Development and evaluation of a clinical-biochemical nomogram for differentiating pleomorphic adenoma and Warthin’s tumor in the parotid gland

Objectives: This study aims to construct and evaluate an integrated nomogram model that combines clinical characteristics and biochemical blood markers to enable accurate and cost-effective differentiation between PA and WT.

Methods: A retrospective analysis was carried out on patients diagnosed with PA and WT at the First Affiliated Hospital of Guangxi Medical University between 2013 and 2022. The participants were randomly allocated to the training set or validation set in a 3:1 ratio. Independent predictors were identified using the Least Absolute Shrinkage and Selection Operator (LASSO) and logistic regression. Based on these predictor variables, a nomogram model was developed. The study systematically evaluated the diagnostic performance, calibration effects, and clinical utility of the model through the application of receiver operating characteristic curves (ROC), calibration curves, decision curve analysis (DCA), and clinical impact curves (CIC).

Results: The study cohort included 249 patients with PA and 173 with WT. The participants were split into a training set of 316 and a validation set of 106. Six independent predictors were identified: sex, age, smoking history, bilateral parotid involvement, white blood cell (WBC) count, and postprandial blood glucose at 2 hours (PBG2h). The AUC of the nomogram on the training set was 0.951, and the sensitivity, specificity and accuracy were 0.91, 0.906 and 0.908, respectively; on the validation set, these metrics were 0.934, 0.918, 0.844 and 0.887, respectively. The calibration curves were in close alignment with the ideal diagonals, indicating that the model has excellent calibration performance. Significant net clinical benefit was observed in the DCA and CIC analyses when the threshold probabilities were in the range of 0.05 to 0.97 for the training set and 0.07 to 0.96 for the validation set.

Conclusion: This study presents the first nomogram model integrating clinical features and biochemical indicators for preoperative differentiation between PA and WT. The model offers an accurate, cost-effective tool to support personalized treatment decisions.

1 Introduction

Parotid gland tumors are broadly categorized into benign and malignant lesions, with benign parotid gland tumors accounting for approximately 80%, most frequently observed as pleomorphic adenoma (PA) and Warthin’s tumor (WT) (1). Although these two tumors exhibit similar clinical presentations, their biological differences significantly influence treatment strategies. Pleomorphic adenoma has a reported recurrence rate of 14.1% to 42% at 5 years (2, 3) and 4% to 31.4% at 10 years (3, 4) after surgery, and untreated cases carry a 5.9% risk of malignant transformation, necessitating radical surgical intervention (5). In contrast, Warthin’s tumor, with a malignancy rate of less than 1%, is typically managed through conservative approaches or local excision (6). Recent epidemiological studies have shown an increasing incidence of WT, with it now surpassing PA as the most common benign parotid tumor in certain populations. For example, data from Germany indicate that WT accounted for 42.1% of benign tumors over a 42-year period (7) and increased to 53.4% in a more recent decade-long study (8). This rise may be attributed to factors such as aging populations, chronic inflammation, smoking, obesity, metabolic syndrome, and improvements in imaging techniques (7, 9–11). Thus, accurately differentiating PA from WT before surgery is critical for selecting the appropriate treatment plan and optimizing patient outcomes.

While imaging techniques such as ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and fine-needle aspiration cytology (FNAC) are often employed for preoperative diagnosis, these methods have limitations. The diagnostic accuracy of FNAC ranges from 72% to 86%, with a higher false-negative rate for WT (12, 13), and may cause complications such as pain, bleeding, infection, or transient facial paralysis (14, 15). Imaging techniques provide basic tumor information but are limited by high costs, technical requirements, and potential radiation risks, particularly in patients with renal insufficiency (16–18). Notably, WT frequently presents with cystic formations on imaging—especially bilateral or multifocal cases—characterized by smooth walls, papillary projections, cholesterol crystals (hyperintense on T1-weighted MRI), and rapid washout on dynamic contrast enhancement (19, 20). However, these features are not pathognomonic and can overlap with other cystic salivary gland lesions. Although radiomics models show promise in differentiating PA from WT, their application remains limited due to small sample sizes, inconsistent feature selection, and the lack of external validation (21).

Research has shown that WT is more common in elderly male smokers (22), and immune-inflammatory markers such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammation index (SII) can assist in distinguishing benign from malignant parotid tumors (23, 24). Additionally, the development of WT may be associated with metabolic syndrome and obesity (9). However, predictive models based on these markers remain underdeveloped. To address this gap, the present study utilizes a 10-year retrospective cohort from a tertiary hospital in Southern China to analyze the clinical features, sociodemographic data, and preoperative inflammatory, immune, and metabolic indicators of PA and WT patients. A simple, accurate diagnostic model is constructed to provide a low-cost, non-invasive, and efficient preoperative diagnostic tool aimed at optimizing treatment strategies and improving patient prognosis.

