com
Review
Management of advanced parotid cancer. A systematic review
J.-P. Jeannon*, F. Calman, M. Gleeson, M. McGurk, P. Morgan,
M. O’Connell, E. Odell, R. Simo
Department of Head & Neck Oncology, Guy’s, King’s & St Thomas’ Cancer Centre, Guy’s & St Thomas’ NHS Foundation Trust,
3rd Floor, Thomas Guy House, London SE1 9RT, UK
Abstract
Background: Primary adenocarcinomas of the parotid gland are rare and account for less than 5% of all head and neck malignant
neoplasms. There is considerable variation in biological behaviour within this group; low-grade tumours exhibit slow growth rates with
minimal or no local invasion. High-grade tumours, however, show a high incidence of local recurrence and distant metastasis.
Aim: The purpose of this paper is to analyse the important prognostic indicators for this cancer.
Methods: A systematic review was performed involving 19 published studies from 1987 to 2005 which included 4631 patients. T stage,
grade of tumour, N stage and adjuvant radiotherapy on overall (5 year) survival were analysed as prognostic indicators.
Results: T stage ( p ¼ 0.041, hazard ratio 1.8 (confidence interval 1.2e2.9)), N stage ( p ¼ 0.05, hazard ratio 1.1 (0.2e1.8)), and high-grade
( p ¼ 0.001, hazard ratio 2.1 (1.5e2.7)) were associated with a significantly worse survival. The effect of adjuvant radiotherapy was to
improve overall survival: p ¼ 0.002, hazard ratio 2.9 (1.5e4.7). The mean 5 year survival for advanced high-grade parotid cancer was 35%.
Conclusion: High-grade advanced parotid cancers are associated with a poor survival. Adjuvant radiotherapy is indicated in these tumours
and this improves survival.
Ó 2008 Elsevier Ltd. All rights reserved.
0748-7983/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ejso.2008.10.009
J.-P. Jeannon et al. / EJSO 35 (2009) 908e915 909
Table 1
WHO International histopathological classification of parotid cancers
WHO 1972 WHO 1991 WHO 2005
Mucoepidermoid tumour Acinic cell carcinoma Acinic cell carcinoma
Acinic cell tumour Mucoepidermoid carcinoma Mucoepidermoid carcinoma
Adenoid cystic carcinoma Adenoid cystic carcinoma Adenoid cystic carcinoma
Adenocarcinoma Polymorphous low-grade adenocarcinoma Polymorphous low-grade adenocarcinoma
Epidermoid carcinoma Epithelial myoepithelial carcinoma Epithelial myoepithelial carcinoma
Undifferentiated carcinoma Basal cell adenocarcinoma Clear cell carcinoma
Carcinoma in pleomorphic adenoma Sebaceous carcinoma Basal cell adenocarcinoma
Papillary cystadenocarcinoma Sebaceous carcinoma
Mucinous adenocarcinoma Cystadenocarcinoma
Oncocytic carcinoma Low-grade cribriform cystadenocarcinoma
Salivary duct carcinoma Mucinous adenocarcinoma
Adenocarcinoma Oncocytic carcinoma
Myoepithelial carcinoma Salivary duct carcinoma
Carcinoma ex pleomorphic adenoma Adenocarcinoma NOS
Squamous cell carcinoma Myoepithelial carcinoma
Small cell carcinoma Carcinoma ex pleomorphic adenoma
Undiffererniated carcinoma Carcinosarcoma
Other carcinomas Metastasising pleomorphic adenoma
Squamous cell carcinoma
Small cell carcinoma
Large cell carcinoma
Lymphoepithelial carcinoma
Sialoblastoma
Materials and methods high-grade and advanced cases in such series was too small
for meaningful results.
Literature review
Statistical analysis
A literature search was used using Medline, Embase and
the Cochrane databases. The following key search words The following factors were extracted from studies and
were used: ‘parotid’, ‘cancer’, ‘advanced’ and ‘survival’. analysed: T stage, grade of tumour, N stage and adjuvant
radiotherapy on overall (5 year) survival.
