The objective of this study was to analyze the oncological and functional outcomes after the surgical treatment of parotid cancer. We reviewed 80 primary parotid carcinomas retrospectively. A superficial parotidectomy was performed in 10 patients; 27 patients underwent total parotidectomy and 43 patients underwent radical parotidectomy. A facial–facial nerve anastomosis was chosen for the facial nerve reconstruction in eight patients, while an interpositional graft was selected in 24 patients. The overall N-positive rate of pathology was 21.3%. The rate of occult metastasis was 8.1%. High-grade carcinoma and lymphovascular emboli were independent factors for nodal metastasis. The 5-year disease-free survival and overall survival rates were 79.7% and 78.8%, respectively. Preoperative facial nerve palsy and extraparenchymal invasion were the independent factors associated with poor disease-free survival. Of the 41 patients in the facial nerve preservation group, 13 (31.7%) had transient facial nerve paresis. In the facial nerve sacrifice group of 39 cases, (sub)total recovery (House–Brackmann grade I/II) occurred in 14 (35.9%), partial recovery (House–Brackmann grade III/IV) in 13 (33.3%), and no recovery (House–Brackmann grade V) in 12 (30.8%). Facial nerve palsy upon presentation and extraparenchymal invasion indicate a grave prognosis. Facial nerve function after proper reconstruction is tolerable.
Primary malignant parotid tumours are relatively rare, and their biological behaviour varies. The prognosis following treatment for these tumours has been very difficult to interpret. In addition, these tumours have a prolonged risk of loco-regional recurrence and distant metastasis, often occurring years after diagnosis. Although surgical resection followed or not by adjuvant radiotherapy has been the treatment of choice, factors such as those mentioned above complicate treatment. Radical parotidectomy is advocated for the surgical treatment of high-grade malignancies (particularly those with facial nerve invasion). Unfortunately, the aesthetic and functional deficits created by sacrificing the facial nerve can be emotionally and physically traumatizing to the patient. Therefore, when the facial nerve is sacrificed, some form of facial nerve reconstruction is imperative. As in head and neck squamous cell carcinoma, metastases in the cervical lymph nodes are a major factor in therapy and for the prognosis of parotid cancer. Despite these facts and because of the low incidence of parotid carcinoma, the controversy about neck dissection remains a common issue.
The objective of this study was to analyze the oncological and functional outcomes after the surgical treatment of parotid cancer within the practice of a single experienced surgeon.
Materials and methods
This study was reviewed and approved by the university institutional review board. The medical records of 96 patients who underwent primary surgical management (by YSR) for parotid cancer between 1997 and 2012 were reviewed retrospectively. Patients with metastasis from other sites ( n = 2), lymphoma ( n = 2), or insufficient data ( n = 12) were excluded. Hence, in this retrospective study, 80 primary parotid carcinomas were analyzed.
The study group was composed of 45 men and 35 women (mean age 51 years, range 12–89 years). The follow-up period ranged from 6 to 216 months (mean 39.8 months). The tumours were classified as stage I in 18 patients, stage II in 14 patients, stage III in five patients, stage IVa in 42 patients, and stage IVb in one patient. The TNM stages of the patients are summarized in Table 1 . Thirty-eight patients were treated with surgery only. Surgical treatment was followed by postoperative radiotherapy in 33 patients (for high-grade tumours, advanced stage (III–IV), or facial nerve invasion) and chemoradiation in nine patients (for extracapsular nodal spread, perineural invasion, vascular embolism, or positive margins).
A superficial parotidectomy was performed for small, movable superficial lumps or in cases of false-negative findings from preoperative fine-needle aspiration biopsies. Radical parotidectomy was defined as a surgical procedure aimed at the eradication of malignant tumours involving the parotid gland and adjacent structures including any facial nerve branch or the main trunk of the facial nerve ( n = 38), temporal bone ( n = 12), mandible ( n = 4), lingual or glossopharyngeal nerve ( n = 5), or parapharyngeal node resection ( n = 5). A superficial parotidectomy was performed in 10 patients (12.5%); 27 (33.8%) patients underwent total parotidectomy and 43 (53.7%) patients underwent radical parotidectomy.
A neck dissection was performed within the same operation for 60 patients (75%). Overall, 23 therapeutic neck dissections and 37 elective neck dissections (I–III in 29 patients, I–IV in three, II–III in two, and II in three) were performed. Patients were reconstructed immediately at the time of ablation using a sternocleidomastoid muscle flap ( n = 51) for contour restoration, as well as a pectoralis major myocutaneous flap ( n = 7), latissimus dorsi myocutaneous flap ( n = 2), and trapezius myocutaneous flap ( n = 1) for skin coverage, contour restoration, and mandible reconstruction.
Facial nerve function before surgical management
Nineteen patients (23.8%; 19/80) had facial nerve paralysis upon presentation. In 14 patients with no preoperative facial deficit (22.9%; 14/61), a neural infiltration was found intraoperatively.
Management of the facial nerve
Facial nerve preservation was possible in 41 (51.2%) of the 80 patients. In the other 39 (48.8%) patients, the facial nerve was sacrificed based on the preoperative and operative clinical situation of the nerve. For cases in whom the facial nerve was affected preoperatively, the branches involved were sacrificed, at the least. If the main trunk of the nerve was involved in the tumour, the peripheral branches were first identified, and after the removal of the tumour, a trunk–branches anastomosis with neural grafts was attempted. If the nerve was intact preoperatively but was intraoperatively found to be grossly involved in the tumour, the branches involved were sacrificed. In these cases, the nerve was sacrificed until the proximal and distal stump were tumour-free, as confirmed by frozen section analysis. If the nerve was intact preoperatively and was found to be in close relation with the tumour intraoperatively, an attempt was made to preserve it. In such cases, additional radiotherapy was recommended, regardless of the malignancy type.
