hair, wearing earrings, and sleeping on the operative side because of both anesthesia and paresthesia.
The impact of GAN sacrifice morbidity on patient quality of life is tolerable and improves during the first postoperative year. However, we feel that GAN morbidity may be bothersome enough to warrant efforts to preserve the posterior branch of the GAN when possible and appropriate.
Conventionally, surgeons have sacrificed the great auricular nerve (GAN) during parotidectomy to facilitate access to the parotid gland.1 Studies2- 4 have shown that GAN sensory loss can lead to anesthesia, paresthesia, discomfort, functional deficits (eg, difficulties wearing earrings and handling the telephone), an increased risk of traumatic injury, and an increase risk of neuromas. However, the extent, timing, and patients' perspectives of GAN morbidity and recovery after nerve sacrifice are not fully clear.
Some surgeons advocate the preservation of the posterior branch of the GAN with the intention of reducing unwanted sequelae. Retrospective3,5,6 and prospective7- 11 studies (including 2 randomized clinical trials) have found that GAN preservation is effective in increasing the likelihood and speed of return of nerve function. Conversely, other prospective evidence shows that GAN preservation may be unnecessary with the findings that sensory loss and recovery are similar between sacrificed and preserved GANs.12 Furthermore, although patients may lose some sensory function from GAN division, their quality of life is largely unaffected.13 Given that authors6,10 who advocate preservation of the posterior branch of the GAN report that GAN sacrifice is at times necessary, we need a better understanding of the consequences of GAN morbidity.
The GAN originates from the anterior rami of the second and third cervical nerves, bends around the posterior aspect of the sternocleidomastoid muscle, and ascends to just posterior of the tail of the parotid gland, where it divides into anterior and posterior branches.14 The anterior branch supplies sensation to the facial skin over the parotid gland and the ascending rami of the mandible. The posterior branch supplies sensation to the skin overlying the mastoid process and the skin of the lobule and helix.15 For GAN sacrifice, surgeons traditionally divide the trunk proximal to the bifurcation. Authors3,5,11 who perform posterior branch preservation divide the anterior branch for skin flap elevation.
This study is an effort to prospectively assess GAN morbidity after nerve sacrifice during parotidectomy during the first postoperative year. We hoped to determine, with respect to time after nerve sacrifice, (1) the extent of GAN morbidity and (2) the patients' perspectives of their neurological dysfunction and the resulting function problems.
This study received approval from the institutional review board's Human Subjects Ethics Committee at Stanford University, Stanford, Calif.
The study sample consisted of 27 consecutive patients who underwent parotidectomy with GAN sacrifice for any indication performed by 3 surgeons at the Department of Otolaryngology–Head and Neck Surgery, Stanford University, from March 2003 to February 2004. We excluded patients who also had had a neck dissection, could not communicate effectively enough to undergo sensory testing (eg, patients who were non–English speaking or who had dementia), had a generalized neurological illness, or could not commit to follow-up appointments. We obtained informed consent and permission to take photographs from each patient enrolled in the study.
We met with each patient preoperatively, and at 3, 6, 9, and 12 months postoperatively. We allowed for a month leeway for each postoperative follow-up appointment to ensure scheduling success. During these appointments, the patients first completed a questionnaire. Patients scored (on a 1-10 visual analogue scale) their assessment of the return of sensation, current paresthesia, satisfaction with the current sensation, the importance of full normal sensation, current pain, current discomfort, and how bothersome the dysfunction was for them all in the GAN sensory area. In addition, the patients answered yes or no to questions regarding the incidence of injury to the GAN sensory area, their feeling of being at risk for injury to the area because of decreased or absent sensation, and current difficulties with using the telephone, shaving, combing their hair, and wearing earrings or hearing aids. We also asked an open-ended question about any other potential problems associated with GAN function.
Figure 1 shows the 8 GAN sensory regions we defined for the purposes of this study: the preauricular, infra-auricular, postauricular, mandible body, lobule, inferior helix, superior helix, and concha regions. Preoperatively, the operator of the anesthesiometer (W.R.R.) applied 1 touch pressure to each of the 8 sensory regions areas to confirm intact sensation. All patients enrolled in the study had full light touch sensation according to the anesthesiometer testing and were without any subjective neurological dysfunction in the head and neck area prior to their operation.