2 Materials and methods

2.1 Research objects

This study adheres to the TRIPOD guidelines (25), and data were collected from patients who underwent parotid tumor surgery in the Department of Otorhinolaryngology-Head and Neck Surgery at the First Affiliated Hospital of Guangxi Medical University between January 2013 and December 2022. The inclusion criteria were defined as: preoperative diagnosis of a parotid mass and subsequent surgical resection; availability of preoperative clinical data, peripheral blood biochemical test results within one week before surgery, and complete imaging records (MRI or CT); and postoperative pathological diagnosis of PA or WT. The exclusion criteria included malignant tumors from other locations, active infections or chronic inflammation, autoimmune diseases, or the use of corticosteroids or antibiotics within one week before surgery. The Institutional Review Board approved the study (No. 2023-E339-01), and each patient provided informed consent. The research process is shown in Figure 1.

Figure 1

Figure 1. Flowchart of the study design.

2.2 Data collection

One week before surgery, preoperative peripheral blood tests (including complete blood count, liver function, renal function, and blood glucose) were performed, and sociodemographic data, clinical symptoms, signs, and other relevant information were gathered from the hospital’s computerized medical records. Potential biomarkers were calculated, including body mass index (BMI), NLR, PLR, SII, and lymphocyte-to-monocyte ratio (LMR). Specific parameters are listed in Table 1. The largest tumor diameter was measured from CT or MRI images. Patients were categorized into married and unmarried groups (including divorced individuals) based on their marital status. Smoking was defined as having smoked continuously for at least six months prior to surgery, with a daily intake of more than one cigarette. Drinking was defined as consuming alcohol continuously for at least six months prior to surgery, with a weekly alcohol intake of 72 grams or more. A BMI of ≥25.0 kg/m² was considered overweight. Fasting blood glucose (FPG) was measured from venous blood samples taken before breakfast, and PBG2h was measured two hours after breakfast. Other blood samples were taken from 6 to 8 in the morning. The blood samples were analyzed in the hospital lab.

Table 1

Table 1. Baseline characteristics of patients with WT and PA in the total dataset.

2.3 Model construction and statistical analysis

Continuous variables with outliers (defined as deviations from the mean by two standard deviations or more) and a small number of missing values were addressed using median interpolation. The chi-squared test was used to evaluate differences in categorical variables between groups. For variables with a normal distribution, the mean ± standard deviation was employed as a means of presenting the data, with the use of an independent samples t-test. On the other hand, for non-normally distributed variables, where the data were represented using medians and interquartile ranges, the appropriate statistical test was the Mann-Whitney U test.

Participants were randomly assigned to the training and validation sets in a 3:1 ratio, and the balance of the datasets was assessed. A LASSO regression method based on ten-fold cross-validation was used in the feature selection process with the parameter λ=100. Stepwise backward logistic regression was employed to identify independent factors and build the model. The discriminative ability of the model was evaluated using the ROC, with the optimal cut-off value established by the maximum Youden index. This was followed by the calculation of the AUC, sensitivity, specificity, and accuracy. Differences in AUC values were compared using the DeLong test. Bootstrap resampling (boot = 500) was applied to obtain confidence intervals for the AUC values and to generate calibration curves for both datasets. The goodness of fit of the model was assessed using the Hosmer-Lemeshow test. Additionally, DCA and CIC were used to evaluate the applicability and net benefit of the model in clinical practice. All statistical analyses were performed using SPSS 23.0 and R 4.2.3 software, with the significance level set at p < 0.05.

3 Results

3.1 Baseline characteristics

The research sample comprised 249 patients with PA (excluding 26 cases) and 173 patients with WT (excluding 20 cases), resulting in a PA-to-WT ratio of 1.44:1. With the exception of nine additional markers, including pain and soreness, and the history of diabetes, all other characteristics exhibited significant differences when the two groups of individuals were compared (p < 0.05, Table 1).

3.2 Model development

In accordance with the 3:1 random allocation principle, 316 patients were allocated to the training dataset, while the remaining 106 patients were assigned to the validation dataset. No notable differences were observed between the sets regarding any of the variables (all, p > 0.05; Table 2). Using the LASSO method, six potential predictors were identified: gender, age, smoking history, bilateral parotid involvement, WBC, and PBG2h (Figure 2). These variables were confirmed as independent predictors through logistic regression (Table 3; Figure 3).

