Available online at www.sciencedirect.

com

ScienceDirect

British Journal of Oral and Maxillofacial Surgery 57 (2019) 1039–1043

Utility of neck dissection for management of carcinoma of

the parotid gland
A. Kaura a,∗,1 , R.A. Kennedy b,∗∗,1 , S. Ali a , E. Odell b , R. Simo a , J.P. Jeannon a , R. Oakley a
a Department of Otolaryngology, Head and Neck Surgery, Guy’s and St Thomas’ NHS Foundation Trust, Great Maze Pond, London, SE1 9RT
b Department of Head and Neck Histopathology, Guy’s and St Thomas’ NHS Foundation Trust, Great Maze Pond, London, SE1 9RT

Accepted 2 September 2019

Available online 26 September 2019

Abstract

To validate the use of neck dissection as part of the management of patients with parotid carcinomas, we retrospectively reviewed pathological
and clinical data from the head and neck pathology archive at Guy’s and St Thomas’ Hospital on all patients who had primary parotid carcinomas
resected between 1992 and 2014. The main outcome measure was the incidence of metastatic disease. A total of 54 of the 82 patients identified
had neck dissections. Nodal metastases were detected in 10 with high-grade, invasive carcinoma ex-pleomorphic adenomas, two with salivary
duct carcinomas, one with a high-grade adenocarcinoma not otherwise specified (NOS), one with an adenoid cystic carcinoma, and one with
a high-grade acinic cell carcinoma. No metastases were found in those with a low-grade acinic cell carcinoma, low-grade mucoepidermoid
carcinoma, epithelial-myoepithelial carcinoma, or non-invasive carcinoma ex-pleomorphic adenoma. The findings of this study support the use
of routine neck dissection for the treatment of high-grade, invasive carcinoma ex-pleomorphic adenoma, salivary duct carcinoma, high-grade
adenocarcinoma NOS, adenoid cystic carcinoma, and high-grade acinic cell carcinoma.
© 2019 Published by Elsevier Ltd on behalf of The British Association of Oral and Maxillofacial Surgeons.

Keywords: Parotid; carcinoma; neck; dissection; lymph; node; prognosis

Introduction and site within both the minor and major glands, has limited
the construction of an evidence base to support the methods
Salivary gland carcinoma is a rare disease that can arise within of treatment.
any of the three major, or the minor glands. The parotid gland The United States National Cancer Database 1998-20122
remains the most common site for newly-diagnosed cancer of has shown that lymph node metastases are associated with
the major salivary glands in England, and its incidence rose a high T stage and high grade of malignancy, and high-
from 0.4 to 0.6/100 000 people between 1994 to 2013.1 This grade cases with metastatic disease are associated with a poor
rarity, combined with the varied entities, types of behaviour, prognosis. This finding has also been supported by other US
studies.3,4 Data from UK populations, however, are limited.
The aim of this study therefore was to establish the
∗ Corresponding author at: UCL Ear Institute, 332 Grays Inn Road, Lon-
frequency with which neck dissection was undertaken for
don WC1X 8EE. parotid carcinoma within a London hospital, and to vali-
∗∗ Corresponding author.
date its use by establishing the incidence of metastasis to
E-mail address: akaura0@doctors.org.uk (A. Kaura).
1 These authors have contributed equally to this manuscript and are co- the cervical lymph nodes.
first authors.

https://doi.org/10.1016/j.bjoms.2019.09.002
0266-4356/© 2019 Published by Elsevier Ltd on behalf of The British Association of Oral and Maxillofacial Surgeons.
1040 A. Kaura, R.A. Kennedy, S. Ali et al. / British Journal of Oral and Maxillofacial Surgery 57 (2019) 1039–1043

Table 1
Pathological staging of parotid carcinomas in patients treated by neck dissection.
Type of carcinoma and grade T1N0 T1N+ T2N0 T2N+ T3N0 T3N+ T4aN0 T4aN+ T4bN0 T4bN+ Total
Acinic cell:
Low 3 – 6 – – – – – – – 9
High – – – 1† – – – – 1 – 2
Adenocarcinoma not otherwise specified:
High – 1 – – 1 – 1 – – – 3
Adenoid cystic 2 – 2 – – – 1† – – 1† 6
Basal cell adenocarcinoma 1 – – – – – – – – – 1
Carcinoma ex-pleomorphic adenoma (non-invasive) 1 – 1 – 1 – – – – – 3
Carcinoma ex-pleomorphic adenoma (invasive):
Low 1 – – – 1† – – – – – 2
High – – 4† 2 – 4†† – 4†† – – 14
Carcinosarcoma ex-pleomorphic adenoma – – – – – – – – – 1 1
Epithelial-myoepithelial 1 – 2 – – – – – – – 3
Secretory – – – – 1 – – – – – 1
Mucoepidermoid:
Low – – – – 2 – – – – – 2
Intermediate – – 2 – – – – – – – 2
High – – – – 1 – – – – – 1
Salivary duct – – 2 – – – – 1 – 1 4
Total 9 1 19 3 7 4 2 5 1 3 54
† indicates one death related to the parotid carcinoma.

