Preservation of Salivary Function Following Extracapsular Dissection for Tumors of the Parotid Gland.

PURPOSE: This study evaluated salivary function after extracapsular dissection (ECD) compared with partial superficial parotidectomy (PSP) and classic superficial parotidectomy (CSP) of benign parotid gland tumors. The authors hypothesized that ECD would be superior to PSP and CSP in preserving postoperative salivary function. MATERIALS AND METHODS: Retrospective analyses were performed for 43 consecutive patients who underwent parotidectomies of benign parotid tumors performed by a single experienced surgeon. Clinical data and pre- and postoperative whole salivary flow rates were compared among the operative procedures. Pearson χ2 and Fisher exact tests were used to compare categorical variables. Kruskal-Wallis, Mann-Whitney U, and Wilcoxon signed rank tests were used to compare means. A P value less than .05 was considered significant throughout the study. RESULTS: Tumor sizes did not differ among groups. Operative times, amounts of drainage, and hospital days for ECD were markedly decreased compared with CSP and tended to be decreased compared with PSP. Resection margins were exposed in 0, 12.5, and 6.7% of patients who underwent ECD, PSP, and CSP, respectively. Postoperative complications occurred less often (but not meaningfully) after ECD. Postoperative basal salivary flow rates in the ECD, PSP, and CSP groups were 0.39, 0.32, and 0.14 mL/minute, respectively (P = .05). Stimulated salivary flow rates remained stable for the ECD and PSP groups but decreased in the CSP group. CONCLUSION: ECD is a safe and time-efficient surgical approach, offering early recovery from parotid tumors and better preservation of salivary function. ECD should be considered a surgical approach for parotid tumors, especially those in the parotid tail region, such as Warthin tumors.

Is surgical excision of lipomas arising from the parotid gland systematically required?

Lipomas arising from the parotid gland are very rare. We report a 10-year experience in a single institution (La Timone University Hospital of Marseille, France). Among 614 parotidectomies for neoplasms performed from 1998 to 2008, 12 lipomas were identified. A retrospective analysis based on medical records was made. Evaluation, analysis and current management of lipomas of the parotid gland are described. Lipomas accounted for 2% of all parotid neoplasms and 2.6% of benign tumors in our series. The median age of patients was 60 years with a M/F sex ratio of 5-1. The main presentation was a soft asymptomatic, slow-growing, mobile mass although 30% had an indurated mass on palpation. Diagnosis of lipoma, based on the results of imaging, was made preoperatively in all cases. The mean tumor duration prior to excision was 11.5 months. The surgical decision was made regarding increased swelling with functional/esthetic discomfort in 83% of cases. Partial parotidectomy was performed in most cases. Postoperative complications occurred in 16% of cases although no permanent complication was observed. No recurrence was observed in our series. Histologically, 92% of tumors were classic lipomas. Lipomas can be clinically misleading since 30% of patients in our series showed an indurated mass on palpation. Preoperative imaging, especially MRI, is the cornerstone of their management as it allows very accurate lipoma diagnosis. Since in our series, diagnosis of lipoma had been made preoperatively in all cases, the surgical excision could be delayed and finally surgical decision has been made for esthetic and/or functional considerations in more than 80% of cases.

[Intraoperative facial nerve monitoring in parotid gland surgery].

OBJECTIVE: To re-evaluate the value and the methods of intraoperative facial nerve monitoring in parotid gland surgery. METHODS: Sixty-five cases received intraoperative facial nerve monitoring in parotidectomy (test group) since 2000 - 2008. The facial nerve was identified through central trunk method (n = 18), branch method (n = 35) and mixed method (n = 12). Most patients accepted general anesthesia by incubation. The operating duration and minimum electronic stimulation threshold values of EMG in evoked facial muscle were recorded. Facial nerve was identified though branch method (n = 44) and no intraoperative facial nerve monitoring was performed in parotidectomy (control group). RESULTS: There were four cases (6.1%) of mild temporary paralysis and no permanent post-operative paralysis of facial nerve in the test group. The average operating duration was 1.8 hour. The minimum reactive electronic stimulation threshold of EMG in evoked facial muscle was 0.08 mA. The range of suitable electronic stimulation threshold of EMG was from 0.2 mA to 1.0 mA. While there were nine cases (20.5%) of mild temporary paralysis and two cases (4.5%) of permanent post-operative paralysis of facial nerve in the control group and the average operating duration was 3.0 hours. CONCLUSION: Intraoperative facial nerve monitoring (IFNM) in parotidectomy can assist a surgeon to confirm and identify the facial nerve and exercise precautions so as to shorten operating duration and prevent potential surgical complications.

