American Journal of Otolaryngology–Head and Neck Medicine and Surgery 39 (2018) 65–70

Contents lists available at ScienceDirect

American Journal of Otolaryngology–Head and Neck

Medicine and Surgery
journal homepage: www.elsevier.com/locate/amjoto

Survival in patients with parotid gland carcinoma – Results of a

multi-center study☆
Keigo Honda a,⁎, Shinzo Tanaka b, Shogo Shinohara c, Ryo Asato d, Hisanobu Tamaki e, Toshiki Maetani f,
Ichiro Tateya g, Morimasa Kitamura g, Shinji Takebayashi c, Kazuyuki Ichimaru h, Yoshiharu Kitani i,
Yohei Kumabe j, Tsuyoshi Kojima k, Koji Ushiro l, Masanobu Mizuta g, Koichiro Yamada a, Koichi Omori g
a
Department of Otolaryngology, Japanese Red Cross Wakayama Medical Center, Wakayama, Japan
b
Department of Otolaryngology - Head & Neck Surgery, Japanese Red Cross Osaka Hospital, Osaka, Japan
c
Department of Otolaryngology - Head & Neck Surgery, Kobe City Medical Center General Hospital, Hyogo, Japan
d
Department of Otolaryngology - Head & Neck Surgery, National Hospital Organization Kyoto Medical Center, Kyoto, Japan
e
Department of Otolaryngology - Head & Neck Surgery, Kurashiki Central Hospital, Okayama, Japan
f
Department of Otolaryngology - Head and Neck Surgery, Medical Research Institute Kitano Hospital, Osaka, Japan
g
Department of Otolaryngology - Head & Neck Surgery, Kyoto University Graduate School of Medicine, Kyoto, Japan
h
Department of Otolaryngology - Head & Neck Surgery, Kokura Memorial Hospital, Fukuoka, Japan
i
Department of Otorhinolaryngology - Head & Neck surgery, Shizuoka General Hospital, Shizuoka, Japan
j
Department of Otolaryngology - Head & Neck Surgery, Hyogo Prefectural Amagasaksi General Medical Center, Hyogo, Japan
k
Department of Otolaryngology, Tenri Hospital, Nara, Japan
l
Department of Otolaryngology, Japanese Red Cross Otsu Hospital, Shiga, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Background: Parotid gland carcinoma is a rare malignancy, comprising only 1–4% of head and neck carcinomas;
Received 9 August 2017 therefore, it is difficult for a single institution to perform meaningful analysis on its clinical characteristics. The
aim of this study was to update the clinical knowledge of this rare disease by a multi-center approach.
Keywords: Methods: The study was conducted by the Kyoto University Hospital and Affiliated Facilities Head and Neck Clin-
Parotid gland carcinoma
ical Oncology Group (Kyoto-HNOG). A total of 195 patients with parotid gland carcinoma who had been surgical-
Multi-institutional retrospective study
ly treated with curative intent between 2006 and 2015 were retrospectively reviewed. Clinical results including
Clinical results
Prognostic factor
overall survival (OS), disease-free survival (DFS), disease-specific survival (DSS), local control rate (LCR), regional
control rate (RCR), and distant metastasis-free survival (DMFS) were estimated. Univariate and multivariate
analyses were performed to identify prognostic factors.
Results: The median patient age was 63 years old (range 9–93 years), and the median observation period was
39 months. The OS, DFS, DSS, LCR, RCR, and DMFS at 3 years were 85%, 74%, 89%, 92%, 88%, and 87%, respectively.
Univariate analysis showed age over 74, T4, N+, preoperative facial palsy, high grade histology, perineural inva-
sion, and vascular invasion were associated with poor OS. N+ and high grade histology were independent factors
in multivariate analysis. In subgroup analysis, postoperative radiotherapy was associated with better OS in high
risk patients.
Conclusion: Nodal metastases and high grade histology are important negative prognostic factors for OS. Postop-
erative radiotherapy is recommended in patients with advanced high grade carcinoma.
© 2017 Elsevier Inc. All rights reserved.

