The aim of this study was to analyse signs and symptoms present in patients with parotid gland carcinoma and to assess their prognostic value. A retrospective study of data from 131 patients who were treated surgically was performed. Evaluation of prognostic factors was possible in 109 patients who completed a minimum 5 year follow up. The most common sign and symptoms were parotid mass (96.9%), pain (40.4%), enlarged cervical lymph nodes (32.0%), facial nerve palsy (20.6%) and overlying skin infiltration (19.8%). In 20% of all cases there were no symptoms of tumour malignancy. The average duration of symptoms suggesting malignancy was 4 months. In univariate analysis, the strongest prognostic value was found for facial nerve palsy; it reduced nearly tenfold (9.7) the 5-year disease-free survival. The subsequent poor prognostic factors were: skin infiltration, enlarged cervical lymph nodes, tumour fixation and tumour size (>4 cm). Pain and the dynamics of tumour growth were not statistically significant for survival rate. Significant difference in 5-year disease free survival rate was found between the groups of patients, according to the number of symptoms suggesting malignancy. The multivariate analysis showed that only facial nerve palsy and skin infiltration were independent prognostic factors.
Salivary gland tumours are rare. They constitute 2–3% of all head and neck tumours. Most salivary gland tumours are located in the parotid gland, mainly in the superficial lobe, and malignancy is observed in about 15%. Surgery is the treatment of choice for all parotid tumours and radiotherapy is applied as supplementary treatment or in inoperable tumours.
Parotid gland carcinoma is one of the most heterogenic human tumours for its histological structure and clinical course. Despite continuous progress in diagnostics and treatment it is still a challenge for oncologists and head and neck surgeons, especially because of its relative scarcity. As parotid benign tumours are much more common, their sign and symptoms are frequently underestimated and treatment is delayed. The signs and symptoms of malignancy are usually related to the risk of advanced clinical stage and poor prognosis for survival and quality of life. Therefore it is of great clinical value to know the individual symptoms, and to understand their prognostic value. Most published papers do not describe in detail the clinical signs and symptoms in patients with parotid gland carcinoma.
Patients and methods
131 patients treated at the Department of Otolaryngology, Medical University of Gdansk, from 1978 to 2008, for primary parotid gland carcinoma, were studied. Sign and symptom analysis was carried out for all patients, but treatment results were evaluated in 109 patients, because a 5-year disease-free period was established as a criterion for therapy success. For this reason 12 patients were excluded who survived, but their follow up period was shorter than 5 years, and 10 patients were lost to follow up.
Parotid gland carcinoma was classified clinically according to the TNM classification of UICC from 2002. Reclassification was based on the data from medical records containing results of clinical examination, imaging examinations (ultrasonography, CT scan, MRI was performed in 81, 29 and 9 patients, respectively), histopathological examinations and surgical protocols. Tumour grading was based on histopathological reports.
In the 131 analysed patients, there were 70 women (53.4%) and 61 men (46.6%); the female/male ratio was 1.1:1. The patients’ age ranged from 11 to 90 years (median age was 58 years).
In this study, the second (T2) and fourth (T4) tumour stages were predominant, 43 (32.8%) and 54 (41.2%) patients, respectively. T1 stage was found in 12 (9.2%), and T3 in 22 (16.68%) patients. In 89 patients the neck lymph nodes were not enlarged (N0). N1 and N2 stages were found in 22 (16.8%) and 20 (15.3%) patients, respectively. The authors did not observed stage N3 in any patient.