Rare Epididymal-Derived Myoepithelial Carcinoma: A Case Report and Review of Literature.

Myoepithelial carcinoma (MC), also known as malignant myoepithelial neoplasm, is more common in the parotid glands of the head and neck. The main clinical manifestation is the growth of a nonspecific, painless mass at the primary site. We report the first case of MC derived from the epididymis. The current reports of non-parotid MC are still rare, and epididymal-derived MC has not been reported previously. Simultaneously, we explore the role of EWSR1 fusion as a predicting marker and further reveal the origin of MC to provide new ideas for its diagnosis and treatment.

Risks of Neoplasia and Malignancy in Surgically Resected Cystic Parotid Lesions.

OBJECTIVES: To evaluate the risks of neoplasm and malignancy in surgically treated cystic parotid masses compared with solid or mixed lesions and to evaluate the performance of fine-needle aspiration (FNA) in parotid cysts. STUDY DESIGN: Retrospective cross-sectional study. SETTING: Single-institution academic tertiary care center. SUBJECTS AND METHODS: Patients without a history of human immunodeficiency virus or head and neck cancer who underwent parotidectomy for parotid masses and had preoperative imaging to characterize lesions as cystic, solid, or mixed (ie, partially cystic and partially solid). We assessed the risks of neoplasia and malignancy, adjusting for age, sex, race/ethnicity, facial nerve weakness, and history of malignancy. We also evaluated the sensitivity and specificity of FNA. RESULTS: We included 308 patients, 27 of whom had cystic parotid masses (5 simple and 22 complex). Cystic masses were less likely to be neoplastic compared to solid or mixed masses (44% vs 97%; odds ratio [OR], 0.03; 95% confidence interval [CI], 0.01-0.07); however, there was no difference in the risk of malignancy (22% vs 26%; OR, 0.81; 95% CI, 0.32-2.10). Cystic masses were more likely to yield nondiagnostic FNA cytology results, but for diagnostic samples, FNA was 86% sensitive and 33% specific for diagnosing neoplasia and 75% sensitive and 83% specific for diagnosing malignancy. CONCLUSION: In our population, cystic masses undergoing surgery were less likely to be neoplastic but had a similar risk of malignancy as solid masses. The risk of malignancy should be considered in the management of cystic parotid masses.

Parotidectomy-related facial nerve lesions: proposal for a modified Sunnybrook Facial Grading System.

OBJECTIVE: Facial nerve injury, affecting mainly the marginal mandibular branch, is the most frequent neurologic complication from parotidectomy. To test a modified Sunnybrook Facial Grading System as a new tool to assess the facial nerve function following parotidectomy, emphasizing the marginal mandibular branch. METHODS: We reviewed the medical records of 73 post-parotidectomy patients (40 female, 18-84 years old, mean age 53.2 years) with facial nerve sparing, referred to the Department of Physical Therapy. All patients had parotid neoplasms or advanced skin cancer, and were followed by the principal author between 2006 and 2014. RESULTS: The muscles innervated by the marginal mandibular branch were the most frequently affected (72.6%), particularly in patients undergoing neck dissection (p = 0.023). The voluntary movement scores obtained with the modified system were significantly lower compared with the original version (p < 0.001). The best and worst scores were observed in patients with benign parotid tumors and skin cancer, respectively. Patients requiring neck dissection (p = 0.031) and resection of other structures (p = 0.021) had the lowest scores, evidenced only with the modified version. Patients with malignant tumors had significantly worse ratings, regardless of the Sunnybrook system version. The post-physiotherapy analysis involved 50 patients. The worst facial rehabilitation outcomes were related to the marginal mandibular branch function. CONCLUSION: The modified Sunnybrook Facial Grading System improved the marginal mandibular branch assessment, preserving the evaluation of other facial nerve branches.

Auriculotemporal Nerve Involvement in Parotid Bed Malignancy.