At each postoperative follow-up visit, after completion of the questionnaire, one operator (W.R.R.) performed sensory testing over the area of GAN sensory distribution. For sensory testing, we used Semmes-Weinstein pressure microfilament anesthesiometers (Smith & Nephew Rolyan, Germantown, Wis), which are flexible plastic fibers connected to a hard plastic rod. When pressing the fiber against the skin of a patient to the point at which the fiber bends, the operator transmits a standardized, constant, and reproducible amount of force to the skin.16,17 We used the 4.31-J anesthesiometer because this amount of force represents the threshold of normal protective sensation.16 For the purposes of this study, we define anesthesia at this sensory threshold.
With the patient's eyes closed, the operator applied the anesthesiometer multiple times in a sequential fashion, but with an irregular rhythm, throughout the entire test area. The operator asked each patient to respond with the phrase “touch” if he or she felt the force of the anesthesiometer and to respond with the phrase “up” if he or she felt a paresthesia-like sensation. We defined paresthesia as any sensation that felt electrical, burning, or tingling. Based on the responses, the operator delineated the areas of anesthesia and paresthesia with green and red felt-tip markers, respectively. These ink delineations created a sensory dysfunction map on the skin of each patient (Figure 2 and Figure 3). The operator determined the presence of referred sensation in each of the 8 predefined sensory regions by asking the patient (whose eyes were again closed) to show with his or her finger the point where the operator had applied the anesthesiometer fiber. We defined referred sensation as a light touch sensation where the patient's finger designation was more than 2 cm away from the actual spot stimulated by the anesthesiometer. The operator then took digital photographs of the ink map on the face, neck, and ear of each patient with a digital camera (PowerShot S400 Digital Elph, 4.0 megapixel; Canon, Lake Success, NY) and traced the ink maps onto a translucent piece of paper. On the translucency, the operator made black ink outlines of the auricle, the lateral corner of the eye, and the lateral corner of the mouth to show anatomic reference landmarks.
We recorded whether patients had diabetes mellitus or any other disease with possible effects on peripheral nerve function. At each follow-up point, we also recorded any event of neoplasm recurrence, salivoma, hematoma, cellulitis, neuroma, or a numbness-related injury.
We determined the percentages of patients who had anesthesia, paresthesia, and referred sensation in any and in each of the 8 predefined sensory regions. We analyzed the presence of sensory dysfunction in the different regions to show the change in the area of anesthesia and paresthesia over time and to identify the degree of sensory dysfunction in different specific regions. Ultimately, we removed the concha sensory results when determining the average percentage of sensory regions with anesthesia or paresthesia because GAN sacrifice had minimal effects on the sensation of the concha. We determined the average score and ranges for each of the questionnaire questions.
We stratified the patients according to the following attributes: older than 60 years, sex, type of operation (enucleation vs superficial vs total), incision type (modified face-lift vs modified Blair incision), use of adjuvant therapy (radiation therapy or chemotherapy), comorbidities, and postoperative complications to reveal any trends among these factors and the extent, timing, and patients' perspectives of GAN function morbidity.
One patient underwent a bilateral parotidectomy. We counted this patient as 2 patients for the purposes of this study.
Twenty-two (81%) of 27 patients completed follow-up. One 101-year-old study subject died during the follow-up period. (We had operated on this patient, using local anesthesia, because he had intractable pain related to a neuroendocrine carcinoma. He lived 6 more months free of pain, ultimately dying of heart disease.) We were unable to contact 1 patient after enrollment. Three patients dropped out of the study owing to lack of interest. Of the 22 patients who completed 12 months of follow-up, 1 missed the 3-month follow-up point, and 3 missed the 9-month follow-up point.Table 1 displays the characteristics and operative outcomes of the 22 patients with 1-year follow-up.
Table 2 shows the percentage of patients with anesthesia in any and in each of the 8 predefined areas supplied by the GAN at different follow-up points.Table 3 shows the percentage of patients with paresthesia in any and in each of the 8 predefined areas supplied by the GAN at different follow-up points. Patients at 1 year had an average of 3.4 regions of paresthesia, although these areas of paresthesia had much less contiguity and generally occupied less space within each region. Three patients reported that they felt the paresthesias only with the anesthesiometer testing and not in their daily life. Reportedly, the paresthesias would often last several seconds after a stimulus; interestingly, many patients applied pressure to the area of paresthesia with the full palm of their hand to relieve the discomfort. The prevalence of patients who experienced referred sensations was 29% (6 of 21) at 3 months, 68% (15 of 22) at 6 months, 90% (17 of 19) at 9 months, and 91% (20 of 22) at 12 months.Table 4 shows the patients' questionnaire scores (from 1 to 10) for various aspects of GAN sensory function.