There was a prevalence of high grade tumours, 89 cases (67.9%), tumours of intermediate grade were found in 8 (6.1%) patients, and low grade tumours in 34 (25.9%) patients. The most frequent tumours were adenocarcinoma not other specified, and carcinoma ex pleomorphic adenoma, 25 cases (19.0%) of each tumour, then mucoepidermoid carcinoma, 24 (18.3) cases, and adenoid cystic carcinoma, 19 (14.5%) cases. Histological types and grade of carcinomas are presented in Table 1 .
|Histological type||HG||IG||LG||Whole group|
|Carcinoma ex pleomorphic adenoma||25||25||19.0|
|Adenoid cystic carcinoma||19||19||14.5|
|Acinic cell carcinoma||11||11||8.4|
|Undifferentiated carcinoma||7||2 *||9||6.9|
|Squamous cell carcinoma||7||7||5.3|
|Salivary duct carcinoma||3||3||2.3|
|Basal cell adenocarcinoma||1||1||0.8|
|Clear cell carcinoma NOS||1||1||0.8|
|Total||89 (67.9)||8 (6.1)||34 (25.9)||131||100|
All patients underwent surgical treatment. In 55.7% postoperative radiotherapy was applied. In most patients who underwent combined treatment there was at least one indication for postoperative irradiation. Mainly it was locally advanced disease (T3/T4) and high grade of tumour malignancy, 59 and 60 patients respectively (45.0% and 45.8%), the presence of regional lymph nodes metastases (37 cases, 28.2%) and positive or uncertain surgical margin (29 cases, 22.1%). Conservative parotidectomy (with preservation of the facial nerve) was performed in 42.7% patients, semiconservative (with preservation of some facial nerve branches) in 16.0%, and radical parotidectomy (with facial nerve sacrifice) in 41.2%. In 108 patients (82.4%) the parotid surgery was followed by neck dissection, mainly selective neck dissection of levels II, III and V (67 patients, 51.1%), which were performed in patients with N0 neck clinically. In the remaining patients, radical or radical modified neck dissection was carried out.
Owing to the fact that the analysed material comes from a long period (30 years), the treatment outcomes were divided into three decades and compared (1978–1987 vs 1988–1997 vs 1998–2008). The authors did not find significant difference in overall survival and disease-free survival between these groups of patients. They did not find significant differences in the 5-, 10- and 15-year disease-free survival and overall survival rates between patients with adenoid cystic carcinoma and patients with the remaining histological types of tumours. For the whole group of patients, the 5-year overall survival rate was 57.0%, and the disease-free survival rate was 50.0%.
Treatment failure was related to local recurrence (16.5% of cases), multifocal recurrence (15.6%), and distant metastases (14.7%). Isolated nodal recurrence was found only in 3.7% of cases.
A detailed assessment of the clinical sign and symptoms was made, and to estimate their prognostic value a univariate and multivariate analysis was performed. The survival expectation was estimated by the Kaplan–Meier test. For comparison of survival expectation between two patients groups a Wilcox test by Gehan was used. For multivariate survival rate analysis a regression of proportional Cox hazard was used. Statistically significant level was estimated at p < 0.05. For assessment of the prognostic factors the effect strength was measured by odds ratio and hazard ratio.
A parotid mass was found in 127 patients (96.9%), in the remaining patients clinical examination found no tumour. In 8 cases the tumour was multifocal, in 3 patients it appeared at the site where a polymorphic adenoma had previously been removed, and in another 3 patients the tumour had a form of infiltration. Rapid tumour growth was reported by 61 patients (46.5%). Tumour fixation to the underlying tissue was found in 44 patients (33.6%). Pain was the second frequent symptom and was reported by 53 patients (40.4%). In 50 of them (38.2%) the pain was located in the tumour and face, in 10 patients (7.6%) it was earache (otalgy). Neck lymphadenopathy was found in 42 patients (32.0%). Another frequent symptom was facial nerve palsy, which was observed in 27 patients (20.6%). In 23 patients (17.5%) all the facial nerve branches were disabled. In 10 patients (7.6%) there was total facial nerve paralysis. Overlying skin infiltration was found in 26 patients (19.8%), and ulceration was found in 8 of them (6.1%). Facial paresthesia was reported by 13 patients (9.9%). The more scarce signs and symptoms were: trismus (5 patients), infiltration of the parotid duct outlet (2 patients), bleeding from the parotid duct outlet (2 patients), and glossopharyngeal nerve palsy (2 patients). Asymmetry of the pharynx (dislocation of the tonsil towards midline) was found in 4 patients, and in 1 of them the parotid tumour was not clinically present. A cheek mass was present in 3 patients. Clinical signs and symptoms are summarized in Table 2 .