OBJECTIVE: To identify and evaluate patients with parotid bed malignancy demonstrating radiographic findings of auriculotemporal (AT) nerve involvement. METHODS: A retrospective review of patients with parotid bed malignancy was performed to identify patients with imaging findings of AT nerve involvement and record associated clinical findings, symptoms, and pathology information. Independent, blinded review of radiographic images by a senior neuroradiologist was performed to identify imaging characteristics and categorize patients into highly likely or possible involvement groups. RESULTS: Of 547 patients identified with parotid bed malignancy, 23 patients exhibited radiographic findings suggestive of AT nerve involvement. Thirteen patients met criteria for highly likely involvement, and 10 patients met criteria for possible involvement. Cutaneous malignancy with metastasis to the parotid bed accounted for 11 of 23 patients, and the most common histology was squamous cell carcinoma (9 patients). Primary parotid malignancy accounted for 12 of 23 patients, and the most common histology was salivary ductal carcinoma (3 patients). All 13 highly likely patients reported periauricular pain, and 11 of 13 demonstrated facial weakness. Features suggesting advanced disease included radiographic findings of intracranial involvement (10/23 patients), nonsurgical primary treatment (13/23 patients), and positive margins on pathology report (7/10 patients). CONCLUSION: AT nerve involvement is an uncommon but important phenomenon that often occurs in the setting of advanced disease and is commonly associated with periauricular pain and coexisting facial weakness. Awareness of the associated clinical features and imaging patterns can allow for appropriate identification of this pattern of spread and help to optimize treatment planning.

Differentiation of benign and malignant parotid tumors using deconvolution-based perfusion CT imaging: feasibility of the method and initial results.

AIM: We evaluated the feasibility of perfusion CT (CTP) of the parotid gland and attempted to differentiate benign from malignant tumors. MATERIALS AND METHODS: CTP was performed in 17 patients with benign tumors and 10 patients with malignant parotid tumors. Data were postprocessed by using deconvolution-based perfusion analysis. Postprocessing-generated maps showed blood flow (BF), blood volume (BV), mean transit time (MTT), and capillary permeability surface product (PS). Regions of interest were placed through the tumor site and the contralateral healthy parotid tissue. Ratios of the perfusion values between the tumors and the contralateral healthy structures were also calculated. Pearson correlation coefficients were determined to compare the agreement between the two readers. RESULTS: Perfusion maps of all tumors were successfully obtained. High Pearson correlation coefficients comparing the two readers' visually measured abnormalities were observed (r=0.79-0.86, P=0.001) for all perfusion maps, The MTT and PS values between malignant and benign tumors were not significantly different. The BF and BV values were statistically significant different between the benign and malignant tumors (0.00

Interleukine-6 (IL-6) may be a link between myasthenia gravis and myoepithelioma of the parotid gland.

Myoepithelioma is a rare benign neoplasm of the salivary glands occurring more frequently in the parotids. Myasthenia gravis (MG) is a chronic, T-cell dependent, antibody and complement-mediated autoimmune neuromuscular transmission disorder. Interleukine-6 (IL-6) is an immune protein belonging to the family of the hematopoietins, liberated in response to infection, burns, trauma, and neoplastic diseases. It seems that an overproduction of IL-6 might play an important role in the pathophysiology of MG. Moreover, it has been discussed the possible role of IL-6 as a modulating factor either in proliferation or in differentiation of pleomorphic adenoma cell line into myoepithelioma. The authors present a rare case of parotid myoepithelioma occurred in a patient affected by myasthenia gravis and suppose a possible IL-6 mediated relationship between myasthenia gravis and parotid myoepithelioma.

Sebaceous lymphadenoma identified by fine needle aspiration biopsy: a case report.

BACKGROUND: Sebaceous lymphadenoma of the parotid gland is a rare benign neoplasm. This is the first reported case of fine needle aspiration biopsy (FNAB) findings for sebaceous lymphadenoma of the parotid gland. CASE: A 60-year-old male presented with painless, bilateral parotid swelling noted for 5 months. The swelling was more pronounced on the right. Examination revealed bilaterally prominent parotid glands with diffuse firmness but no discrete masses. There was no evidence of facial nerve dysfunction. Laboratory evaluation was negative for infectious and autoimmune etiologies. Magnetic resonance imaging revealed bilateral cystic parotid masses. FNAB of the right parotid was obtained to assist with preoperative counseling. It revealed lymphoid and salivary gland parenchymal cells. The patient underwent a right superficial parotidectomy. The surgical specimen of the parotid mass confirmed the diagnosis of sebaceous lymphadenoma on the tissue section. The contralateral parotid mass had not been excised at this writing. CONCLUSION: This report is the first to describe the FNAB findings of the unusual benign parotid neoplasm sebaceous lymphadenoma. Though the definitive diagnosis of any parotid mass requires tissue, generally obtained via parotidectomy, an FNAB diagnosis can be useful in counseling a patient prior to definitive biopsy.