Study design
Survival for stage T1 and T2 tumours was compared to
T3 and T4. High-grade tumours according to the Word
This is a systematic review of published cases of pri-
Health Organisation classification (1991) were compared
mary adenocarcinomas of the parotid gland. The analysis
to low-grade ones. Survival of node positive (Nþ) were
involved 19 published studies from 1987 to 2005 which
compared to node negative (N0) patients. Survival of
involved 4631 patients.2e20
patients that received radiotherapy was compared to those
Selection criteria that did not. Statistical analysis utilised Stats Direct soft-
ware. 2 2 tables were constructed in order to analyse
Only primary adenocarcinomas of the parotid gland
were included in the study. All stages and grades of Table 2
carcinoma were included. Benign parotid tumours were ex- American Joint Committee on Cancer (AJCC)/UICC staging by TNM
cluded. Secondary squamous carcinomas and lymphomas classification1
were excluded. Adenoid cystic carcinomas were also Primary tumour (T)
excluded as these cancers can have a prolonged and unpre- TX Primary tumour cannot be assessed
dictable natural history. Sub-mandibular, sublingual and T0 No evidence of primary tumour
minor salivary tumours were also excluded. Data extrac- T1 Tumour 2 cm or less in greatest dimension without
tion was performed and only T stage, N stage, effect of extraparenchymal extension
T2 Tumour more than 2 cm but not more than 4 cm in
adjuvant radiotherapy and high-grade parotid cancers
greatest dimension without extraparenchymal extension
were analysed. T3 Tumour having extraparenchymal extension without
As the histopathological categorisation and grading seventh nerve involvement and/or more than 4 cm but
system has changed over the last two decades, only papers not more than 6 cm in greatest dimension
published in the last 20 years were included. Series report- T4 Tumour invades base of skull, seventh nerve, and/or
exceeds 6 cm in greatest dimension
ing less than 100 cases were also excluded as the number of
910 J.-P. Jeannon et al. / EJSO 35 (2009) 908e915
Table 3 odds ratios and 95% confidence intervals (CI). Forest plots
Overall survival (OS; 5 years) for each study (n ¼ 4631, 19 papers) graphs were used to display results. Sensitivity analysis was
Study n 5 year OS undertaken. Bias assessment plots were calculated and tests
Bhattacharyya and Fried, 20053 651 45% of heterogeneity were performed. We used fixed and ran-
Mendenhall et al., 20044 224 44% dom effects models to assess heterogeneity across studies.
Luukkaa et al., 20045 237 23%
Terhaard et al., 20056 498 35%
Spiro et al., 19897 463 25% Results
Renehen et al., 19998 103 45%
Goode et al., 199810 234 37%
Lima et al., 20059 126 36% Initial literature search revealed 379 possible papers.
Pedersen et al., 199211 110 28.5% However, using the selection criteria quoted in the method-
Boahene et al., 200412 128 65% ology, only 19 studies were found to be suitable for analy-
Magnano et al., 199913 126 34% sis. The analysis involved 19 published studies from 1987
Charabi et al., 200014 494 28%
Kirkbride et al., 200115 184 40%
to 2005 which involved 4631 patients (Table 3).2e20
Harbo et al., 200216 152 15%
Pohar et al., 200517 163 24%
Prognostic indicators
Carinci et al., 200218 134 33%
Ball et al., 199519 271 28%
Frankenthaler et al., 199120 178 34% T stage
Dorairajan et al., 200424 155 22% Overall, T stage was seen to be a significant factor in
determining overall survival: p ¼ 0.041, hazard ratio 1.8
(CI 1.2e2.9) (Figure 1).
Figure 3. Forest plot effect of Nþ status on overall survival in advanced parotid cancer.
not specify standardised diagnostic criteria or categories. Established practice for managing advanced high-grade
However, a number of factors render the study group parotid cancers involves combined modality treatment.
relatively homogeneous. Adenoid cystic carcinoma is ex- Wide surgical excision including total conservative paroti-
cluded on grounds of its protracted clinical course. This dectomy is the procedure of choice.
is the only common high-grade carcinoma with a distinct High-grade cancers are more likely to metastasise
growth pattern and its diagnostic criteria have not changed locally and relapse with distant metastasis compared to
during the review period. Reclassification from the first to low-grade tumours.
the second WHO system produced little change in the rel- Loco-regional metastasis to the neck is also associated
ative proportion of high-grade carcinomas. The main con- with worsening survival and neck dissection is mandatory
sequence of the 1991 classification was to subtype within for patients with clinically involved nodes.
the low-grade or high-grade group. A few low-grade le- The role of elective neck dissection for N0 disease
sions were moved from benign to malignant or vice versa. was not formally assessed in this systematic review.
Rare entities, for which diagnosis is likely to be most con- The majority of studies show >20% incidence of occult
tentious, were excluded from the largest study.3 For the cervical lymph node metastasis in high-grade and ad-
purposes of this analysis, high-grade carcinomas can be vanced parotid cancers. The risk of nodal relapse in
considered to have been accurately identified throughout patients with clinically node-negative necks is highest
the study period on the basis of aggressive growth pattern, in patients with locally advanced or high-grade malignan-
cytological pleomorphism, mitotic activity and propensity cies and is significantly reduced by prophylactic nodal
for metastasis.21,22 irradiation.23 Retrospective series comparing elective neck
The statistical tests of heterogeneity and the Forest plot dissection against observation/therapeutic neck dissection
graphs confirm that the studies show relative homogeneity. for N0 parotid cancers does show an increase in
J.-P. Jeannon et al. / EJSO 35 (2009) 908e915 913
Figure 4. Forest plot effect of adjuvant radiotherapy on overall survival for parotid cancer.
