See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/51372405

Clinical features of cystadenolymphoma

(Warthin's tumor) of the parotid gland: A
retrospective comparative study of 96 cases
ARTICLE in ORAL ONCOLOGY JULY 2006
Impact Factor: 3.03 DOI: 10.1016/j.oraloncology.2005.10.017 Source: PubMed

CITATIONS

DOWNLOADS

VIEWS

30

400

215

3 AUTHORS, INCLUDING:
Afshin Teymoortash

Jochen A Werner

Philipps University of Marburg

Philipps University of Marburg

122 PUBLICATIONS 762 CITATIONS

576 PUBLICATIONS 5,055 CITATIONS

SEE PROFILE

SEE PROFILE

Available from: Jochen A Werner

Retrieved on: 17 July 2015

Oral Oncology (2006) 42, 569 573

available at www.sciencedirect.com

journal homepage: http://intl.elsevierhealth.com/journals/oron/

Clinical features of cystadenolymphoma

(Warthins tumor) of the parotid gland:
A retrospective comparative study of 96 cases
A. Teymoortash *, Y. Krasnewicz, J.A. Werner
Department of Otolaryngology, Head and Neck Surgery, Philipps University, Deutschhaus Str. 3,
35037 Marburg, Germany
Received 13 August 2005; accepted 14 October 2005

KEYWORDS

Summary The details of the etiopathogenesis of cystadenolymphoma (Warthins tumor) are

still unclear. To explore the possible risk factors for the development of this tumor, medical
records of 81 patients with 96 Warthins tumors of the parotid glands were compared retrospectively with those of 91 patients with pleomorphic adenoma. The medical history and clinical
tumor characteristics of all patients were similar. There were no significant differences
between these two patient groups with respect to concomitant diseases, regular medications,
and preoperative laboratory findings. However, a significant male predominace of patients with
Warthins tumor could be noted (P < 0.05). The male to female ratio was 3.3:1 in patients with
Warthins tumor. Multifocal Warthins tumor were detected in five cases (6.2%), and 10 patients
(12.3%) had bilateral lesions. The odds ratio for the incidence of Warthins tumor among current smokers compared with never smokers was 8.3 (P < 0.0001). Compared with never smokers, clearly higher odds of Warthins tumor was observed in heavy smokers (more than 30
pack-years) (odds ratio = 24.1, P < 0.0001) than patients who smoked less than 30 pack-years
(odds ratio = 4.9, P < 0.0001).
c 2005 Elsevier Ltd. All rights reserved.

Warthins tumor;
Pleomorphic adenoma;
Risk factors;
Gender;
Bilaterality;
Multifocality;
Smoking

Introduction

* Corresponding author. Tel.: +49 6421 2866478; fax: +49 6421

2866367.
E-mail address: teymoort@med.uni-marburg.de (A. Teymoortash).
URL: www.ent-marburg.de (A. Teymoortash).

The average incidence rate of all salivary gland tumors is 5

per 100,000.1 According to the Hamburg Salivary Gland Registry, about 65% of those tumors are benign. Cystadenolymphoma (Warthins tumor) is ranked second in frequency
after pleomorphic adenoma of the parotid gland. Warthins
tumor accounts for about 15% of all epithelial tumors of the
parotid gland.2

1368-8375/$ - see front matter c 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.oraloncology.2005.10.017

570
In 1910 Albrecht and Arzt reported two tumors of the
upper neck region which they interpreted as confused tissue in the entodermal pharyngeal anlage, particularly that
of salivary glands, in the lymph nodes.3 After these authors
a large number of studies drew the conclusion that this tumor develops due to salivary gland heterotopia in peri- and
intraparotideal lymph nodes.4,5 Although various theories
have been put forward to explain the development of
Warthins tumor,6 only two have ultimately remained. The
first is the mentioned hypothesis of heterotopia; the second
is the theory that this tumor is an adenoma with concomitant lymphocytic infiltration. According to the latter theory,
when they are small and have a short history, Warthins tumors consist mainly of epithelial components, while when
they are large they show, in addition to their epithelial component, a lymphoid stroma. Since they are observed first,
the epithelial components are therefore thought to be the
fundamental neoplastic elements.7 This theory was disproved by a recent study performed by Honda et al.,8 who
showed that the epithelial tumor components, like the lymphocytic infiltrations,9 are polyclonal. If, however, neoplasia is defined as a monoclonal process, this kind of tumor
cannot be considered to be a true neoplasm. The almost total lack of recurrence and malignant transformation of this
tumor further supports this view.10 Multicentricity at first
excision and growth from a new focus seem to be responsible for the cases of recurrence reported in the literature.
Malignant transformation of this tumor, if it ever occurs,
is extremely rare.11
The details of the pathogenesis of Warthins tumor are
still unclear. However, because of the arguments against a
true neoplastic origin of this tumor, we favour a hypothesis
combining immunological interactions between tumor cells
and lymphocytic infiltrations with heterotopia and classify
this tumor in the group of tumor-like lesions.10 One important question still remains: Which triggers cause the tumorous changes and favour their development?
In the present study, we retrospectively analysed clinical
data from patients with Warthins tumor to explore possible
risk factors for development of this disease and for further
characterisation of this tumor.

