Introduction
Warthin's tumour, also known as papillary cystadenoma lymphomatosum, adenolymphoma and lymphomatous adenoma, is a benign growth in the parotid gland; it is the second most prevalent type of benign tumour of the parotid gland following pleomorphic adenoma.Reference Sarradin, Siegfried, Uro-Coste and Delord1 According to the World Health Organization (WHO) Classification of Head and Neck Tumours, Warthin's tumour accounts for 2–15 per cent of all primary epithelial tumours within the parotid gland.Reference Sarradin, Siegfried, Uro-Coste and Delord1
Warthin's tumours typically manifest as painless, slowly expanding, solid lumps in men aged 50–70 years, with a higher incidence in smokers. The tumours, on average, measure 2–4 cm upon detection.Reference Párraga-Linares, Aguirre-Urízar, Berini-Aytés and Gay-Escoda2
While Warthin's tumours are predominantly unilateral, the WHO records that bilateral Warthin's tumours appear in 5–14 per cent of patients.Reference Maiorano, Lo Muzio, Favia and Piattelli3 However, there is substantial divergence in the medical literature concerning the percentage of bilateral Warthin's tumours. Bilateral Warthin's tumours may occur synchronously or metachronously, more often the former.Reference Maiorano, Lo Muzio, Favia and Piattelli3–Reference Yu, Ma, Zhang, Peng, Cai and Gao5
Historically, the male-to-female ratio of Warthin's tumour has been reported to be as high as 10:1, but more recent estimates suggest a ratio of 5:1, and even as low as 2:1.Reference Tunç, Gönüldaş, Arslanhan and Kanlıkama6,Reference Stryjewska-Makuch, Kolebacz, Janik and Wolnik7
Clinical assessment, imaging and cytology inform the diagnosis of Warthin's tumour, but histology is the only definitive confirmation. Of Warthin's tumours, 0.3 per cent undergo malignant transformation, with malignant lymphoma being the typical alteration and epithelial malignancy being extraordinarily uncommon.Reference Yaranal and Umashankar8 Unless there is diagnostic ambiguity or symptomatic complications, Warthin's tumour tends to be a sole parotid lump, which is managed conservatively.
Investigations
In the UK, ultrasonography with fine needle aspiration cytology (FNAC) was previously the ‘gold standard’ for initial Warthin's tumours diagnosis.Reference Brennan, Herd, Howlett, Gibson and Oeppen9 However, ultrasound-guided core biopsy has now emerged as the principal diagnostic technique for assessing patients with salivary gland tumours, especially when malignancy is suspected. Core biopsy shows higher sensitivity and specificity than FNAC, and has shown impressive sensitivity and specificity, 92–94 per cent and 99–100 per cent respectively, for diagnosing Warthin's tumour.Reference Zahran, Alsedra, Cope and Youssef10,Reference Song, Song, Sung, Roh, Choi and Nam11 The parotid mass undergoes histological examination to confirm the diagnosis.
If there is clinical suspicion of deep lobe involvement or malignancy before surgery, a magnetic resonance imaging (MRI) scan is the preferred investigation; however, if that is contraindicated, a computed tomography (CT) scan with contrast may be considered. An MRI scan is also employed for disease surveillance.Reference Thoeny12 Challenges with all imaging methods include differentiating malignant from benign tumours and visualising the facial nerve.Reference Brennan, Herd, Howlett, Gibson and Oeppen9 Dynamic contrast-enhanced MRI, a novel form of MRI, is being explored as a diagnostic tool for Warthin's tumours, with promising results.Reference Mikaszewski, Markiet, Smugała, Stodulski, Garsta and Piątkowski13
Management
With its slow growth and low malignant transformation rate (0.3 per cent), Warthin's tumour management is predominantly conservative (watch and wait). Surgical intervention might be contemplated if there is diagnostic ambiguity, facial nerve involvement, facial pain or aesthetic concerns.Reference Yaranal and Umashankar8,Reference Reddy, Thangarajah, Castellanos-Arango and Panarese14 However, bilaterality does not favour the decision to operate. The surgical approach depends on the tumour location. For superficial lobe Warthin's tumour, a superficial parotidectomy or extra-capsular dissection is often performed. Extra-capsular dissection is typically chosen for benign parotid superficial lobe tumours measuring less than 4 cm.Reference Mehta and Nathan15 If Warthin's tumours are situated in the deep lobes, a total or near total parotidectomy may be performed, depending on tumour location. Warthin's tumours can also present at the parotid gland tail, where a partial parotidectomy may be carried out.