None of the patients sustained an injury to the GAN sensory area at any time in the first postoperative year. At each follow-up, a percentage of patients felt they were at risk for injury owing to the abnormal sensation on the operative side of the face: 10% (2 of 21) at 3 months, 18% (4 of 22) at 6 months, 5% (1 of 19) at 9 months, and 23% (5 of 22) at 12 months. Patients reported activities with potential to cause injury to be shaving, using curling irons, and sleeping on the operative side.
Patients reported problems with handling a telephone on the operative side throughout the first postoperative year: 81% (17 of 21) at 3 months, 64% (14 of 22) at 6 months, 32% (6 of 19) at 9 months, and 59% (13 of 22) at 12 months. Specific complaints included discomfort, increased sensitivity, strange sensations, inability to place telephone correctly, and that “the ear doesn't feel like my ear.” Patients with telephone difficulties generally used the telephone on the nonoperative ear instead.
Male patients reported problems with shaving on the operative side throughout the first postoperative year: 40% (4 of 10) at 3 months, 30% (3 of 10) at 6 months, 30% (3 of 10) at 9 months, and 20% (2 of 10) at 12 months. Discomfort and concern for potential injury were the main complaints in shaving.
Patients reported problems with combing their hair on the operative side throughout the first postoperative year: 10% (2 of 21) at 3 months, 18% (4 of 22) at 6 months, 11% (2 of 19) at 9 months, and 18% (4 of 22) at 12 months. Discomfort and hitting the ear with the comb were the main complaints in combing hair.
Patients reported problems with wearing earrings on the operative side throughout the first postoperative year: 70% (7 of 10) at 3 months, 30% (3 of 10) at 6 months, 78% (7 of 9) at 9 months, and 40% (4 of 10) at 12 months. Specific complaints included discomfort, increased sensitivity, difficulty putting on and keeping on earrings, and the need for a mirror to put on earrings. Although asked about it in the questionnaire, none of the patients wore hearing aids.
Patients reported the following problems associated with GAN dysfunction: discomfort when sleeping on the operative side, difficulty putting on and wearing glasses (owing to uncertain placement on ear), discomfort when being kissed or caressed in the operative area, difficulty with the placement of a stethoscope in the operative side ear (2 physicians were among the study sample), difficulty using a curling iron (for fear of ear injury), inability to sense whether or not long hair was tucked behind the ear, and a sense of reduced hearing on the operative side. Most patients reported discomfort when sleeping on the operative side, especially in the first 3 months.
Ninety-one percent (20 of 22) had some type of functional deficit in daily activities at some point in the postoperative course. Overall, there was a steady improvement in the numbers of patients without subjective functional deficits in daily activities: 18% (4 of 22) at 3 months, 24% (5 of 21) at 6 months, 32% (6 of 19) at 6 months, and 45% (10 of 22) at 12 months. Some patients with early functional deficits regained their use at a later point after surgery. Some patients developed problems later in the postoperative course. For some patients, the main problem was discomfort related to anesthesia, usually at the 3-month point; for others, the main problem was discomfort related to paresthesia, usually at the 6-month point.
We did not expect to have, nor did we have, enough study subjects to achieve suitable statistical power for determining statistically significant influential variables. Figure 2 and Figure 3 show the GAN sensory dysfunction in 2 different patients at 3, 6, 9, and 12 months postoperatively.
Anesthesia from GAN morbidity subsided continually during the first postoperative year both in the percentage of patients and amount in each patient (Table 2). The presence and span of paresthesia increased during the first postoperative year (Table 3). However, the distribution of paresthesia at 12 months in patients was less contiguous and occupied less space within each region compared with that of the 6-month period. In conjunction, the patients also, on average, reported less paresthesia (3.1 regions of 10) at 12 months than at 6 months (4.8 regions of 10) (Table 4). We conclude from this data that on average the paresthesia ultimately lessened in severity in the later half of the first postoperative year.