Material and methods

A retrospective analysis of 81 patients with Warthins tumor
of the parotid gland who were diagnosed and treated between April 1998 to April 2005 was performed. For comparison, a second control population of 91 patients with
pleomorphic adenoma of the parotid gland was selected
who were treated within the same period of time. All patients with Warthins tumor and pleomorphic adenoma were
treated by either superficial or subtotal parotidectomy. The
diagnosis of all tumors including bilateral and multifocal lesions based on histopathological examinations of tumor
specimens after parotidectomy.
Tumor characteristics were reviewed and details of signs
and symptoms at presentation and duration of illness until
admission were noted and analysed. Staging of parotid gland
tumors was defined by sonography and magnetic resonance
imaging (MRI) of the parotid glands in all cases. Tumor sublocalisations, tumor size, bilaterality and multifocality were

A. Teymoortash et al.
assessed according to radiological and histopathological
findings.
The medical records of all patients were reviewed with
respect to age, gender, concomitant diseases and malignancies and regular medications. The complete preoperative
laboratory findings, including hematological studies, hepatic profiles, renal function and general metabolic functions
were studied and abnormalities of laboratory tests were
analysed. Each result was classified as normal, increased,
or decreased, according to reference normal values. Lifetime exposure to tobacco smoke was recorded and patients
were classified according to their smoking habit into current
(130, or P30 pack-years), former, or never smokers. Individual periods of smoking were summed to calculate total
cumulative duration of smoking and pack-years of smoking
(calculated as number of packs per day times the cumulative number of years smoked). Patients who had stopped
smoking up to 5 years before the diagnosis of the parotid
gland tumor were classified as former smokers and excluded
from smoking analysis. Odds ratios and 95% confidence
intervals for all patients with Warthins tumor were calculated with a logistic regression model adjusted for smoking
and compared with those patients with pleomorphic adenoma. Different categories of patients data were compared
by Chi-squared test and Fishers exact test to analyse the
statistical significance. Statistical analysis was performed
computerized with the software package SPSS (version
11.5, SPSS Inc., Chicago, USA). P-values 6 0.05 were considered to indicate statistical significance.

Results
Patient population and medical history
Warthins tumors had a significant higher incidence in the
male than in the female population (P < 0.05). The male
to female ratio for Warthins tumor and pleomorphic adenoma was 3.3:1 (62 versus 19) and 1:1.6 (35 versus 56),
respectively. The mean age at diagnosis was 57.9 years
(range, 1682) for Warthins tumor and 50.3 years (range,
1582) for pleomorphic adenoma. The average duration of
complaints at diagnosis of Warthins tumor was 29.5 months
(range, 0360 months) compared to 31 months (range, 0
360 months) for pleomorphic adenoma. The most commonly
reported symptoms at diagnosis were painless swelling in
the tumor area (92.6%) and tenderness (10.5%) for Warthins
tumor. Thirteen Warthins tumors were detected accidentally by sonography and MRI of the parotid glands (13.5%).
Similarly, the most commonly reported symptoms at diagnosis were painless swelling in the tumor area (94.6%) and tenderness (3.7%) for pleomorphic adenoma.
There were no significant differences between the Warthins tumor and the pleomorphic adenoma patients group
with respect to concomitant diseases and malignancies, regular medications, and complete preoperative laboratory
evaluations. Tables 13 depict these clinical data of both
groups.
Evaluation of smoking habits showed that 79% of patients
with Warthins tumor had a history of tobacco use; for pleomorphic adenoma, only 30.8 % of patients had a history of
tobacco use. Duration of smoking in the great majority of