1. Introduction
Although the parotid gland is the site most frequently affected by
major salivary gland carcinoma, parotid gland carcinoma comprises
☆ Conflict of interest: This study did not receive any grant support. There are no conflicts
of interest to declare.
⁎ Corresponding author at: Japanese Red Cross Wakayama Medical Center, Department
of Otolaryngology, Komatsubaradori 4-20, Wakayama City, Wakayama Prefecture, Japan.
E-mail address: kegohonda@kankyo.ne.jp (K. Honda).
only 1–4% of all head and neck cancers [1–2]. In Japan, the annual inci-
dence of malignant tumor of the parotid gland was approximately 3 in
1,000,000 in 2014 [2]. A variety of histological types are observed in pa-
rotid gland carcinoma. A total of 24 histologies are listed in the 2005
World Health Organization (WHO) classification. Because of the rarity
and variety of parotid gland carcinoma, it is difficult for a single institu-
tion to conduct clinical studies with an adequate number of cases. Lon-
ger time periods allow for an increase in the total number of included
cases, but the results can be affected by changes in treatment practice
over the study period. A multi-center approach is an effective method

https://doi.org/10.1016/j.amjoto.2017.10.012
0196-0709/© 2017 Elsevier Inc. All rights reserved.
66 K. Honda et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 39 (2018) 65–70

to overcome this problem. Because a large number of patients during a Table 1

relatively short period can be enrolled, evaluation, including multivari- Patient characteristics.