In most situations, the immediate reconstruction of the facial nerve was the treatment of choice after such a transaction. A facial–facial nerve anastomosis was selected for eight patients with short lesions that allowed for tensionless, end-to-end sutures of the nerve stumps. An interpositional graft was selected in 24 patients who were not eligible for tensionless reconstruction. The greater auricular nerve was utilized for grafting in 14 patients, the sural nerve in six, and the cervical plexus in four. In seven patients, the facial nerve was resected without reconstruction (minor buccal branch, n = 5; revision surgery with long-term preoperative palsy, n = 1; impossible to identify facial nerve stump, n = 1) ( Table 2 ). Postoperative facial nerve function was reported using the House–Brackmann (H–B) facial grading system ( Table 3 ) after a minimum follow-up period of 6 months.
|Facial nerve reconstruction||Number|
|Preservation of the facial nerve||41 (51.2%)|
|Sacrifice of the facial nerve without reconstruction||7 (8.8%)|
|Minor buccal branch||5|
|Revision surgery with long-term preoperative palsy||1|
|Impossible to identify facial nerve stump (extensive disease)||1|
|Sacrifice of the facial nerve with reconstruction||32 (40%)|
|Greater auricular nerve||14|
|Nerve end-to-end anastomosis||8|
|II||Mild dysfunction||Slight weakness noticeable on close inspection, may have slight synkinesis||Normal symmetry and tone||Moderate to good function||Complete closure with minimum effort||Slight asymmetry|
|III||Moderate dysfunction||Obvious but not disfiguring difference between two sides. Noticeable but not severe synkinesis, contracture, or hemifacial spasm||Normal symmetry and tone||Slight to moderate movement||Complete closure with effort||Slightly weak with maximum effort|
|IV||Moderately severe dysfunction||Obvious weakness and disfiguring asymmetry||Normal symmetry and tone||None||Incomplete closure||Asymmetrical with maximum effort|
|V||Severe dysfunction||Only barely perceptible movement||Asymmetry||None||Incomplete closure||Slight movement|
|VI||Total paralysis||No movement||No movement||None||None||None|
The overall and disease-specific survival rates were determined using the Kaplan–Meier actuarial life-table method with the log rank test for statistical comparisons. Cox regression was also performed for multivariate analysis. Univariate analysis was performed using Fisher’s exact test to analyze the relationships among the clinical variables. Multivariate analysis was performed by logistic regression using SPSS statistical package software (SPSS Inc., Chicago, IL, USA). A P -value of less than 0.05 was considered significant.
Histopathology and lymph node metastases
Histopathological results are reported in Table 4 . Carcinoma ex pleomorphic adenoma was the most common type ( n = 24), followed by mucoepidermoid carcinoma ( n = 15), adenoid cystic carcinoma ( n = 11), acinic cell carcinoma ( n = 7), salivary duct carcinoma ( n = 5), and adenocarcinoma ( n = 4). The high-grade carcinomas included intermediate and high-grade mucoepidermoid carcinoma, adenocarcinoma, squamous cell carcinoma, solid type adenoid cystic carcinoma, salivary duct carcinoma, and carcinoma ex pleomorphic adenoma. High- and low-grade malignant tumours were observed in 48 (60%) and 32 (40%) cases, respectively.
|Histopathological types of tumours||Number|
|Carcinoma ex pleomorphic adenoma||24|
|Adenoid cystic carcinoma||11|
|Acinic cell carcinoma||7|
|Salivary duct carcinoma||5|
|Epithelial myoepithelial carcinoma||3|
|Squamous cell carcinoma||2|
|Malignant oncocytic carcinoma||2|
|Basal cell adenocarcinoma||1|
The overall N-positive rate was 21.3% (17 out of 80 patients) in the final pathological report. Fifteen out of 43 patients (34.9%) had nodal metastases in the radical parotidectomy group. Ipsilateral level IIa was the most common site for nodal metastasis. Ipsilateral multiple level involvement was observed in 14 (17.5%) patients. There were no cases of isolated level III, IV, or V metastasis. The rate of ipsilateral occult cases was 8.1% (three out of 37 patients). Six (7.5%) tumours were reported as having positive margins (cancer or carcinoma in situ at any resection margin). Among the clinical variables analyzed for nodal metastasis, high-grade carcinoma ( P = 0.001), pT4 stage ( P = 0.01), preoperative facial nerve palsy ( P = 0.011), and lymphovascular emboli ( P < 0.001) were correlated with nodal metastasis upon univariate analysis. Multivariate analysis revealed that high-grade carcinoma ( P = 0.043) and lymphovascular emboli ( P = 0.023) were independent factors for nodal metastasis ( Table 5 ).
|Factor||LN negative||LN positive||P -value (univariate)||P -value (multivariate)||HR||95% CI|
|High-grade carcinoma||0.001 *||0.043 *||10.056||1.080–93.641|
|pT stage||0.01 *||0.695|
|Preoperative facial nerve palsy||0.011 *||0.546|
|Lymphovascular emboli||<0.001 *||0.023 *||6.865||1.303–36.169|