Warthins tumor
Table 1

571

The most concomitant diseases in patients with Warthins tumor and pleomorphic adenoma

Concomitant disease

Number of patients with WT

Concomitant disease

Number of patients with PA

No concomitant disease
Hypertension
Diabetes mellitus type II
Coronary heart disease
COPD
Cardiac arrhythmia
Cardiac failure

19
26
12
8
7
6
6

No concomitant disease
Hypertension
Thyroid dysfunction
Cardiac arrhythmia
Coronary heart disease
Osteoporosis
Prostatic hypertrophy

35
22
20
8
4
4
4

PA = pleomorphic adenoma, WT = Warthins tumor.

Table 2

The most regular medications in patients with Warthins tumor and pleomorphic adenoma

Regular medications

Number of
patients with WT

Regular medications

Number of patients
with PA

No regular medication
Antihypertensive/coronary drugs
Anticoagulants
Antidiabetics
Diuretics
Antiasthmatics/broncholytics

27
37
10
10
9
7

No regular medication
Antihypertensive/coronary drugs
Thyroid therapeutics
Sex hormones and their inhibitors

34
24
20
11

Anticoagulants

PA = pleomorphic adenoma, WT = Warthins tumor.

Table 3

The most pathologic laboratory parameters in patients with Warthins tumor and pleomorphic adenoma

Pathologic parameters

Number of patients with WT

Pathologic parameters

Number of patients with PA

No pathologic parameter
Glukose
MCV
MCH
c-glutamyltransferase
Hematokrit
C-reactive protein
Leukocytes
Chlorid
Uric acid

9
24
16
15
14
14
13
10
9
9

No pathologic parameter
Hematocrit
Glucose
MCH
Thrombocytes
MCV
C-reactive protein
Chlorid
Protein
c-glutamyltransferase

16
18
17
10
10
9
8
7
6
6

PA = pleomorphic adenoma, WT = Warthins tumor, increased, decreased, MCV = mean corpuscular volume, MCH = mean corpusclar
hemoglobin.

Table 4

Smoking characteristics in patients with Warthins tumor and pleomorphic adenoma

Number of patients with WT

Number of patients with PA

Never smokers

17 (21)

63 (69.2)

Current smokers
Intensity of smoking in pack-years
<30 pack-years
P30 pack-years

32 (39.5)
22 (27.2)

20 (22)
4 (4.4)

Duration of smoking in years

110
1120
2130
>30

2 (2.5)
7 (8.6)
23 (28.4)
22 (27.2)

7
9
3
5

Former smokers

10 (12.3)

4 (4.4)

PA = pleomorphic adenoma, WT = Warthins tumor.

(7.7)
(9.9)
(3.3)
(5.5)

572
smokers with Warthins tumor (83.3%) were over 20 years
(Table 4). Current smokers had an odds ratio of 8.3 (95%
confidence intervall [CI], 3.8418.32; P < 0.0001) than nerver smokers for Warthins tumor compared with pleomorphic adenoma. Compared with nerver smokers, clearly
higher odds of Warthins tumor was observed in heavy smokers (more than 30 pack-years). The odds ratio for the incidence of Warthins tumor among patients who smoked less
than and over 30 pack-years compared with nerver smokers
was 4.9 (95% CI 2.799.11; P < 0.0001) and 24.1 (95% CI
7.8283.06; P < 0.0001), respectively.

Clinical tumor characteristics

Ninety-six Warthins tumors were found in 81 patients. The
rate of bilateral Warthins tumor was 12.3% (10 out of 81
patients). Eight of these tumors were detected synchronously and two were detected metachronously after 19
and 52 months. One patient with bilateral Warthins tumor
had also synchronously a pleomorphic adenoma of the
parotid gland. Multifocality could be shown in 5 out of 81
patients with Warthins tumor (6.2%). These patients had
synchronously two lesions in the same parotid gland. All synchronously found multifocal and bilateral Warthins tumors
(13 out of 96) were detected accidentally by sonography and
MRI of the parotid glands. One patient with pleomorphic
adenoma developed a metachronous pleomorphic adenoma
of the contralateral parotid gland after 108 months. Salivary
gland malignancies in association with Warthins tumor
could not be detected. However, in one patient a pleomorphic adenoma and an acinus cell carcinoma were found synchronously in the same parotid gland.
The average size of Warthins tumors at diagnosis, as
determined at preoperative sonography, was 25.6
9.9 mm (range, 950 mm) in greatest diameter. Similarly,
the average maximum diameter of pleomorphic adenoma
was 21.9 8.2 (range, 1850 mm). All Warthins tumors except for two were located in the lower pole of the superfacial lobe of the parotid gland (96.8%). The great majority of
pleomorphic adenoma were also found in the same sublocation (83.7%).