ate analyses, can be performed without being affected by changes in Variable No. of patients
treatment method. Gender
The aim of this multi-center retrospective study was to update the Male 113
clinical information regarding treatment results and prognostic factors Female 82
of parotid carcinoma. Age (years)
Median (range) 59 (9–93)
Observation period (months)
2. Materials and methods Median (range) 35 (1–128)
Preoperative facial nerve palsy
The study was conducted by the Kyoto University Hospital and Affil- Partial 22
iated Facilities Head and Neck Clinical Oncology Group (Kyoto-HNOG), Total 5
Fine needle aspiration cytology
to which a total of 12 tertiary care hospitals belong. A retrospective Class I–II 53
chart review was performed on patients with parotid gland carcinoma Class III 24
who had been surgically treated with curative intent between 2006 Class IV 24
and 2015. Patients who received palliative treatment were excluded in Class V 52
Insufficient material 8
this study. The database included information regarding patient charac-
Not performed 34
teristics, treatment modality, pathologic findings, and clinical outcomes. Histology
Clinical results including overall survival (OS), disease free survival Mucoepidemoid carcinoma 48
(DFS), disease specific survival (DSS), local control rate (LCR), regional Carcinoma ex-pleomorphic adenoma 25
control rate (RCR), and distant metastasis free survival (DMFS) were es- Acinic cell carcinoma 21
Salivary duct carcinoma 21
timated using the Kaplan-Meier method. LCR, RCR, and DMFS were cal-
Adenocarcinoma, NOS 16
culated from the date of surgery to the date of detection of local, Adenoid cystic carcinoma 16
cervical, and distant recurrences, respectively. In order to identify prog- Squamous cell carcinoma 15
nostic factors, candidate factors were selected, which included age over Epithelial myoepithelial carcinoma 12
Cystadenocarcinoma 4
74, sex, T classification, N classification, preoperative facial palsy, high
Basal cell carcinoma 3
grade histology, pathological perineural invasion, pathological vascular Myoepithelial carcinoma 3
invasion, and postoperative radiotherapy (PORT). Mammary analogue secretory carcinoma 3
Three subgroup analyses were performed to evaluate the efficacy of Undifferentiated carcinoma 3
prophylactic resection of the facial nerve in patients with cT1–3 disease, Adenosquamous carcinoma 1
Carcinosarcoma 1
elective neck dissection (END) in patients with cN0 disease, and PORT in
Clear cell carcinoma 1
a high risk group. Sebaceous carcinoma 1
In statistical evaluation, pairwise comparison between plotted Unclassified carcinoma 1
curves was performed using the log-rank test. The Cox proportional Histology, grade
Low grade 81
hazard model was used in multivariate analyses. A p-value b 0.05 was
Intermediate grade 28
considered to be statistically significant. Statistical analysis was per- High grade 86
formed with EZR (Saitama Medical Center, Jichi Medical University, Disease stage
Saitama, Japan), which is a graphical user interface for R (The R Founda- I 42
tion for Statistical Computing, Vienna, Austria). This study was ap- II 45
III 33
proved by the institutional review board of every facility. Obtaining
IVA 70
patient consent was waived by the review board due to the retrospec- IVB 5
tive nature of this study. The study details were announced at every T classification
facility. T1 43
T2 48
T3 38
3. Results T4 66
N classification
3.1. Patient characteristics (Table 1) N0 149
N1 17
N2 29
A total of 217 patients with parotid gland carcinoma received initial
N3 0
treatment, of which 195 patients were treated surgically with curative Surgery
intent and included in this study. Among the included 195 patients, Parotid resection
113 were male and 82 were female. The mean age at diagnosis was Partial 91
59 years (range 9–93 years), and the median observation period was Total 104
Facial nerve resection
35 months (1–128 months). The mean observation period of patients Preserved 119
still alive at the end of the study was 39 months. Partial 29
Preoperative facial palsy was present in 27 patients, 5 of whom had Total 47
total palsy. The most common histology was mucoepidermoid carcino- Neck dissection
Yes 106
ma. High grade histology was as common as low grade histology. Nodal
No 89
metastases were infrequent, and there were no patients with N3 Postoperative radiotherapy
disease. Yes 113
The facial nerve was preserved in 119 patients and partially or totally No 82
resected in 76 patients. Therapeutic neck dissection was performed in
35 patients with clinically positive nodes (cN+), which showed patho-
logically positive nodes in 32 patients. END was performed in 54 of 160 Postoperative adjuvant radiotherapy was performed in 82 patients,
patients with clinically negative necks (cN0), and occult metastases including 59 patients with high grade carcinoma and 23 patients with
were found in 12 patients. low to intermediate grade carcinoma.
 K. Honda et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 39 (2018) 65–70
3.2. Fine needle aspiration cytology and histologic grade
Fine needle aspiration cytology (FNAC) was performed in 161 pa-
tients. Valid results were obtained in 153 patients, and insufficient
aspirated material resulted in no results in 8 patients. In patients
with low grade carcinoma, the FNAC results were most commonly
class I–II (29/62), while in patients with high grade carcinoma, the
results were most commonly class V (40/69). However, class I–II re-
sults in cytology did not exclude the possibility of high grade carci-
noma (12/53).
3.3. Clinical results
The OS, DSS, and DFS rates for the entire cohort were 85%, 100%,
and 74% at 3 years. Fig. 1a shows OS by tumor stage. The OS rate at
3 years was 93% for stage I, 100% for stage II, 93% for stage III, and
69% for stage IV. Survival was significantly worse in stage IV pa-
tients compared to stage I–III patients (p = 8.9 ∗ 10 − 8 ). Fig. 1b
shows OS by histological grade. The OS rate at 3 years was 97% for
low grade, 86% for intermediate grade, and 74% for high grade car-
cinoma. Survival was significantly worse in patients with high
grade carcinoma than in patients with low to intermediate grade
carcinoma (p = 0.00055).
3.4. Univariate and multivariate analysis of prognostic factors
The results of univariate analysis of the clinicopathological
factors affecting OS, DSS, DFS, LCR, RCR, and DMFS are listed in
Table 2. Candidate factors that showed a significant association
with clinical outcomes included advanced age, T4 local disease, pos-
itive nodal disease (N +), preoperative facial palsy, histological
grade, perineural invasion, vascular invasion, and PORT. In order to
identify independent prognostic factors, the factors that were signif-
icant in univariate analyses were tested in multivariate analysis
(Table 3). N + and high grade histology were independent negative
prognostic factors for both OS and DSS. Significant negative associa-
tions were also found between advanced age and DFS and between
N + and DMFS. The rest of the tested factors lost their significance
in multivariate analysis, and no independent prognostic factors
were identified for LCR and RCR in this study.
Fig. 1. a Overall survival (OS) of parotid carcinoma patients by stage. b Overall survival (OS) by histological grade.
67
3.5. Subgroup analyses
Three subgroup analyses were conducted to obtain additional infor-
mation useful for optimizing the treatment strategy for parotid gland
carcinoma.
Firstly, in order to evaluate the clinical impact of elective facial nerve
resection as a safety margin, an analysis was conducted in patients with
T1–3 high grade carcinoma. In 38 patients included in this sub-analysis,
the facial nerve was resected in 7 and preserved in 31. There was no sig-
nificant difference in OS between the two groups.
Secondly, in order to evaluate the efficacy of END, an analysis was
performed in cN0 patients. Of 160 patients, 54 had undergone END
and 106 had not. The comparison between the two groups showed no
significant difference in OS. The univariate analysis of clinicopathologi-
cal factors for regional recurrence in 106 patients treated without END
showed that T4 (p = 0.0081) and high grade histology (p = 0.010)
were significantly associated with regional recurrence. These two fac-
tors maintained their significance in multivariate analysis (T4, hazard
ratio = 4.0, p = 0.037; high grade histology, hazard ratio = 3.9, p =
0.036).
Lastly, to clarify whether PORT improves clinical outcomes in pa-
tients with poor prognostic factors, an analysis was performed in 55 pa-
tients with stage IV, high grade carcinoma. PORT was administered to 39
patients and not administered to 16 patients. The OS at 3-year was 76%
in patients with PORT and 26% in patients without PORT. Patients with
PORT had significantly better OS (p = 0.0064) (Fig. 2).
3.6. Recurrence
A total of 41 patients, mainly with high grade carcinoma (30/41) ex-
perienced tumor recurrence during the observation period. Recurrence
had occurred within 2 years after initial treatment in 78% (32/41) of pa-
tients. Salvage surgery was possible in 16 patients, only 4 of whom
remained disease free during the observation period. The OS rate was
33% at 3 years after recurrence, with a median survival of 22 months.
4. Discussion
Parotid gland carcinoma is a rare head and neck malignancy, whose
incidence in the Japanese population was 3.2 per 1,000,000 in 2014 [1].
Parotid cancer is often observed as a subauricular nodule. Swelling of
the cervical lymph nodes, facial paresis, or palsy can also be an initial
68 K. Honda et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 39 (2018) 65–70