Through sensory testing, we found most often the following evolution of findings: early anesthesia followed by paresthesia and referred sensation followed by normal sensation in later follow-up points. Recovery of GAN function may occur for a number of reasons, including the regeneration of the GAN nerve fibers; collateral innervation by the lesser occipital nerve, auriculotemporal nerve, trigeminal nerve, and transverse cutaneous nerve; and the patient's own psychological adaptation to the sensory loss. Paresthesia and referred sensations may be the result of an abundant or immature reinnervation of the skin overlying the sectioned branch. The differences in the scope and severity of paresthesias and hypoesthesia found in the patients in this study are likely due to biologic vagaries of wound healing.
The patient questionnaire scores for sensation and paresthesia (Table 4) generally correlated well with the results of the anesthesiometer testing. As expected, the satisfaction scores generally correlated with the sensation scores as they improved. The importance of full normal sensation being, on average, approximately a score of 7 to 8 throughout the first postoperative year may demonstrate that patients have only a moderate need for full sensation. On average, the pain, discomfort, and bothersome morbidity scores (Table 4) for patients were low at 3 months and continued to drop over the subsequent follow-up visits. These low scores show the low impact that GAN sensory dysfunction has on patients' lives. Scores for discomfort and bothersome morbidity being slightly higher than pain scores perhaps indicates the more appropriate descriptions for GAN sensory dysfunction.
Other authors6- 8,13 have reported similar daily activity functional deficits with susceptibilities to injuries and burns; difficulties with telephone use, shaving, combing their hair, wearing earrings, and facial contact during intimacy; and discomfort wearing glasses. In this study, patients proportionately had the most problems with telephone use and wearing earrings and the least problems with shaving and combing their hair. For both telephone use and earring wearing, the percentage of patients with problems lessened in the middle of the first postoperative year and then increased at the later part of the year. Patients explained that this phenomenon was due to the annoyance and discomfort of the anesthesia subsiding but then being replaced by that of the paresthesia.
This study serves as a basis for comparison with studies of patients in whom the GAN was preserved. These prospective7- 11 and retrospective studies,3,5,6,18 using various methods to assess nerve function, have shown that sensory loss is less, recovery of function is earlier, and that quality of life is higher in patients who had their GAN posterior branch preserved. Studies9 have reported that all patients who underwent preservation had no anesthesia by 6 months8 or by 12 months. Some evidence showed that the group who had undergone sacrifice of the GAN achieved a sensory recovery plateau by 6 months postoperatively8 and had dysfunction up to 2 years later.10 Another study9 showed that 90% of patients who had undergone sacrifice of the GAN still had anesthesia at a minimum of 8 years of follow-up. The patients in our study did not experience this type of recovery course. Half of the patients had no anesthesia at 12 months. Furthermore, at 12 months, patients, on average, had a fifth of the amount of anesthesia they had at 3 months. No sensory plateau occurred in our study; instead, we observed a continual abatement in sensory loss. Overall, on average, daily activity function also continually improved.
Preservation and posterior retraction of the nerve trunk and posterior branch seem technically possible and achievable approximately 70% of the time.6,10 Preservation should not be performed when the tumor is in close proximity to the posterior branch of the GAN. In addition, the posterior branch is not always present to be preserved. Because preservation is not possible in nearly a third of cases, more information about the neurological ramifications of GAN sacrifice, as presented in this study, is important.
To our knowledge, no authors have attempted anterior branch preservation. Dividing the anterior branch is necessary to separate parotid fascia from the sternocleidomastoid muscle, raise the anterior skin flap, remove the nerve branch that is in close proximity to the tumor, and avoid injury to the marginal mandibular nerve.8,12,18 In one study6 supporting posterior branch preservation, there was no difference in the return of cheek skin sensation (GAN anterior branch distribution) between patients with sacrificed and preserved posterior branches of the GAN; 50% of both groups experienced significant sequelae.6 However, most studies have ignored altogether the anesthesia in the anterior branch distribution in their analysis.
Evidence does exist that suggests that GAN preservation may be unnecessary. Porter and Wood12 presented prospective controlled evidence that suggested no difference in the postoperative recovery whether or not the GAN was preserved. Also, Patel et al13 concluded that GAN sacrifice does not have a very large impact on patients’ quality of life. In patients whose GAN was sacrificed, 23 (77%) reported only a little or no bother caused by the symptoms; 27 (90%) reported no or almost no interference with their daily lives.13 Authors of the preservation studies9,10 have explained that the temporary neurapraxia of the GAN is caused by intraoperative manipulation and devascularization of the posterior branch. Scar tissue associated with surgery may ultimately entrap the GAN and compromise or obliterate its function regardless of operative efforts.