Aim
This study aimed to examine and evaluate the presence of contralateral Warthin's tumours among patients with positive histology results following surgical excision over the past 16 years at a high-volume head and neck unit in Kent, England. We conducted the first (to the best of our knowledge) systematic literature review on bilateral parotid Warthin's tumour, which aimed to enhance our comprehension of the prevalence, sex ratio and age at which this disease occurs.
Materials and methods
We conducted a single-centre retrospective analysis of patients diagnosed with Warthin's tumour based on histology following surgery between 1 January 2005 and 31 December 2020. A total of 290 patients (178 males and 112 females) underwent surgical excision of what was subsequently confirmed on histology to be a Warthin's tumour.
The study included patients with positive histology results following surgical excision. The data were digitally retrieved from the histology database. Contralateral Warthin's tumours were evaluated using ultrasound-guided FNAC, and MRI or CT scanning, with some doctors opting for surgical management or a histological diagnosis.
Search strategy, inclusion and exclusion criteria
The systematic review was conducted as per the standard outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (‘PRISMA’) statement. This study was registered with the International Prospective Register of Systematic Reviews (‘PROSPERO’) (registration number: CRD42022326846).
We carried out searches using the PubMed database and the Cochrane Library, to identify published studies describing patients, at any age, who were diagnosed with bilateral Warthin's tumour. The search terms (including Medical Subject Headings (MeSH terms)) and Boolean operators for PubMed were: ((“Warthin tumour”[All Fields] OR “adenolymphoma”[MeSH Terms] OR “adenolymphoma”[All Fields] OR (“Warthin”[All Fields] AND “tumor”[All Fields]) OR “Warthin tumor”[All Fields]) OR (“Bilateral”[All Fields] AND (“Warthin tumour”[All Fields] OR “adenolymphoma”[MeSH Terms] OR “adenolymphoma”[All Fields] OR (“Warthin”[All Fields] AND “tumor”[All Fields]) OR “Warthin tumor”[All Fields])) OR ((“lymphoma”[MeSH Terms] OR “lymphoma”[All Fields] OR “lymphomatous”[All Fields]) AND (“cystadenoma, papillary”[MeSH Terms] OR (“cystadenoma”[All Fields] AND “papillary”[All Fields]) OR “papillary cystadenoma”[All Fields] OR (“papillary”[All Fields] AND “cystadenomas”[All Fields]) OR “papillary cystadenomas”[All Fields])) OR “cystadenolymphoma”[All Fields]) AND “Bilateral”[All Fields]).
Screening, data extraction and quality assessment
An initial search was conducted using the abovementioned search strategy; discrepancies in terms of inclusion and exclusion decisions were discussed, and refinement of the application of these criteria was undertaken. The full screening process was independently carried out by two authors (RHKN and AG), with any conflicts being resolved by our co-first author (ZS). The full texts of potentially relevant articles for the review were retrieved and independently reviewed by RHKN and AG.
Eligibility criteria during data screening incorporated studies written in English language that described the prevalence of bilateral Warthin's tumour, and were not non-human studies, systematic reviews, literature reviews, or conference papers or abstracts. The systematic review identified 157 papers, with 14 meeting the inclusion criteria (Figure 1).
Figure 1. Identification of studies via databases and registers. WT = Warthin's tumours
All data screening and extraction processes were implemented in Microsoft™ Excel® software. In this study, we extracted: (1) first author's last name and publication year; (2) study timescale; (3) study country; (4) total sample size of the Warthin's tumour patient cohort; (5) number of bilateral Warthin's tumours; and (6) sex ratio. We used the 13-item Case Report (‘CARE’) guideline to evaluate the quality of case series studies, and the Newcastle-Ottawa Quality Assessment Scale for case–control and cohort studies.