Table 2
Univariate analysis of clinicopathological factors for OS, DSS, DFR, LCR, RCR and DMFS.

Variable (number) OS p DSS p DFS p LCR p RCR p DMFS p

(3 years) (3 years) (3 years) (3 years) (3 years) (3 years)

Age (years) 0.0035⁎ 0.063 0.0042⁎ 0.016⁎ 0.14 0.22

75 or higher (37) 66% 74% 57% 82% 80% 78%
b75 (158) 89% 92% 77% 94% 90% 89%
Gender 0.18 0.11 0.14 0.86 0.28 0.22
Male (113) 82% 86% 71% 92% 88% 84%
Female (82) 90% 93% 78% 91% 89% 91%
T stage b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.017⁎ 0.021⁎ 0.012⁎
T4 (66) 68% 75% 54% 85% 81% 78%
T1–3 (129) 94% 96% 84% 95% 92% 91%
N stage b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.09 0.0002⁎ b0.0001⁎
N+ (46) 60% 63% 44% 85% 71% 59%
N− (149) 93% 97% 83% 94% 93% 94%
Facial palsy 0.012⁎ 0.0043⁎ 0.0072⁎ 0.11 0.27 0.0025⁎
Yes (27) 62% 68% 53% 80% 82% 72%
No (168) 82% 92% 76% 94% 89% 89%
Grade b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.17 0.0020⁎ 0.0002⁎
High (86) 74% 99% 57% 90% 81% 73%
Low-intermediate (109) 94% 77% 86% 94% 94% 97%
Perineural invasion 0.0006⁎ 0.0006⁎ b0.0001⁎ 0.35 0.062 0.0005⁎
Yes (47) 65% 72% 46% 87% 79% 70%
No (148) 92% 94% 82% 93% 91% 92%
Vascular invasion 0.0012⁎ 0.0006⁎ b0.0001⁎ 0.26 0.021⁎ 0.0004⁎
Yes (32) 61% 68% 37% 85% 74% 65%
No (163) 90% 93% 81% 93% 91% 91%
Postoperative radiotherapy 0.47 0.22 0.19 0.77 0.83 0.029⁎
Yes (82) 85% 87% 72% 91% 89% 82%
No (113) 86% 91% 75% 93% 88% 91%

OS, overall survival; DSS, disease specific survival; DFS, disease free survival; LCR, local control rate; RCR, regional control rate; DMFS, distant metastasis free survival.
⁎ Statistically significant.