Acute facial paralysis due to primary squamous cell carcinoma of the parotid gland.

The incidence of primary squamous cell carcinoma in salivary gland tumours is rare. The first presenting symptom of a salivary gland tumour is usually a painless lump. Acute facial paralysis is uncommon to be the initial presentation. This case report describes a patient of acute facial paralysis due to primary squamous cell carcinoma of parotid gland who was initially incorrectly labelled as a case of Bell's palsy. He was managed successfully with total parotidectomy, neck dissection, facial nerve sacrifice with immediate cable graft reconstruction and fascia lata sling followed by postoperative radiotherapy. The sacrifice of the extra-temporal part of the facial nerve in rare cases can be unavoidable during parotid resections for malignancy when the nerve is grossly involved by the tumour. It is important in this situation that some form of nerve repair and/or facial rejuvenation is undertaken to avoid the unpleasant sequele of facial paralysis. His postoperative facial symmetry improved at rest and one year later he regained satisfactory eye closure and smile. Not all facial paralysis is Bell's palsy. All patients who are labelled as Idiopathic facial paralysis or Bell's palsy should have adequate clinical evaluation with detailed history and proper physical examination.

Total versus superficial parotidectomy for stage III melanoma.

BACKGROUND: The primary purpose of this study was to describe the parotid recurrence rates after superficial and total parotidectomy. METHODS: A retrospective cohort study was performed on patients with cutaneous melanoma metastatic to the parotid gland who underwent parotidectomy from 1998 through 2014. Primary outcome was parotid bed recurrence. Secondary outcomes were facial nerve function postoperatively and at last follow-up. RESULTS: One hundred twenty-nine patients were included in the study. Thirty-four patients (26%) underwent a total parotidectomy and 95 patients underwent superficial parotidectomy. Twelve patients (13%) developed parotid bed recurrence after superficial parotidectomy alone versus zero after total parotidectomy (P = .035). Facial nerve function, clinically detected disease, stage, and adjuvant treatment were not statistically different between the groups (P = .32, .32, .13, and 0.99, respectively). CONCLUSION: Parotid bed melanoma recurrence was more common after superficial parotidectomy compared to total parotidectomy, and recurrence resulted in significant facial nerve functional deficit. Our results support total parotidectomy when metastatic melanoma involves the parotid nodal basin.

Facial nerve function in carcinoma of the parotid gland.

AIM: To analyse, for patients with carcinoma of the parotid gland, the prognostic value for treatment outcome of the function of the facial nerve (NVII), and determining facial nerve dysfunction after treatment. METHODS AND MATERIALS: In a retrospective study of the Dutch head and Neck cooperative group (NWHHT), data of 324 patients with parotid carcinoma were analysed. The function of N VII before treatment was intact in 77%, partially and completely impaired in 14% and 7%, respectively. Eighty-eight percent of the patients were treated surgically, and 77% of them were treated by a combination of postoperative radiotherapy. In 21% NVII was sacrificed, a reconstruction was performed in one of three. RESULTS: Independent risk factors for N VII dysfunction before treatment were tumour localisation, positive neck nodes at presentation, pain, increasing age, and perineural invasion. Regional, not local, control was significantly impaired for complete facial paralysis. N VII dysfunction was an independent factor for disease free survival, and was 69%, 37% and 13% for normal, partially and completely impaired function, respectively. After treatment 22% of the patients experienced a partial paralysis, and 13% of the patients experienced a complete paralysis of N VII. CONCLUSION: For patients with parotid carcinoma, facial nerve function before treatment is a strong prognostic factor for disease free survival.

Schwannomalike mixed tumor of the parotid gland: a case report.

BACKGROUND: Schwannomalike mixed tumor is a rare benign tumor of the parotid glands. CASE: A 75-year-old woman presented with a tumor 3 cm in diameter localized in her left parotid gland. Fine needle aspiration (FNA) of the lesion revealed 2 types of cells: 1 with elongated, wavy, hyperchromatic nuclei and a scant cytoplasmic border and 1 with larger, weakly staining nuclei and more abundant cytoplasm. Morphologic examinations were performed. CONCLUSION: FNA cytology was very useful in the diagnosis of this rare tumor of the parotid gland.

Parotid lymphomas--clinical and computed tomographic imaging features.