Material and methods The five most common histological types of primary
parotid carcinoma were high-grade invasive carcinoma
This is a retrospective validation study for the use of neck ex-pleomorphic adenoma (n = 16), low-grade acinic cell
dissection in the management of parotid carcinoma. No inter- carcinoma (n = 14), low-grade mucoepidermoid carcinoma
ventions were made on human or animal subjects. (n = 8), adenoid cystic carcinoma (n = 7), and non-invasive
To validate the use of neck dissection in the treatment (intracapsular) carcinoma ex-pleomorphic adenoma (n = 6).
of carcinoma of the parotid gland, we searched the oral and Less common types (five patients or fewer) included high-
head and neck pathology database at Guy’s and St Thomas’ grade acinic cell, high-grade adenocarcinoma, basal cell ade-
Hospital for patients who had primary resection of parotid nocarcinoma, invasive low-grade carcinoma ex-pleomorphic
carcinomas at the head and neck unit between 1992 and 2014. adenoma, carcinosarcoma ex-pleomorphic adenoma, low-
Pathological data were retrieved from text reports and fur- grade cystadenocarcinoma, epithelial-myoepithelial, low-
ther clinical information from the Trust’s electronic patient grade salivary carcinoma not otherwise specified (NOS),
record system. One patient was excluded as it was not pos- intermediate and high-grade mucoepidermoid, low-grade
sible to establish the T stage. Intraparotid and periparotid myoepithelial, high-grade salivary duct carcinoma, and low-
lymph nodes were included with level II. Mucoepidermoid grade secretory carcinoma.
carcinomas were graded according to the system published
by Brandwein et al,5 and staging was according to the Union
for International Cancer Control (UICC) 7th edition. From Treatments
2007 patients were discussed at head and neck oncology
multidisciplinary team meetings. Fifty-four patients had neck dissections and 28 did not. Of
Statistical analysis was done with the help of GraphPad those who did not, the record of the preoperative diagno-
Prism 6.0g for Windows (GraphPad Software), and Welch’s sis was not available in one, and the results of preoperative
t tests were done to assess the significance of differences fine needle aspiration cytology were equivocal or incorrectly
between groups. benign in 21. For the remainder who did not have neck dissec-
tions, the preoperative diagnoses were acinic cell carcinoma
(n = 2), mucoepidermoid carcinoma (n = 3), and carcinoma
ex-pleomorphic adenoma (n = 1).
Forty-three of the 54 who had neck dissections had had a
Results preoperative diagnosis of carcinoma, and seven had had a pre-
operative diagnosis that was suspicious of malignancy. One
Histological types and patients’ demographics who had a preoperative diagnosis of benign pleomorphic ade-
noma had an associated enlarged lymph node which, although
The group comprised 82 patients (42 women and 40 men, benign on fine needle aspiration cytology, was excised as part
median (range) age 57 (10 – 84) years). of a neck dissection limited to level II. A preoperative diag-
 A. Kaura, R.A. Kennedy, S. Ali et al. / British Journal of Oral and Maxillofacial Surgery 57 (2019) 1039–1043 1041