Statistical analyses
IBM SPSS® software was used for all statistical comparisons. A p-value of less than 0.05 was considered statistically significant.
Results
Of the 290 cases, 71 had bilateral or contralateral Warthin's tumours, yielding a prevalence rate of 24.48 per cent over the 16-year period in our case series study. Among the 290 cases, there were 178 males and 112 females, resulting in a sex ratio of 1.6:1. Thirty-nine males and 32 females were diagnosed with bilateral Warthin's tumour. Table 1 illustrates that there was no significant difference between males and females in terms of unilateral and bilateral tumours, as the chi-square test p-value was 0.199 (p < 0.05). Eight of the patients underwent contralateral parotidectomies.
Table 1. Bilateral versus unilateral Warthin's tumour divided by sex
Data represent numbers of cases
Throughout the analysed study period, an increasing number of patients were diagnosed with Warthin's tumours in both unilateral and bilateral forms, as depicted in Figure 2. The average age at diagnosis for both unilateral and bilateral Warthin's tumours was higher for females compared to males in our cohort. The mean values for both groups are summarised in Table 2. Regarding age distribution, our data revealed that most patients diagnosed with Warthin's tumour were aged 60–70 years. Furthermore, Figure 3 shows that the peak diagnostic age for Warthin's tumours ranged between 50 and 80 years.
Figure 2. Trend of bilateral (‘B’) and unilateral (‘U’) tumour frequencies from 2005 to 2021.
Table 2. Average age at diagnosis of bilateral versus unilateral Warthin's tumour divided by sex
Data represent mean patient age at diagnosis, in years
Figure 3. Frequencies of ages at diagnosis.
As seen in Table 3, 14 papers from 1980 to 2019 were included in our systematic review.Reference Maiorano, Lo Muzio, Favia and Piattelli3,Reference Seifert, Bull and Donath16–Reference So, Sahovaler, Nichols, Fung, Yoo and Weir28 These papers included patients with at least one histological diagnosis of Warthin's tumours. The largest study had 1084 patients and the smallest had 25 patients. There was significant variation in the reported bilateral nature of the disease and the sex ratio. There was variation even among studies conducted within the same country.
Table 3. Studies identified from systematic review reporting bilaterality of Warthin's tumours and sex ratios
n/a = not applicable
In 2013, Xu et al.Reference Xu, Yong, Yu, Chen, Gao and Zhou23 reported a bilateral tumour rate of 40 per cent, with a male to female ratio of 7.3:1. In 2018, Xu et al.Reference Xu, Lu, Zhu, Ruan, Zhang and Yang27 performed a larger study, which showed a 0.65 per cent bilaterality and 11:1 male to female ratio. German studies conducted by Teymoortash et al.Reference Teymoortash, Krasnewicz and Werner19 and Klussmann et al.Reference Klussmann, Wittekindt, Preuss, al Attab, Schroeder and Guntinas-Lichius20 showed some consistency; however, Seifert et al.Reference Seifert, Bull and Donath16 reported half the rate of bilateral Warthin's tumours.
Discussion
The prevalence of bilateral Warthin's tumour is not currently agreed upon. The WHO Classification of Head and Neck Tumours states that the prevalence of bilateral involvement can be between 5 and 14 per cent.Reference Sarradin, Siegfried, Uro-Coste and Delord1 In our systematic review, the prevalence of reported bilateral Warthin's tumours varied widely, from as low as 0.65 per cent to 40 per cent, as seen in Table 3. The mean reported prevalence was 12.5 per cent, which aligns with the WHO classification of tumours.Reference Sarradin, Siegfried, Uro-Coste and Delord1
The primary finding from our study is the contralateral Warthin's tumour rate of 24.5 per cent. This should prompt us to discuss the bilateral nature of this disease with patients who are diagnosed with Warthin's tumours or who undergo resection of this tumour.