OBJECTIVE: To review the clinical presentation and computed tomography (CT) imaging characteristics of all parotid lymphomas diagnosed at the study institution over a 7-year period. DESIGN: Retrospective chart review of parotid lymphomas diagnosed between 1997 and 2004. SUBJECTS: A total of 121 patients with parotid lesions were identified. After retrospective chart review, a total of 10 patients with histologically proven parotid lymphoma were included in the study, 8 of whom had CT scans available for assessment. RESULTS: Ten patients with histologically proven lymphoma of the parotid gland were identified from among 121 patients with parotid neoplasms, an incidence in this series of 8.3%. All lymphomas were of non-Hodgkin's type. All patients presented with a painless unilateral parotid swelling. Most patients had a short history of less than 4 months' duration, of whom 3 presented with a rapidly evolving swelling of less then 1 month's duration. No patient had a background of Sjögren's disease or any other autoimmune disorders. The commonest finding noted on CT was of a unilateral, single mass of relative soft-tissue homogeneity with poorly defined, indistinct tumour margins. Associated loco-regional lymphadenopathy was identified in 2 cases, 1 clinically and another radiologically; multiple ipsilateral lesions were noted in 2 cases. No cases of contralateral disease were observed. CONCLUSION: Lymphoma has a clinical presentation similar to other neoplasms arising within the parotid gland. A unilateral, non-tender swelling was a universal finding. A history of less than 4 months may suggest the possibility of lymphoma. CT scanning is a useful adjunctive investigation to determine the site and extent of the disease, loco-regional nodal status and contralateral gland and neck status. Multifocality and associated adenopathy are associated with, but not exclusive to, parotid lymphoma. Although poor tumour boundary definition on CT imaging is a strong predictor of malignancy, no pathognomonic finding specific for lymphoma has been identified. The potential diagnosis of parotid lymphoma should be considered in all patients who present with a parotid mass.

[Granular cell tumours (Abrikossoff's tumour) of the parotid region]. / Tumeur à cellules granuleuses (tumeur d'Abrikossoff) de la loge parotidienne.

BACKGROUND: Granular cell tumour (Abrikossoff's tumour) was first described by Abrikossoff in 1926. These tumours are rare and usually presents as a solitary lesion, located mainly in the subcutaneous tissue of the head and neck, and in the oral cavity (tongue). CASE REPORT: We report a rare case of a granular cell tumor of the parotid gland, in a 55-year old woman, who was referred with a left preauricular mass that had rapidly increased in size over 2 months. There was no cervical lymph adenopathy. RMI demonstrated a solitary lesion of the parotid gland. Surgical resection was performed. CONCLUSION: We discuss the classification, pathophysiology and the treatment of granular cell tumours through a review of the literature.

Xanthogranulomatous sialadenitis: a case report and literature review.

Xanthogranulomatous tissue reaction is an uncommon but well-documented process that occurs at many sites in the body. It is most often recognized in the kidney and gallbladder, where its etiology is believed to involve an outflow obstruction. We report the case of a man with a parotid mass that exhibited features consistent with an inflammatory process on fine-needle aspiration biopsy. The mass persisted despite medical management, and the patient subsequently underwent a superficial parotidectomy. Histologic examination of the resected specimen identified a xanthogranulomatous tissue reaction adjacent to a Warthin's tumor. We compare the features of this case with those of the 2 previously reported cases of xanthogranulomatous sialadenitis, and we discuss its possible etiologies.

Ultrasound and CT imaging features in a patient with salivary duct carcinoma of the parotid gland: a case report with literature review.

The aim of this paper was to present the ultrasound (US) and computed tomography (CT) appearance of a patient with salivary duct carcinoma of the parotid gland. US showed a voluminous mass of the parotid gland, with multiple calcifications. Furthermore, it revealed regional multiple lymph nodes with malignant characters. Sonoelastography of the lesion and lymph nodes detected increased rigidity. Contrast enhanced CT scan of the neck completed the data description regarding the mass expansion and invasion of surrounding tissues. US and CT imaging features played a key role in establishing the malignant character of the mass and lymph nodes.

An analysis of facial nerve function in irradiated and unirradiated facial nerve grafts.