Table 2

Table 2. Baseline characteristics of individuals in training set and validation set.

Figure 2

Figure 2. LASSO regression analysis was used for feature selection. (A) As λ of the LASSO algorithm changed, the LASSO coefficient profiles of the 35 features in the group were observed. (B) The optimal penalty coefficient lambda (λ) was identified in the LASSO model, and a 10-fold cross-validation was carried out in the group.

Table 3

Table 3. Logistic regression analysis for the risk of PA.

Figure 3

Figure 3. The nomogram for differential diagnosis of PA and WT.

3.3 Model evaluation

3.3.1 Discriminative ability

The nomogram model, based on six variables, demonstrated an AUC of 0.951 (95% confidence interval [CI]: 0.928-0.973) in the training set, and an AUC value of 0.934 (95% CI: 0.881-0.987) in the validation set. The Delong test showed that the discriminative ability of the model was significantly better than that of a single predictor variable (p < 0.05; Figure 4; Table 4).

Figure 4

Figure 4. Comparison of ROC curves of the nomogram and six independent predictors in the training set (A) and validation set (B).

Table 4

Table 4. Comparison of the diagnostic efficacy of the nomogram with six independent predictors.

3.3.2 Calibration

In the training set, the calibration curve showed an excellent degree of consistency between predictive and actual results, as well as in the validation set (Figure 5). The Hosmer-Lemeshow test yielded the χ² value for the training set as 9.492 (p = 0.303) and the χ² value for the validation set as 15.429 (p = 0.051), indicating that the model has good calibration and is suitable for clinical use.

Figure 5

Figure 5. Calibration curves of nomogram in the training set (A) and validation set (B).

3.3.3 Clinical applicability

DCAs showed that the nomogram model offered a greater net benefit than any single predictor, particularly for threshold probabilities within the range of 0.05 to 0.97 in the training set as well as 0.07 to 0.96 in the validation set (Figure 6). Furthermore, CIC indicated that the model performed well in predicting high-risk PA patients, with results that closely aligned with actual incidence rates (Figure 7).

Figure 6

Figure 6. Comparison of DCA of the nomogram with 6 independent risk factors in the training set (A) and validation set (B).

Figure 7

Figure 7. CIC of the nomogram in training set (A) and validation set (B).

4 Discussion

This study introduces, for the first time, a multidimensional, data-driven nomogram model that leverages clinical characteristics and hematological indicators to accurately differentiate between PA and WT preoperatively. Utilizing LASSO regression and logistic regression analyses, independent predictors—including sex, age, smoking history, bilateral parotid involvement, WBC, and PBG2h—were identified for model construction. The results indicated that the model achieved an AUC value of 0.951 in the training set. The validation set demonstrated consistent performance, with an AUC reaching 0.934, confirming high sensitivity, specificity, and overall accuracy. The results of the calibration curve and the Hosmer-Lemeshow goodness-of-fit test showed that there was a high degree of agreement between the predicted probabilities of the nomogram model and the actual observed frequencies, which confirmed the excellent calibration performance of the model. Moreover, the DCA and CIC further verified that the model has important application value in clinical practice. Compared to FNAC and imaging techniques, this nomogram is a non-invasive, affordable, and easy-to-use diagnostic tool. It is especially useful in places without many resources, as it only uses normal blood tests and medical history. The standardization of hematological indicators and clinical data effectively minimized human bias, enhancing the model’s reliability and reproducibility. This study provides a new perspective for preoperative diagnosis of parotid tumors and offers a reference for developing predictive models for other tumors.

Within this investigation, a higher proportion of males (95.38%) were observed in the WT group compared to the PA group (58.63%), a finding that is consistent with those reported in other studies (22, 26). Considering that smoking serves as a standalone risk factor for WT (27), the gender disparity between PA and WT may be related to smoking, as women in China and Turkey smoke less than those in Western countries (7, 28, 29). Additionally, the WT group’s median age was 58 years, which is higher than the PA group’s median age of 38 years, as reported elsewhere (22, 28, 30, 31).

WT is often characterized by multifocal growth and bilateral parotid involvement, observed in approximately 2.10% to 27% and 3.8% to 30% of cases, respectively (28, 32). This study found a higher prevalence of bilateral parotid involvement in the WT group (10.40% vs. 0.40%) compared to the PA group, supporting bilateral involvement as a key distinguishing feature between PA and WT. While some studies suggest a link between bilateral involvement and Epstein-Barr virus (EBV) infection (33, 34), others have failed to detect EBV-DNA in WT cells, necessitating further investigation into its pathogenesis (35, 36).