Additionally, our study found an increasing incidence of both unilateral and bilateral Warthin's tumours over our 16-year study period, as seen in Figure 2, which may represent an increasing incidence in the general population. The incidence is difficult to assess from this study as a large cohort are not diagnosed histologically. Interestingly, the increasing rates of diagnosis may also be the result of an ageing population. The increasing incidence of Warthin's tumours has been mentioned in earlier studies, possibly due to an increasing elderly population and improved diagnostic capability.Reference Tunç, Gönüldaş, Arslanhan and Kanlıkama6,Reference Stryjewska-Makuch, Kolebacz, Janik and Wolnik7,Reference Xu, Lu, Zhu, Ruan, Zhang and Yang27 The increase specifically in female patients may also be a result of an increase in the numbers of women smoking.Reference Ebbs and Webb17
The prevalence of bilateral Warthin's tumour varies in the literature. Three studies from Germany presented significantly different rates of bilateral Warthin's tumour prevalence, ranging from 7.5 per cent to 16.7 per cent.Reference Seifert, Bull and Donath16,Reference Teymoortash, Krasnewicz and Werner19,Reference Klussmann, Wittekindt, Preuss, al Attab, Schroeder and Guntinas-Lichius20 Franzen et al.Reference Franzen, Coordes, Franzen and Guenzel29 reported a bilaterality of 41 per cent in a German cohort, but this is not comparable to our study, as those authors specifically investigated only multiple parotid tumours and excluded all solitary tumours. Only two studies in the systematic review presented higher bilaterality than our study: Xu et al.,Reference Xu, Yong, Yu, Chen, Gao and Zhou23 who reported 40 per cent, and Patel and Morton,Reference Patel and Morton24 who reported 30 per cent. However, the sample size for both studies was small, at 25 and 41 patients respectively; therefore, their results are less likely to represent the population at large.
The significant variation reported is due to a lack of research into Warthin's tumours. This may be the result of a lack of interest in Warthin's tumours, as they are benign tumours and, in many cases, can be treated with a single operation. Even the origin of the tumour is not well understood. The current hypothesis is that it arises from proliferating salivary gland ductal cells that were entrapped in parotid lymph nodes. Therefore, Warthin's tumours are tumour-like lesions rather than adenomas.Reference Teymoortash and Werner30
The mean age at diagnosis for unilateral and bilateral Warthin's tumours in our study was 65.2 and 67.8 years old respectively. This is in accordance with the WHO classification of Head and Neck Tumours, which reports that Warthin's tumours are most often diagnosed in people aged 50–70 years.Reference Sarradin, Siegfried, Uro-Coste and Delord1 Seifert et al.Reference Seifert, Bull and Donath16 (1980) reported similar findings, with most Warthin's tumours, of 275 cases, being diagnosed in patients in their sixth decade of life. Xu et al.Reference Xu, Lu, Zhu, Ruan, Zhang and Yang27 reported, in a sample of 1084 patients, an average age at diagnosis of 56.48 years, with 82 per cent of cases being diagnosed in patients aged 50–70 years. Klussman et al.Reference Klussmann, Wittekindt, Preuss, al Attab, Schroeder and Guntinas-Lichius20 also reported a median age at diagnosis of 60 years in 185 cases. It is interesting to note that there were 5 patients diagnosed at 30–40 years of age and 20 patients diagnosed at 40–50 years of age. Although Warthin's tumour tends to affect older people, it must still be considered as a diagnosis in younger patients presenting with a parotid gland mass.