PURPOSE: The effect of high-dose radiation therapy on facial nerve grafts is controversial. Some authors believe radiotherapy is so detrimental to the outcome of facial nerve graft function that dynamic or static slings should be performed instead of facial nerve grafts in all patients who are to receive postoperative radiation therapy. Unfortunately, the facial function achieved with dynamic and static slings is almost always inferior to that after facial nerve grafts. In this retrospective study, we compared facial nerve function in irradiated and unirradiated nerve grafts. METHODS AND MATERIALS The medical records of 818 patients with neoplasms involving the parotid gland who received treatment between 1974 and 1997 were reviewed, of whom 66 underwent facial nerve grafting. Fourteen patients who died or had a recurrence less than a year after their facial nerve graft were excluded. The median follow-up for the remaining 52 patients was 10.6 years. Cable nerve grafts were performed in 50 patients and direct anastomoses of the facial nerve in two. Facial nerve function was scored by means of the House-Brackmann (H-B) facial grading system. Twenty-eight of the 52 patients received postoperative radiotherapy. The median time from nerve grafting to start of radiotherapy was 5.1 weeks. The median and mean doses of radiation were 6000 and 6033 cGy, respectively, for the irradiated grafts. One patient received preoperative radiotherapy to a total dose of 5000 cGy in 25 fractions and underwent surgery 1 month after the completion of radiotherapy. This patient was placed, by convention, in the irradiated facial nerve graft cohort. RESULTS: Potential prognostic factors for facial nerve function such as age, gender, extent of surgery at the time of nerve grafting, preoperative facial nerve palsy, duration of preoperative palsy if present, or number of previous operations in the parotid bed were relatively well balanced between irradiated and unirradiated patients. However, the irradiated graft group had a greater proportion of patients with pathologic evidence of nerve invasion (p = 0.007) and unfavorable type of nerve graft (p = 0.04). Although the irradiated graft cohort had more potentially negative prognostic factors, there was no difference in functional outcome (H-B Grade III or IV) between irradiated and unirradiated graft patients. H-B Grades III, IV, V, and VI were the best postoperative facial nerve functions achieved in 35%, 39%, 13%, and 13% of patients, respectively. The patient with preoperative radiotherapy never recovered any facial nerve function (H-B Grade VI). Median time to best facial nerve function after surgery was longer in the irradiated patients (13.1 vs. 10.8 months), but this was not statistically significant (p = 0.10). Presence of preoperative facial nerve palsy (p = 0.005), duration of preoperative palsy (p = 0.003), and age greater than 60 years at the time of grafting (p = 0. 04) were all negative prognostic factors for achieving a functional facial nerve on univariate analysis. Analysis of age as a continuous variable (p = 0.12) and pathologic evidence of nerve invasion (p = 0. 1) revealed a trend toward negative prognostic factors. Gender, number of previous operations in the parotid bed, extent of surgery at the time of nerve grafting, and type of grafting procedure were not significant prognostic factors. Whether radiotherapy was delivered less than 6 weeks after nerve grafting or more than 6 weeks had no impact on achievement of a functional facial nerve. CONCLUSION: Negative prognostic factors for achieving a functional facial nerve in our series include the presence of preoperative facial nerve palsy, duration of preoperative palsy, and age greater than 60 years. Radiotherapy was not a negative prognostic factor. Comparing irradiated and unirradiated grafts revealed no difference in best facial nerve function achieved, despite the presence of a greater proportion of negative prognostic factors in

Preoperative facial muscle imaging predicts final facial function after facial nerve grafting.

BACKGROUND AND PURPOSE: Our goal was to determine whether preoperative MR imaging of facial muscles predicts facial function after facial nerve grafting. METHODS: A retrospective review of all patients undergoing facial nerve grafting between 1997 and 2001 revealed 26 patients. Twelve of the patients had adequate preoperative MR images available for review and had undergone clinical follow-up for at least 12 months. Eight had malignant parotid tumors, and four had benign skull base or parotid tumors. Preoperative facial muscle MR imaging appearance was categorized as symmetrical or asymmetrical. The asymmetrical images were further classified into mild or pronounced asymmetry. Preoperative facial function was classified by using the House-Brackmann scale. Postoperative function was graded with the May scale. RESULTS: Four patients had symmetrical facial muscles shown by preoperative MR imaging, three had mild asymmetry, and five had pronounced asymmetry. No or mild asymmetry had an 86% positive predictive value for good to excellent functional outcome. Eighty percent of patients with pronounced asymmetry experienced poor functional outcomes. Six of eight patients with malignant and perineural tumors at surgery had asymmetrical facial muscles revealed by preoperative MR imaging studies. CONCLUSION: Symmetrical or mildly asymmetrical facial muscles are predictive of good facial function after nerve grafting. Pronounced asymmetry of facial muscles on MR images is predictive of poor facial function after grafting. Asymmetric facial muscles on preoperative MR images are associated with perineural tumor spread in patients with malignant disease.