According to current consensus, inflammatory reactions are closely linked to the development of WT (37). Accumulating evidence suggests that chronic inflammation is not only a hallmark of WT but also a key driver in its pathogenesis. Recent studies (37, 38) have highlighted that chronic inflammation plays a pivotal role in WT pathogenesis through multiple mechanisms: (1) persistent inflammatory microenvironment promotes mitochondrial dysfunction in epithelial cells leading to oncocytic transformation; (2) senescence-associated secretory phenotype (SASP) from aged cells perpetuates inflammatory responses; and (3) lymphoid stroma in WT actively maintains chronic inflammation through germinal center formation.

Chronic inflammation is also closely related to aging, and the risk of WT increases significantly with age, with an estimated 9.4% increase in incidence per additional year of age (OR = 1.094) (39). Aging cells secrete inflammatory factors (such as IL-1α and IL-6) through SASP, which intensifies local inflammatory response and promotes the proliferation of tumor cells by activating NF-κB and STAT3 pathways. In addition, mitochondrial dysfunction related to aging may further amplify inflammatory signals and form a vicious circle (37). These findings suggest that WT should be classified as an age-related disease, where aging-associated inflammation contributes to the initiation and progression of the tumor. Despite WBC levels staying within the normal range, the WT group in this study had considerably higher WBC and inflammatory markers such as NLR, LMR, and SII than the PA group. A persistent low-grade inflammatory condition could be indicated by the rise in WBC. In contrast to PA, which is mainly composed of fibrous tissue and epithelial cells, WT, which are made up of tumor epithelium and lymphoid stroma, is more likely to cause inflammatory reactions. This idea is further supported by WBC being identified as an independent predictor.

Kadletz et al. (9) reported a significant increase in BMI in WT patients (29.1 vs. 26.2, p < 0.0001), as well as a higher prevalence of metabolic syndrome-related comorbidities in WT patients (62.4% vs. 35.2%, p < 0.0001), suggesting obesity as a contributing factor to the development of WT. This is consistent with studies showing that components of metabolic syndrome—such as obesity, hypertension, and hyperglycemia—promote chronic inflammation through adipocytokine secretion and transforming growth factor-beta (TGF-β) pathway activation (40, 41). Gontarz et al. (39) served similar findings. However, due to the interplay between hypertension, overweight/obesity, hyperglycemia, and inflammation, this study compared the differences in metabolic indicators such as hypertension, diabetes history, BMI, and blood glucose levels between PA and WT patients. The WT group exhibited significantly higher rates of hypertension, overweight, fasting blood glucose, and postprandial glucose at 2 hours. LASSO regression analysis identified PBG2h as an independent risk factor for differentiating PA from WT. Based on existing literature, we speculate that obesity-induced increase of inflammatory factors in adipose tissue (such as macrophage, TNF-α and IL-6) and systemic inflammatory markers (such as CRP and NLR) (42) may provide a pro-inflammatory microenvironment for WT and participate in the pathogenesis of WT. It is worth noting that obesity and metabolic syndrome are themselves associated with accelerated biological aging and increased levels of systemic inflammaging markers, which may further exacerbate the age-related inflammatory milieu conducive to WT development. Overweight or obesity may promote fibrosis and inflammation, leading to parotid dysfunction (43). Additionally, abnormal levels of sialic acid, phosphate, and peroxidase activity in obese individuals could contribute to these effects (44). Furthermore, hyperglycemia may enhance the expression of TGF-β2 and its signaling pathways, promoting extracellular matrix protein deposition in the parotid gland, thus reducing salivation (45). Hyperglycemia can also damage the parasympathetic nerves of the parotid gland, leading to vasodilation, decreased salivary flow rate, and increased total protein levels in saliva (46).

Despite the fact that this study obtained rich data sources and set high standards for factor selection to enhance the rigor of the model, there are limitations to this study. There may be a selection bias because the sample of subjects was recruited from only one medical center. The observed variations in hematological parameters may have been due to unnoticed coexisting diseases, differences in laboratory analysis techniques, or lifestyle factors such as smoking, which was not specifically considered. Furthermore, some of the predictors are still uncertain in terms of how they may be associated with the pathogenesis of PA and WT. Additionally, individuals who received conservative observation treatment may not have been included in the model, as it is primarily based on pathology data. This study focused only on pleomorphic adenoma (PA) and Warthin tumor (WT), the two most common benign parotid tumors. However, salivary gland tumors are a very heterogeneous group of lesions with many different histological types. For example, oncocytoma, a benign tumor often seen in elderly patients, can also occur bilaterally in the parotid gland, similar to WT. Risk factors by tumor type and demographics may vary in different regions, and therefore, it may be necessary to fine-tune the model to the specific local environment.