This prospective study provides more details about the character and evolution of the sensory sequelae, functional deficits, and patients' perspectives related to GAN morbidity. We feel that the data in this study can also be generalized to other surgical areas such as rhytidectomy, neck dissection, neck mass resection, trauma, or GAN cable graft use, all of which likely would leave patients with postoperative symptoms similar to those described herein. We feel that it is important to be able to provide patients with a more detailed explanation of what to expect with GAN recovery in the first postoperative year. With reassurance, we hope that patients will have increased satisfaction and a decreased amount of anxiety with the results of the operation.
This study used monofilament anesthesiometers to determine multiple points of light touch dysfunction in the GAN-associated area whereas other studies tested only a few representative points.6 We used the Semmes-Weinstein anesthesiometers because of their reliability, reproducibility, and simplicity, as others have.17,19 We did not assess thermal, pain, 2-point discrimination, or vibration sense as other studies have.8- 10 We felt that light touch would be the most important, telling, accurate, reliable, and efficient way of determining the sensory function of the GAN.
Limitations of this study include the following: we experienced some loss of patients at follow-up, which compromised the consecutive character of the original patient cohort. In using 3 different surgeons, significant unknown differences in operative technique may bring bias to the study. Only 1 operator (W.R.R.) performed all of the sensory exams. The test-retest reliability of sensory testing, thus, is a potential source of bias. The subjective nature of patients' comprehension and rating of their sensory function could have brought error to the study. For example, 3 patients reported that they felt the paresthesias only during anesthesiometry testing and not in their daily life. Our use of a 4.31-J anesthesiometer bases our definitions of anesthesia and paresthesia on this amount of force. Other amounts of standardized force could have resulted in different sensory findings. Sensory dysfunction may vary day to day and hour to hour. Our analysis captures only 4 discrete time points for each patient.
The impact of GAN sacrifice morbidity on patient sensory function and quality of life is tolerable and generally improves during the first postoperative year. The prevalence and average area of anesthesia decreased continually during the first year, according to sensory testing and patient scoring. Half of the patients had no anesthesia at 12 months. Paresthesias are a very common part of the postoperative course of a patient who has undergone GAN sacrifice. The prevalence and average area of paresthesia increased during the first year according to sensory testing; however, the contiguity and subjective scoring of paresthesia peaked at 6 months and decreased in subsequent follow-up. Most patients had difficulty using the telephone, shaving, combing their hair, wearing earrings, and sleeping on the operative side because of both anesthesia and paresthesia. The collective prevalence of these functional deficits improved continually during the first postoperative year.
The recovery of the GAN seems to follow the progression from anesthesia to paresthesia with referred sensation to complete sensation. We can offer patients reassurance about any of these symptoms.
We compare the sensory dysfunction of the patients in this study with that in studies of patients in whom the posterior branch was preserved. Although we have not provided evidence showing that preservation of the posterior branch would necessarily result in mitigation of resultant sensory dysfunction, we feel that GAN morbidity is relatively bothersome enough to warrant efforts to preserve the posterior branch of GAN when possible and appropriate. In sparing the posterior branch of the GAN, it may be possible to decrease sensory dysfunction and improve quality of life. Studies of GAN morbidity in the second postoperative year and more randomized-controlled trials comparing nerve sacrifice and preservation are needed.
Correspondence: William R. Ryan, MD (Resident), Department of Otolaryngology–Head and Neck Surgery, Stanford University School of Medicine, 801 Welch Rd, Stanford, CA 94306 (email@example.com).
Submitted for Publication: April 29, 2005; final revision received September 15, 2005; accepted October 19, 2005.
Author Contributions: Dr Ryan had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Financial Disclosure: None.
Funding/Support: This study was funded by a scholarship for medical students from the Stanford University Medical Scientist Scholars Program (Dr Ryan).
Previous Presentation: Dr Ryan presented this study at the American Head and Neck Society Annual Meeting during the Combined Otolaryngology Society Meetings; May 15, 2005; Boca Raton, Fla.
Acknowledgment: We thank Simon Kimm, MD, for creating the illustration in Figure 1.