nosis could not be found for the three remaining patients who or fewer nodes were involved in all but one patient who
had neck dissections. had metastases in 34 nodes. This patient was in his sixth
Nineteen patients did not have radiotherapy. In six, infor- decade and the tumour was a high-grade T4aN2b carcinoma
mation about its provision was unavailable for reasons such ex-pleomorphic adenoma.
as the treatment being continued at other centres. One patient
had had radiotherapy for a separate tonsillar squamous cell
Incidence of occult lymph node metastases
carcinoma (SCC) that had been diagnosed eight months pre-
Eight of the 16 patients with lymph node metastases had
viously. All patients not treated with radiotherapy had N0
been preoperatively diagnosed with nodal disease. Metas-
and T1, 2, or 3, low-grade carcinomas. In one with a T2N0
tases were occult in eight, all of whom had had some form of
non-invasive carcinoma ex-pleomorphic adenoma, tumour
preoperative imaging. All but one patient (who had had mag-
was present at the margin, but this area of the tumour was
netic resonance imaging and ultrasound examination) had
composed of a histologically benign element of pleomorphic
had preoperative computed tomography (CT).
adenoma. The disease-free margin in a T3N0 secretory car-
cinoma could not be established, but in all other cases it was
between 0.1 and 1 mm. No perineural invasion or lympho- Follow up
vascular spread was recorded in the patients not treated with
radiotherapy. Patients treated with radiotherapy
A total of 59 patients were treated with radiotherapy. One
Outcomes of neck dissection had radiotherapy for a separate primary carcinoma of the
head and neck, and in five the data did not show whether
Histological types, T stage, and patients’ demographics radiotherapy had been given. Thirteen patients treated with
Table 1 shows the final staging of patients who had neck dis- radiotherapy died. Nine of them had had neck dissections
section. Lymph node metastases arose in 10/14 patients with (see under neck dissections) and had died of disease. One
invasive, high-grade carcinoma ex-pleomorphic adenoma, died from an unrelated cause, and records of the cause could
the only one with carcinosarcoma ex-pleomorphic adenoma, not be found in two (see under neck dissections). One had had
and two of four with salivary duct carcinoma. Lymph node postoperative radiotherapy but had not had a neck dissection
metastases developed in one of two patients with high-grade (see under no neck dissections), and the cause of death was
acinic cell carcinoma, one of six with adenoid cystic carci- unknown.
noma, and in one of three with high-grade adenocarcinoma One who had had postoperative radiotherapy for a
NOS (Table 1). pT4bN0 high-grade acinic cell carcinoma developed pul-
When all histological types were combined for patients monary metastases (see under neck dissections). There were
who had neck dissections, the carcinomas with and without no further complications or deaths. The mean (range) follow-
lymph node metastases did not differ significantly for age or up time in this group was 44.1 (0.23-142.1) months, median
T stage (p > 0.05). 30.1.
All cases of invasive, high-grade carcinoma ex-
pleomorphic adenoma without lymph node metastases were Patients who did not have radiotherapy
T2. Those with metastases were T2, T3, or T4a. Univariate Of the 19 patients who did not have radiotherapy, follow up
analysis showed that a higher T stage was significantly asso- data were available for all but one. One patient died from an
ciated with lymph node metastases (p = 0.007). There was no unrelated adenocarcinoma of the lung (see under neck dissec-
significant difference in age between patients with metastases tions). The mean (range) follow-up time was 32.1 (1.2 - 70.2)
and those without (p > 0.05). The numbers of patients with months, median 30.1. There were no recurrences, metastases,
other types of carcinoma were too low to allow univariate or or disease-specific deaths.
multivariate analysis.

Site and number of lymph node metastases Patients who did not have neck dissections
Neck dissections included level I in 22 patients, level II in 52, The mean (range) follow up in the 28 patients who did
level III in 33, level IV in 20, and level V in 11 (information not have neck dissections was 41.7 (1.4 - 145.6) months,
regarding levels was insufficient in two who had neck dissec- median 31.2. No information was available on four. No cause
tions). All those with lymph node metastases had involved of death or follow-up information could be found on one
nodes in level II. Metastases were present at level I in three, patient, a man in the ninth decade who had had an epithelial-
II in 16, III in two, IV in two, and V in one. myoepithelial carcinoma (T2N0) treated with postoperative
The involved levels were anatomically sequential in all radiotherapy. One patient developed a recurrence at the pri-
but three cases. They were involved in levels II and IV in two mary site of a pT2N0 low-grade mucoepidermoid carcinoma.
and at levels II and V in the third. This had been incompletely excised initially and was com-
In patients with lymph node metastases, the mean (range) pletely excised 12.5 months after the primary operation. No
number of involved nodes was 6 (1-34), median 3.5. Eleven metastases or disease-specific deaths were recorded.
1042 A. Kaura, R.A. Kennedy, S. Ali et al. / British Journal of Oral and Maxillofacial Surgery 57 (2019) 1039–1043