Our study revealed a male-to-female ratio of 1.6:1. The majority of the papers included suggest a higher male dominance (Table 3). The highest reported male-to-female ratios were from: Maiorano et al.Reference Maiorano, Lo Muzio, Favia and Piattelli3, who reported a 19:1 ratio in their Italian cohort; Xu et al.,Reference Xu, Lu, Zhu, Ruan, Zhang and Yang27 who reported 11:1 in their cohort of 1084 Chinese patients; and Lee et al.,Reference Lee, Yoon, Lee and Lim25 who reported 9:1 in 110 South Korean patients. In comparison, the lowest reported male-to-female ratios come from: Patel and MortonReference Patel and Morton24 with a ratio of 1.56:1 in a cohort from New Zealand; So et al.,Reference So, Sahovaler, Nichols, Fung, Yoo and Weir28 reporting 1.85:1 from their US cohort; and our study. There is considerable variation in the reported sex ratios. However, none of the previous studies showed a higher female to male ratio; this aligns with the WHO classification, which states that male sex is a risk factor for Warthin's tumour.Reference Sarradin, Siegfried, Uro-Coste and Delord1
An interesting detail regarding the sex incidence ratios is their increasing female trend. Ebbs and WebbReference Ebbs and Webb17 found that the proportion of affected females increased over their data collection period. Before 1965, all patients were male, between 1965 and 75 the ratio was 3:1, and between 1975 and 85 the ratio was 1.6:1. They purport that the reason for this was an increase in the prevalence of smoking in females. Lamelas et al.Reference Lamelas, Terry, Alfonao, York and Alfonso31 found the same in their cohort and gave a similar reason of increased smoking in females for the decreasing sex ratio. Furthermore, Franzen et al.,Reference Franzen, Coordes, Franzen and Guenzel29 in addition to their reported increase in the frequency of Warthin's tumour, identified that the incidence ratio between males and females decreased from 5:1 to 2:1 over their data collection period.
• According to World Health Organization, Warthin's tumours are bilateral in 5–14 per cent of cases; however, this systematic review shows a much wider range
• This study found contralateral Warthin's tumours in 24.5 per cent of 290 patients who underwent surgery, based on histology
• Frequency of Warthin's tumour has increased over the last 16 years
• A decreased sex ratio was also evident in our study, as has been mentioned previously, but there are large disparities in the data
The rate of contralateral tumours identified in our study has implications for clinical practice. Although Warthin's tumours rarely become malignant, they do cause distress to patients, with some patients presenting with facial swelling or pain. Xu et al.Reference Xu, Lu, Zhu, Ruan, Zhang and Yang27 recommend performing superficial bilateral parotidectomies for their patients in China; however, practice in the UK is much more conservative. Initial investigation of a Warthin's tumour should consist of a combination of clinical feature examination, imaging, and core biopsy or FNAC in both parotid glands, to ensure that bilateral tumours are not missed. There have been reports of metachronous development of bilateral Warthin's tumour. Patients should be warned of this possibility, and be given advice on what symptoms and signs may suggest metachronous development of a Warthin's tumour. Clinicians may wish to follow up patients for this possibility; however, given its rarity, this may not be a practical solution.
Limitations
The limitations of this study include its retrospective design and the limited selection of data variables collected. Given its retrospective design, the study is at risk of recall bias and selection bias. This is somewhat mitigated by the selection of consecutive patients diagnosed with Warthin's tumour. The limited data collection was intentional, for practical reasons, but it does limit more detailed comparison and analysis in relation to the existing literature. Synchronicity, smoking, surgical complications and recurrence have been measured in some other studies. Furthermore, it can be suggested that our high bilateral tumour rate is the result of data only being collected on patients who underwent surgery, thus it may not be representative of all Warthin's tumours. Future studies should address these limitations by prospectively collecting consecutive patient data with more variables, as well as collecting data on confirmed Warthin's tumours that are not operated on.
Conclusion
In our cohort of patients, who had a histological diagnosis of Warthin's tumour following surgical resection, these tumours were contralateral in 24.5 per cent of cases, and over 30 per cent of cases occurred in female patients. Our systematic review indicates that the rates of bilateral occurrence varies widely. We also provide evidence that the frequency of Warthin's tumour has increased over the last 16 years, with a decrease in the sex incidence ratio over time. We hope that the findings of this study and the systematic review can provide valuable insights for clinicians when discussing the bilateral nature and prevalence of Warthin's tumour with patients.
Competing interests
None declared