sign, although these indicate the presence of advanced disease. The re-
ported frequency of facial palsy is 12–19%, similar to what we observed
in this study (14%, 27/195) [3–5]. When the tumor is resectable, surgical
resection is the main treatment strategy in most cases of salivary carci-
noma. Radiotherapy and chemotherapy are usually provided to patients
with unresectable tumors or those who decline surgery. Partial or total
parotidectomy is the standard surgical method, although locally ad-
vanced disease sometimes requires more radical resection, including
skin or mandibular bony resection followed by flap reconstruction.
Neck dissection, either elective or therapeutic, is also performed as
needed. Surgery alone can be sufficient in patients with early stage
low grade carcinoma, whereas PORT is often administered in patients
with positive surgical margins or high grade carcinoma.
Salivary gland malignancies are rich in variation. The WHO classifies
malignant salivary gland tumors into 24 different types. In this study,
the most frequent histology was mucoepidemoid carcinoma, followed
by carcinoma ex pleomorphic adenoma and adenoid cystic carcinoma,
which was similar to the results of previous reports [6]. Preoperative tis-
sue biopsy is not commonly performed for parotid tumors. Instead,
FNAC is used to estimate the possibility of malignancy. Mallon et al. re-
ported that the sensitivity and specificity of FNAC were 52% and 92%, re-
spectively, in detecting malignant disease [7]. Schmidt et al. reported
that the diagnosis of malignancy by FNAC had lower accuracy than the
diagnosis of neoplasia [8]. In this study, FNAC was performed in 83%
(161/195) of patients, with a 5% rate of insufficient material (8/161).
Importantly, 33% (53/161) of patients were diagnosed with class I–II

Table 3
Multivariate analysis of clinicopathological factors for OS, DSS, DFR, LCR, RCR and DMFS (3-years).

Variable OS DSS DFS LCR RCR DMFS

HR p HR p HR p HR p HR p HR p
(95% CI) (95%CI) (95% CI) (95% CI) (95% CI) (95% CI)

Age N 74 2.2 0.052 − − 1.9 0.049⁎ 2.6 0.07 − − − −

(0.99–4.7) (1.0–3.6) (0.91–7.7)
T4 2.1 0.096 2.2 0.13 1.7 0.18 2.8 0.060 1.3 0.59 0.82 0.68
(0.87–5.1) (0.80–6.0) (0.80–3.4) (0.95–7.9) (0.60–3.3) (0.27–2.4)
N+ 2.9 0.011⁎ 4.4 0.0029⁎ 1.8 0.081 − − 2.6 0.081 4.7 0.0028⁎
(1.3–6.4) (1.7–12) (0.93–3.6) (0.89–7.3) (1.7–13.0)
Facial palsy 0.58 0.25 0.70 0.50 0.67 0.33 − − − − 1.2 0.75
(0.23–1.5) (0.25–2.0) (0.30–1.5) (0.39–3.8)
High grade 2.6 0.034⁎ 5.1 0.016⁎ 1.9 0.067 − − 2.4 0.11 2.1 0.21
(1.1–6.3) (1.4–19) (0.96–3.6) (0.83–7.1) (0.67–6.3)
Perineural invasion 1.5 0.37 1.5 0.47 1.7 0.15 − − − − 2.0 0.22
(0.63–3.5) (0.52–4.2) (0.83–3.4) (0.67–5.9)
Vascular invasion 0.95 0.92 0.70 0.47 1.6 0.18 − − 1.0 0.93 0.94 0.91
(0.40–2.3) (0.26–1.9) (0.79–3.4) (0.37–3.0) (0.34–2.6)
Postoperative radiotherapy − − − − − − − − − − 1.3 0.60
(0.51–3.2)