Discussion
Warthins tumors most commonly present as an asymptomatic, slowly growing mass usually affecting men in the 5th
and 6th decade. The male to female ratio ranges from
2.6:1 to 10:1.6 In the present study the male to female ratio
was 3.3:1. The comparatively significantly greater tumor
incidence in men might indicate a hormone dependence of
this disease. The salivary glands are not considered as the
classical target organs of steroid hormones. The significance
of sex hormones is not clear with regard to their impact on
the salivary glands and their diseases. Only very few studies
have dealt with this question. In some malignant salivary
gland diseases and even in Warthins tumor progesterone
receptors have been found.12 The evidence of progesterone
receptor in Warthins tumor may implicate a potential role
of endocrine factors in the development of this tumor which
might explain the particular predominance of the male sex
regarding this disease.

A. Teymoortash et al.
Warthins tumor is the most common bilateral and multifocal parotid neoplasm. In about 410% of Warthins tumors, there is bilateral tumor development, which is
commonly metachronous. In about 4% of the cases multiple
Warthins tumors may be observed in one parotid gland.2,13
In our study 12.3% of patients with Warthins tumor had
bilateral lesions and multifocal tumors were detected in
6.2% of cases. The relatively high incidence of bilaterality
and multifocality of Warthins tumor of the present study
might be based on the preoperative staging of both parotid
glands by sonography and MRI in all cases as a matter of routine in our department. Bilaterality and multicentric nature
of Warthins tumors can be explained by the hypothesis of
heterotopia in the pathogenesis of these tumors. The main
support for this hypothesis is the detection of salivary duct
inclusions in lymphoid tissue in foetuses and infants. During
the embryogenesis of the parotid gland, epithelial cells
from the oral mucosa happen to penetrate into lymphocyte-rich tissue. The late encapsulation of the parotid gland
explains the occurrence of intraparotideal lymph nodes and
heterotopic salivary gland remnants entrapped in the parotideal lymph nodes. According to this theory, Warthins tumors have their origin in these epithelial inclusions.
Warthins tumors may occur simultaneously with pleomorphic adenomas, various types of carcinoma and malignant lymphomas.14 In the present study, other salivary
gland tumors and malignancies in association with Warthins
tumour could not be detected. Some studies discussed the
importance of immunological reactions during the formation of Warthins tumor. Due to the morphological similarities between Warthins tumor and Hashimoto thyroiditis,
Allegra15 first established the hypothesis that Warthins tumor originates on the basis of an immune reaction of delayed hypersensitivity type. A retrospective study of
possible association of Warthins tumor with autoimmune
pathologies revealed a higher incidence of autoimmune disorders in Warthins tumor patients.16 In the present study no
significant difference between the Warthins tumor and the
pleomorphic adenoma patients group with respect to concomitant diseases could be found.
To explore possible risk factors for development of
Warthins tumor we analysed for the first time the clinical
records of these patients with respect to regular medications and abnormalities of preoperative laboratory tests.
There were no significant differences in pathologic blood
parameters and patients regular medications compared
with patient with pleomorphic adenoma. However, the
great majority of patients with Warthins tumor had a history of over 20 years of smoking. The odds ratio of Warthins
tumor for current smokers compared with never smokers in
this study was 8.3. Compared with nerver smokers, clearly
higher odds of Warthins tumor was observed in heavy smokers (more than 30 pack-years) (odds ratio = 24.1) than
patients who smoked less than 30 pack-years (odds ratio =
4.9). Several studies showed that a significant number of
patients suffering from Warthins tumor are smokers, in
contrast to patients with other salivary gland tumors.17,18
Smoking was discussed as an important etiological factor.
Warthins tumor consists of oncocytic cells containing
numerous mitochondria frequently showing structural
abnormalities and reduced metabolic function. The oncocytic transformation of single striated epithelial cells

Warthins tumor
represents a process within the salivary glands that is frequently observed in advanced age. This fact seems to correlate with age-related metabolic deficiencies. Smoking can
lead to damage to mitochondrial DNA due to the development of numerous reactive oxygen species.19 In this context
a high rate of deleted mitochondrial DNA has been detected
in the oncocytic cells of Warthins tumor.20
In conclusion, comparing the clinical data of Warthins
tumor with pleomorphic adenoma as the most common salivary gland tumors, there were differences in the gender of
patients, tumor bilaterality and multifocality as wells as
smoking habits of these patients ascertainable. Other analysed clinical characteristics of these tumors were similar
without significant differences.