loco-regional failure rates in the observational therapeutic proximity of critical structures and the irregularity of the
groups.24,25 planning target volume, and optimal radiotherapy treatment
Therefore elective treatment of the neck in terms of neck requires either 3-D conformal radiotherapy planning (CRT)
dissection or radiotherapy is indicated. The evidence or intensity-modulated radiotherapy (IMRT). Although
regarding the superiority of type of neck dissection was high-grade cancers are more likely to metastasise locally
not assessed; however, modified radical neck dissection is and relapse with distant metastasis compared to low-grade
reported as the most frequently used. tumours, local control remains a significant problem and
reduced local control leads to decreased rates of survival,
Effect of adjuvant radiotherapy on survival particularly in patients without nodal involvement.23
The role of facial nerve sacrifice was not assessed in this
This study confirms that adjuvant radiotherapy for paper as the individual data from studies was not clear.
parotid malignancy improves overall survival. Patients Surgical management for advanced cases where the facial
with lymph node involvement, positive margins, T3eT4 nerve is functioning involves total conservative parotidec-
tumours or any high-grade malignancy all have a significant tomy (nerve sparing) with modified radical neck dissection.
risk of local relapse following surgical excision and this is If the facial nerve is involved pre-operatively, total radical
reduced by adjuvant radiotherapy.6,26,27 Improved local parotidectomy with nerve resectionereconstruction and
control has been demonstrated with increasing doses of ra- neck dissection is undertaken.28
diation, particularly in patients with positive margins28 and Advances in surgical reconstructive techniques have
doses of 60e66 Gy are recommended. The treatment of sal- made wider resections of adjacent structures such as skin,
ivary tumours with high doses is challenging owing to the mandible and temporal bone possible. Even with an
914 J.-P. Jeannon et al. / EJSO 35 (2009) 908e915
22. Tischendorf L, Herrmann PK, Luttermann T. Implications of the carcinoma of the head and neck treated with surgery and radiation.
reclassification study of tumors of the salivary glands of the head. Int J Radiat Oncol Biol Phys 1995;32(3):619–26. 15.
Deutsche Zeitschr Mund Kiefer Gesichts Chir 1991;15(1):35–7. 29. Huber PE, Debus J, Latz D, et al. Radiotherapy for advanced adenoid
23. Chen AM, Garcia J, Lee NY, Bucci MK, Eisele DW. Patterns of nodal cystic carcinoma: neutrons, photons or mixed beam? Radiother Oncol
relapse after surgery and postoperative radiation therapy for carcino- 2001;59:161–7.
mas of the major and minor salivary glands: what is the role of elective 30. Laramore GE, Krall JM, Griffin TW, et al. Neutron versus photon ir-
neck irradiation? Int J Radiat Oncol Biol Phys 2007;67:988–94. radiation for unresectable salivary gland tumors: final report of an
24. Dorairajan N, Periyasamy S, Muthayya P, Manikandan R, RTOG-MRC randomized clinical trial. Radiation Therapy Oncology
Srinivasan T, Siddharth D. Salivary gland tumors: a 10-year retrospec- Group. Medical Research Council. Int J Radiat Oncol Biol Phys
tive study of survival in relation to size, histopathological examination 1993;27:235–40.
of the tumor, and nodal status. Int Surg 2004;89(3):140–9. 31. Airoldi M, Cortesina G, Giordano C, et al. Update and perspectives on
25. Zbaren P, Schupbach J, Nuyens M. Stauffer E Elective neck dissection non-surgical treatment of salivary gland malignancies. Acta Otorhino-
versus observation in primary parotid carcinoma. Otolaryngol Head laryngol Ital 2003;23:368–76.
Neck Surg 2005;132(3):387–91. 32. Airoldi M, Fornari G, Pedani F, et al. Paclitaxel and carboplatin
26. Chen AM, Granchi PJ, Garcia J, Bucci MK, Fu KK, Eisele DW. for recurrent salivary gland malignancies. Anticancer Res 2000;20:
Local-regional recurrence after surgery without postoperative irradia- 3781–3.
tion for carcinomas of the major salivary glands: implications for 33. Gilbert J, Li Y, Pinto HA, et al. Phase II trial of taxol in salivary gland
adjuvant therapy. Int J Radiat Oncol Biol Phys 2007;67:982–7. malignancies (E1394): a trial of the Eastern Cooperative Oncology
27. Olsen KD, Lewis JE. Carcinoma ex pleomorphic adenoma: a clinico- Group. Head Neck 2006;28:197–204.
pathologic review. Head Neck 2001;23:705–12. 34. Sorensen KB, Godballe C, de Stricker K, Krogdahl A. Parotid carci-
28. Garden AS, Weber RS, Morrison WH, Ang KK, Peters LJ. The noma: expression of kit protein and epidermal growth factor receptor.
influence of positive margins and nerve invasion in adenoid cystic J Oral Pathol Med 2006;35(5):286–91.