OS, overall survival; DSS, disease specific survival; DFS, disease free survival; LCR, local control rate; RCR, regional control rate; DMFS, distant metastasis free survival; HR, hazard ratio; CI,
confidence interval.
⁎ Statistically significant.
 K. Honda et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 39 (2018) 65–70
Fig. 2. Impact of postoperative radiotherapy (PORT) in patients with stage IV high grade
carcinoma (n = 55).
disease by FNAC. This fact indicates that it is difficult to exclude the pos-
sibility of malignancy by FNAC. We did observe a tendency of higher
grade carcinomas to have worse FNAC results. Class V results were asso-
ciated with high grade carcinomas in most cases. Therefore, treatment
options for patients with class V tumors should include those for high
grade carcinomas, such as extended resection combined with recon-
struction, addition of END, and PORT.
Patients with parotid gland carcinoma often present with stage IV
advanced disease [9]. Reported predictive factors for cervical node me-
tastases include high grade histology and advanced T classification [7].
Consistent with the results from previous reports, stage IV was the
most common stage in this study. Patients with higher grade carcino-
mas were diagnosed with more advanced stage than patients with
lower grade carcinomas. N 60% of patients with high grade carcinomas
had stage IV disease. Additionally, the presence of clinically positive
nodes indicated that the tumor was likely to be high grade (31/35
patients).
Various clinical and histological factors have been identified as prog-
nostic factors for parotid gland carcinoma. Reported predictive factors
for OS include histological type, grade, stage, T classification, N classifi-
cation, extracapsular spread, status of surgical margins, skin involve-
ment, gender, and advanced age [1,3,6,10–12]. Reported predictive
factors for recurrence include T classification [4,12,13], N classification
[4], and facial palsy [12]. In this study, most of these factors showed a
significant association with various clinical outcomes in univariate anal-
ysis. In the evaluation of PORT, clinical results were better in patients
who did not receive radiotherapy. This seemingly odd result may be
partly explained by selection bias, in which patients with more ad-
vanced disease were selected for postoperative radiation. In multivari-
ate analyses, only nodal metastases and histological grade remained
statistically significant as independent prognostic factors for OS. Ad-
vanced age was also identified as an independent negative prognostic
factor for DFS, which we believe is at least partly because of the high
rate of death unrelated to parotid carcinoma.
It is known that surgery alone is superior to radiotherapy alone as
primary treatment for parotid gland carcinomas [14]. Additionally, it
has been reported that the facial nerve tends to run closer to the prima-
ry tumor in malignant tumors than in benign tumors [15]. Therefore,
69
resection of the facial nerve is usually recommended if en-bloc resection
of the tumor is required to obtain negative surgical margins. The result
of sub-analysis in this study indicated that elective resection of the facial
nerve in patients with high grade carcinoma did not improve prognosis.
The reported incidence of occult nodal metastases ranges from 10 to
33% [6,16]. There is debate on whether END should be performed in pa-
tients with a clinically negative (cN0) neck. Some researchers recom-
mend END for all patients with parotid carcinomas, while others do
not recommend END for patients with low grade carcinomas [10,11,
17]. Herman et al. reported that there is no benefit of END when the pa-
tient receives PORT [18]. In this study, the survival benefit of END in cN0
patients was not proven. However, it was shown that T4 local disease
and high grade histology were independent risk factors for cervical re-
currence in cN0 patients. Therefore, it is reasonable to perform END in
this subgroup to improve regional control.
Several reports have shown that PORT improved locoregional con-
trol in patients with poor prognostic factors, such as advanced T classi-
fication, facial nerve infiltration, positive margins, perineural invasion,
nodal metastasis, and extracapsular spread [10,12,18]. Armstrong et al.
reported that PORT significantly improves outcome for patients with
stage III and IV disease and patients with lymph node metastases [14],
while Zbären et al. suggest that PORT should be considered even in
T2–4 low-grade carcinomas [19]. In this study cohort, PORT was per-
formed mainly in patients with high grade carcinomas or with locally
advanced low to intermediate grade carcinomas. Although PORT
showed no effectiveness in the cohort as a whole, patients with stage
IV high grade carcinomas who underwent PORT had improved OS.