nosarcoma ex-pleomorphic adenoma in this series also had

lymph node metastases. Although data are limited for this
rare entity, it is known to be aggressive.9 One of the two
patients with high-grade acinic cell carcinoma also had lymph
node metastases, which is in keeping with previously reported
series that showed lymph node metastases in 40% - 49% of
cases.2,10 In our series they were found in 33% of patients
with high-grade adenocarcinoma NOS. Previous series have
found slightly higher incidences ranging from 45% - 65%.2,3
Seventeen per cent of patients with adenoid cystic carci-
noma had lymph node metastases and this is also in keeping
with previous reports.2,11 It is acknowledged that distant,
Fig. 1. Kaplan–Meier graph for overall survival in patients treated by neck rather than lymph node, metastases are more typical in these
dissection for high-grade parotid carcinoma.
cases.11,12
In patients who had neck dissections there was no
Patients who had neck dissections metastatic disease in two with low-grade, and two with
Of the 54 patients who had neck dissections, three had no intermediate-grade, mucoepidermoid carcinoma, and in all
follow-up information. Of those who did, the mean (range) three with epithelial-myoepithelial carcinoma and all three
duration of follow up was 40.7 (0.8 - 167.0) months, median with non-invasive carcinoma ex-pleomorphic adenoma. On
29.6. follow up in those who did not have neck dissections, there
Thirteen patients died. One, who had not had radiotherapy were no recurrences, metastases, or related deaths in three
for the parotid carcinoma, died of an adenocarcinoma of the patients with non-invasive carcinoma ex-pleomorphic ade-
lung. The remaining deaths occurred in those who had had noma, three with epithelial-myoepithelial carcinoma, and two
postoperative radiotherapy: one from an unrelated cause, two with low-grade mucoepidermoid carcinoma. Other patients
from unknown causes (no records could be found); and nine in the group had low-grade carcinomas, but in numbers that
as a result of the parotid tumour, although biopsy confirma- were too low for useful inferences to be drawn.
tion of metastases and spread was not available (Table 1). The indolent behaviour described for non-invasive carci-
These nine all related to high-grade carcinomas with the noma ex-pleomorphic adenoma is consistent with previous
exception of a pT3N0 low-grade but invasive carcinoma ex- studies and reviews.7,13–15 The behaviour of mucoepider-
pleomorphic adenoma. Six of the nine patients had also had moid carcinoma and epithelial-myoepithelial carcinoma is
lymph node metastases. The disease-specific deaths occurred also consistent with previous studies. In an American group
after 5.9 - 79.1 months of follow up. The other complication, that included more than 20 000 patients, the incidence
pulmonary metastases from a pT4bN0 high-grade acinic cell of lymph node metastases among those with epithelial-
carcinoma, was limited to one patient. myoepithelial carcinoma was 6.4%.2 In our group the overall
Overall survival in patients who had neck dissections for incidence in patients with mucoepidermoid carcinoma was
high-grade parotid carcinomas is shown in Fig. 1. Median 20.2%. A high histological grade and high T stage were
survival was 2.11 years and five-year survival between 11% found to be predictive of lymph node metastases. Régis De
and 22%. Ten-year estimated survival was between 0 and Brito Santos et al16 reported a risk of 11.1% for low-grade,
11%. and 16.7% for intermediate-grade, mucoepidermoid carci-
noma. In a clinically N0 group, Lau et al17 reported rates of
metastases of 0% and 10% for low and intermediate-grade
Discussion mucoepidermoid carcinoma, respectively. Our data, although
limited, therefore support previous series and show a low
Key findings and comparison with other studies incidence of metastases from non-invasive carcinoma ex-
pleomorphic adenoma, epithelial-myoepithelial carcinoma,
Combining histological grades, the three most common types and low-grade mucoepidermoid carcinoma.
of carcinoma in this group were carcinoma ex-pleomorphic
adenoma and acinic cell carcinoma, followed by mucoepi- Limitations of study
dermoid carcinoma. Other studies from the UK and US have
found mucoepidermoid carcinoma to be the most common It is acknowledged that the refinement of surgical treatment
histological type of epithelial malignancy.2,6 based on histological type and grade requires accurate preop-
The occurrence of cervical lymph node metastases in 50% erative diagnosis by cytological or histological investigation,
or more of patients with invasive, high-grade carcinoma ex- which may not always be possible. The primary limitation
pleomorphic adenoma and those with salivary duct carcinoma of this study was the size of the group, together with varying
is consistent with previous series2,7,8 and therefore supports numbers of patients with each histological type. It was further
neck dissection in such cases. The single patient with carci- limited by the variation in the levels included in the neck dis-
 A. Kaura, R.A. Kennedy, S. Ali et al. / British Journal of Oral and Maxillofacial Surgery 57 (2019) 1039–1043 1043