References
1. Pinkston JA, Cole P. Incidence rates of salivary gland tumors:
results from a population-based study. Otolaryngol Head Neck
Surg 1999;120:83440.
2. Seifert G. Warthin Tumoren (Adenolymphome). In: Doerr W,
Seifert G, Uehlinger E, editors. Oralpathologie I, Pathologie
sen. Berlin: Springer; 1996. p. 44063.
der Speicheldru
3. Albrecht H, Arzt L. Beitra
ge zur Frage der Gewebsverirrung. I.
Papilla
re Cystadenome in Lymphdru
sen. Frankfurt Z Pathol
1910;4:4769.
re Zystadenolymphome. Die Speicheldru
sen
4. Rauch S. Papilla
des Menschen. Stuttgart: Thieme; 1959., p. 36873.
5. Thompson AS, Bryant HC. Histogenesis of papillary cystadenoma lymphomatosum (Warthins tumor) of the parotid salivary
gland. Am J Pathol 1950;26:80729.
6. Chapnik JS. The controversy of Warthins tumor. Laryngoscope
1983;93:695716.
7. Aguirre JM, Echebarria MA, Martinez-Conde R, Rodriguez C,
Burgos JJ, Rivera JM. Warthin tumor. A new hypothesis
concerning its development. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod 1989;85:603.

573
8. Honda K, Kashima K, Daa T, Yokoyama S, Nakayama I. Clonal
analysis of the epithelial component of Warthins tumor. Hum
Pathol 2000;31:137780.
9. Takezawa K, Jackson C, Gnepp DR, King TC. Molecular
characterization of Warthin tumor. Oral Surg Oral Med Oral
Pathol Oral Radiol Endod 1989;85:56975.
10. Teymoortash A, Werner JA. Tissue that has lost its track:
Warthins tumour. Virchows Arch 2005;446:5858.
11. Foschini MP, Malvi D, Betts CM. Oncocytic carcinoma arising in
Warthin tumour. Virchows Arch 2005;446:8890.
12. Teymoortash A, Lippert BM, Werner JA. Steroid hormone
receptors in parotid gland cystadenolymphoma (Warthins
tumour). Clin Otolaryngol 2001;26:4116.
13. Maiorano E, Lo Muzio L, Favia G, Piattelli A. Warthins tumour:
a study of 78 cases with emphasis on bilaterality, multifocality
and association with other malignancies. Oral Oncol 2002;38:
3540.
14. Seifert G, Bull HG, Donath K. Histologic subclassification of the
cystadenolymphoma of the parotid gland. Analysis of 275 cases.
Virchows Arch A Pathol Anat 1980;388:1338.
15. Allegra SR. Warthins tumor: a hypersensitivity disease? Ultrastructural, light, and immunofluorescent study. Hum Pathol
1971;2:40320.
16. Gallo O, Bocciolini C. Warthins tumour associated with
autoimmune diseases and tobacco use. Acta Otolaryngol
1997;117:6237.
17. Yu GY, Liu XB, Li ZL, Peng X. Smoking and the development of
Warthins tumour of the parotid gland. Br J Oral Maxillofac
Surg 1998;36:1835.
18. Yoo GH, Eisele DW, Askin FB, Driben JS, Johns ME. Warthins
tumor: a 40 years experience at The John Hopkins Hospital.
Laryngoscope 1994;104:799803.
19. Allen JA, Coombs MM. Covalent binding of polycyclic aromatic
compounds to mitochondrial and nuclear DNA. Nature 1980;
287:2445.
20. Lewis PD, Baxter P, Paul Griffiths A, Parry JM, Skibinski DO.
Detection of damage to the mitochondrial genome in the
oncocytic cells of Warthins tumour. J Pathol 2000;191:
27481.