|Signs and symptoms||n||%|
|Facial nerve dysfunction||27||20.6|
|Infiltration, fixation, skin ulceration||26||19.8|
|Bleeding from the parotid duct outlet||2||1.5|
|Infiltration of the parotid duct outlet||2||1.5|
|Glossopharyngeal nerve palsy||2||1.5|
The proportion of each sign and symptoms was different according to the tumour grade. In the high grade tumours the pain (47.1% vs 26.1%), facial nerve palsy (28.0% vs 4.7%), neck lymphadenopathy (34.8% vs 26.1%), skin infiltration (21.3% vs 16.6%), skin ulceration (7.8% vs 2.3%) and tumour/facial paresthesia (14.2% vs 7.8%) were more frequent, compared to intermediate and low grade carcinomas. In the group of patients with adenoid cystic carcinoma, compared to the remaining histological types, the authors found unexpectedly less facial nerve palsy (11% vs 23%), but the difference was not statistically significant ( p = 0.256).
Among all the clinical signs and symptoms those which suggested a malignant character of the parotid tumour were: facial nerve palsy, infiltration or ulceration of the overlying skin, pain, neck lymphadenopathy, rapid tumour growth, trismus, glossopharyngeal nerve palsy and bleeding from the parotid duct outlet.
In the analysed group of patients, there were no signs of malignancy in 27 (20.6%) cases, and in 34 (25.9%) cases there was only one sign of malignancy. In 40 patients (30.5%) there were at least 3 signs of malignancy. Differences between groups of tumour of high, intermediate and low grade were seen. In 28 patients (66.7%) with low or intermediate grade carcinomas and in 33 patients (37.1%) with high grade tumours there were no symptoms, or only one sign of malignancy present. At least three signs and symptoms of malignancy were found in 32 patients (35.9%) with high grade tumour and in 8 patients (19.0%) with low grade tumours. In the whole material, the average number of malignancy symptoms was 1.8 (2.1 in high grade tumours group and 1.4 in intermediate and low grade tumour group). The average duration of symptoms before seeing a doctor was 51.3 months, and the symptoms suggesting the malignancy, 4 months. If the patients with carcinoma ex pleomorphic adenoma were excluded, the average duration of symptoms was 25.4 months. The average symptom duration in the high grade tumour group and intermediate/low grade tumour group was not significantly different.
The relationship between clinical signs and symptoms and treatment results was analysed to assess their prognostic value. The symptom that had the strongest impact on treatment results was facial nerve palsy, as its presence reduced 9.7 fold the 5-year disease-free survival rate. Poor prognosis was also related to skin infiltration, neck lymphadenopathy and tumour fixation, it reduced the survival rate 4.0, 3.4, and 3.1 fold, respectively. Pain and the dynamics of tumour growth did not change the treatment results significantly. The univariate analysis proved a significant difference in survival rate according to the tumour size. The risk for disease recurrence was 2.9 fold higher in patients with tumour size above 4 cm. The relation of clinical signs and symptoms to the treatment results in univariate analysis is summarized in Table 3 . The treatment results were compared in the three groups of patients with different number of symptoms suggesting malignancy. 5-year disease-free survival rate in patients with no malignancy symptoms was observed in 15/21 cases (71.4%), and in 29/56 cases (51.8%) in patients with one or two malignancy symptoms. In patients with three or more malignancy symptoms, the 5-year disease-free survival rate was observed in 10/32 cases (31.2%). The difference between the survival rates in these groups was statistically significant. The results given above are presented in Fig. 1 .