sections, the number of lymph nodes yielded, and the varied PHE publications gateway number: 2016314. Available from URL:
and, in some cases, limited follow-up time. This was par- http://www.ncin.org.uk/publications/. (Last Accessed 16 August 2019).
2. Xiao CC, Zhan KY, White-Gilbertson SJ, et al. Predictors of nodal metas-
ticularly pertinent in the case of salivary carcinomas, which
tasis in parotid malignancies: a National Cancer Data Base study of
have a propensity for delayed local recurrence and metastasis 22,653 patients. Otolaryngol Head Neck Surg 2016;154:121–30.
(sometimes more than 30 years after initial presentation).7 3. Ali S, Palmer FL, DiLorenzo M, et al. Treatment of the neck in carcinoma
The variable behaviour of these carcinomas, and the of the parotid gland. Ann Surg Oncol 2014;21:3042–8.
complex interactions of stage, grade, and patient factors, 4. Stennert E, Kisner D, Jungehuelsing M, et al. High incidence of lymph
node metastasis in major salivary gland cancer. Arch Otolaryngol Head
emphasise the importance of preoperative imaging, biopsy,
Neck Surg 2003;129:720–3.
and discussion by the multidisciplinary team. 5. Brandwein MS, Ivanov K, Wallace DI, et al. Mucoepidermoid carci-
noma: a clinicopathologic study of 80 patients with special reference to
Clinical applicability of the study histological grading. Am J Surg Pathol 2001;25:835–45.
6. Bradley PJ, McGurk M. Incidence of salivary gland neoplasms in a
defined UK population. Br J Oral Maxillofac Surg 2013;51:399–403.
This paper summarises the 22-year experience of the treat-
7. Lewis JE, Olsen KD, Sebo TJ. Carcinoma expleomorphic adenoma:
ment of patients with primary parotid carcinomas in a tertiary pathologic analysis of 73 cases. Hum Pathol 2001;32:596–604.
centre, and provides data from the one of the largest UK 8. Jayaprakash V, Merzianu M, Warren GW, et al. Survival rates and prog-
groups. Our findings are in keeping with those of previ- nostic factors for infiltrating salivary duct carcinoma: analysis of 228
ous authors and, in this context, support the use of neck cases from the surveillance, epidemiology, and end results database. Head
Neck 2014;36:694–701.
dissection when treating patients with high-grade, invasive
9. Staffieri C, Marioni G, Ferraro SM, et al. Carcinosarcoma de novo of
carcinoma ex-pleomorphic adenoma, salivary duct carci- the parotid gland. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
noma, high-grade adenocarcinoma NOS, and those with 2007;104:e35–40.
adenoid cystic carcinoma, and high-grade acinic cell carci- 10. Chiosea SI, Griffith C, Assaad A, et al. The profile of acinic cell carcinoma
noma. Our data concerning low-grade carcinomas are more after recognition of mammary analog secretory carcinoma. Am J Surg
Pathol 2012;36:343–50.
limited and require further evaluation in a multicentre setting.
11. van Weert S, Bloemena E, van der Waal I, et al. Adenoid cystic carcinoma
of the head and neck: a single-center analysis of 105 consecutive cases
over a 30-year period. Oral Oncol 2013;49:824–9.
Conflict of interest 12. Singh FM, Mak SY, Bonington SC. Patterns of spread of head and neck
adenoid cystic carcinoma. Clin Radiol 2015;70:644–53.
13. Brandwein M, Huvos AG, Dardick I, et al. Noninvasive and minimally
We have no conflicts of interest.
invasive carcinoma ex mixed tumor: a clinicopathologic and ploidy
study of 12 patients with major salivary tumors of low (or no?) malig-
nant potential. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
Ethics statement/confirmation of patients’ permission 1996;81:655–64.
14. Weiler C, Zengel P, van der Wal JE, et al. Carcinoma ex pleomorphic
adenoma with special reference to the prognostic significance of his-
This retrospective study used anonymised patient data
tological progression: a clinicopathological investigation of 41 cases.
approved by the audit department of the hospital. Patients’ Histopathology 2011;59:741–50.
permission not applicable- no identifying information 15. Seethala RR. An update on grading of salivary gland carcinomas. Head
included. Neck Pathol 2009;3:69–77.
16. Régis De Brito Santos I, Kowalski LP, Cavalcante De Araujo V, et al. Mul-
tivariate analysis of risk factors for neck metastases in surgically treated
parotid carcinomas. Arch Otolaryngol Head Neck Surg 2001;127:56–60.
References 17. Lau VH, Aouad R, Farwell DG, et al. Patterns of nodal involvement for
clinically N0 salivary gland carcinoma: refining the role of elective neck
1. Girdler R, Odell E, Putnam G. Epidemiology and management irradiation. Head Neck 2014;36:1435–9.
of major salivary gland cancers. Public Health England; 2016.