Based on this result, PORT is recommended for patients with stage IV
high grade carcinoma.
Most recurrences occurred within 2 years after treatment. Nodal
metastasis was identified as an independent risk factor for distant me-
tastases, while no risk factors were identified for local or regional recur-
rence in multivariate analyses. However, the disease-free survival
period of parotid gland carcinoma differs by histology and stage. Like
adenoid cystic carcinoma, some types of carcinoma recur many years
after initial treatment. Therefore, with its relatively short observation
period for low grade parotid carcinomas, this study could not deduce
solid conclusions about recurrence over a long time course.
5. Conclusion
Several independent prognostic factors were identified by multivar-
iate analyses in this retrospective multi-institutional study. Age over 74
was a negative prognostic factor for DFS. Cervical lymph node metasta-
sis was a negative prognostic factor for OS, DSS, and DMFS. Histological
high grade was a negative prognostic factor for OS and DFS. No prognos-
tic factors were identified for local or regional control. The results of this
study support the preservation of the facial nerve whenever possible.
PORT is recommended in patients with stage IV, high grade carcinoma
to obtain better outcomes. The limitation of this study was its relatively
short length of mean observation period. Further study, hopefully a pro-
spective one, is needed to confirm the findings gained by this study.
Conflict of interest
The authors have no conflicts of interest to declare.
References
[1] Kane WJ, MacCafrey TV, Olsen KD, et al. Primary parotid malignancies: a clinical and
pathologic review. Arch Otolaryngol Head Neck Surg 1991;117:307–15.
[2] Report of Head and Neck Cancer Registry of Japan Clinical Statics of Registered
Patients, 2014. Jpn J Head Neck Cancer 2014;42 (suppl.).
[3] Spiro IJ, Wang CC, Montgomery WW. Carcinoma of the parotid gland. Analysis of
treatment results and patterns of failure after combined surgery and radiation ther-
apy. Cancer 1993;71:2699–705.
[4] Vander Poorten VL, Balm AJ, Hilgers FJ, et al. The development of a prognostic score
for patients with parotid carcinoma. Cancer 1999;85:2057–67.
70 K. Honda et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery 39 (2018) 65–70
[5] Johns ME. Parotid cancer: a rational basis for treatment. Head Neck Surg 1980;3:
132–41.
[6] Xiao CC, Zhan KY, White-Gilbertson SJ, et al. Predictors of nodal metastasis in parotid
malignancies: a national cancer data base study of 22,653 patients. Otolaryngol Head
Neck Surg 2016;154:121–30.
[7] Mallon DH, Kostalas M, MacPherson FJ, et al. The diagnostic value of fine needle as-
piration in parotid lumps. Ann R Coll Surg Engl 2013;95:258–62.
[8] Schmidt RL, Hall BJ, Wilson AR, et al. A systematic review and meta-analysis of the
diagnostic accuracy of fine-needle aspiration cytology for parotid gland lesions.
Am J Clin Pathol 2011;136:45–59.
[9] Mendenhall WM, Morris CG, Amdur RJ, et al. Radiotherapy alone or combined with
surgery for salivary gland carcinoma. Cancer 2005;103:2544–50.
[10] Walvekar RR, Andrade Filho PA, Seethala RR, et al. Clinicopathologic features are
stronger prognostic factors than histology or grade in risk stratification of primary
parotid malignancies. Head Neck 2011;33:225–31.
[11] Klussmann JP, Ponert T, Mueller RP, et al. Patterns of lymph node spread and its in-
fluence on outcome in resectable parotid cancer. Eur J Surg Oncol 2008;34:932–7.
[12] Terhaard CH, Lubsen H, Van der Tweel I, et al. Salivary gland carcinoma: indepen-
dent prognostic factors for locoregional control, distant metastases, and overall
survival: results of the Dutch head and neck oncology cooperative group. Head
Neck 2004;26:681–92.
[13] Frankenthaler RA, Luna MA, Lee SS, et al. Prognostic variables in parotid gland can-
cer. Arch Otolaryngol Head Neck Surg 1991;117:1251–6.
[14] Armstrong JG, Harrison LB, Spiro RH, et al. Malignant tumors of major salivary gland
origin. A matched-pair analysis of the role of combined surgery and postoperative
radiotherapy. Arch Otolaryngol Head Neck Surg 1990;116:290–3.
[15] Domenick NA, Johnson JT. Parotid tumor size predicts proximity to the facial nerve.
Laryngoscope 2011;121:2366–70.
[16] Valstar MH, van den Brekel MW, Smeele LE. Interpretation of treatment outcome in
the clinically node-negative neck in primary parotid carcinoma: a systematic review
of the literature. Head Neck 2010;32:1402–11.
[17] Zbären P, Schüpbach J, Nuyens M, et al. Elective neck dissection versus observation
in primary parotid carcinoma. Otolaryngol Head Neck Surg 2005;132:387–91.
[18] Herman MP, Werning JW, Morris CG, et al. Elective neck management for high-
grade salivary gland carcinoma. Am J Otolaryngol 2013;34:205–8.
[19] Zbären P, Schüpbach J, Nuyens M, et al. Carcinoma of the parotid gland. Am J Surg
2003;186:57–62.