5 Conclusions

In conclusion, the six-variable diagnostic model to differentiate PA from WT, based on sex, age, smoking history, bilateral parotid involvement, WBC, and PBG2h, exhibited good predictive performance in both training and validation datasets, as well as strong calibration and clinical applicability. This model constitutes a non-invasive, economical means of early diagnostic aid for appropriate differentiation between PA and WT.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by The Medical Ethics Committee of First Affiliated Hospital of Guangxi Medical University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

XH: Conceptualization, Methodology, Project administration, Supervision, Writing – original draft. FZ: Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft. JH: Data curation, Methodology, Software, Writing – original draft. JS: Formal analysis, Funding acquisition, Resources, Writing – review & editing. ZZ: Conceptualization, Methodology, Project administration, Supervision, Writing – review & editing.

Funding

The author(s) declare financial support was received for the research and/or publication of this article. This work was supported by the Grant Program of the Guangxi Zhuang Autonomous Region Education Department (No. YB2014085).

Acknowledgments

Although this study is retrospective in nature, authors would like to thank all patients with parotid gland tumors who provided blood samples, tumor tissue samples, and data.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Gao M, Hao Y, Huang MX, Ma DQ, Chen Y, Luo HY, et al. Salivary gland tumours in a northern Chinese population: a 50-year retrospective study of 7190 cases. Int J Oral Maxillofac Surg. (2017) 46:343–9. doi: 10.1016/j.ijom.2016.09.021

PubMed Abstract | Crossref Full Text | Google Scholar

2. Wittekindt C, Streubel K, Arnold G, Stennert E, and Guntinas-Lichius O. Recurrent pleomorphic adenoma of the parotid gland: analysis of 108 consecutive patients. Head Neck. (2007) 29:822–8. doi: 10.1002/hed.20613

PubMed Abstract | Crossref Full Text | Google Scholar

3. Redaelli de Zinis LO, Piccioni M, Antonelli AR, and Nicolai P. Management and prognostic factors of recurrent pleomorphic adenoma of the parotid gland: personal experience and review of the literature. Eur Arch Otorhinolaryngol. (2008) 265:447–52. doi: 10.1007/s00405-007-0502-y

PubMed Abstract | Crossref Full Text | Google Scholar

4. Valstar MH, de Ridder M, van den Broek EC, Stuiver MM, van Dijk B, van Velthuysen M, et al. Salivary gland pleomorphic adenoma in the Netherlands: A nationwide observational study of primary tumor incidence, Malignant transformation, recurrence, and risk factors for recurrence. Oral Oncol. (2017) 66:93–9. doi: 10.1016/j.oraloncology.2017.01.004

PubMed Abstract | Crossref Full Text | Google Scholar

5. Choi SY, Choi J, Hwang I, Cho J, Ko YH, and Jeong HS. Comparative longitudinal analysis of Malignant transformation in pleomorphic adenoma and recurrent pleomorphic adenoma. J Clin Med. (2022) 11:1808. doi: 10.3390/jcm11071808

PubMed Abstract | Crossref Full Text | Google Scholar

6. Quer M, Hernandez-Prera JC, Silver CE, Casasayas M, Simo R, Vander Poorten V, et al. Current trends and controversies in the management of Warthin tumor of the parotid gland. Diagnost (Basel). (2021) 11(8):1467. doi: 10.3390/diagnostics11081467

PubMed Abstract | Crossref Full Text | Google Scholar

7. Franzen AM, Kaup Franzen C, Guenzel T, and Lieder A. Increased incidence of Warthin tumours of the parotid gland: a 42-year evaluation. Eur Arch Otorhinolaryngol. (2018) 275:2593–8. doi: 10.1007/s00405-018-5092-3

PubMed Abstract | Crossref Full Text | Google Scholar

8. Saravakos P, Kourtidis S, Hartwein J, and Preyer S. Parotid gland tumors: A multicenter analysis of 1020 cases. Increasing incidence of Warthin’s tumor. Indian J Otolaryngol Head Neck Surg. (2022) 74:2033–40. doi: 10.1007/s12070-020-01981-z

PubMed Abstract | Crossref Full Text | Google Scholar

9. Kadletz L, Grasl S, Perisanidis C, Grasl MC, and Erovic BM. Rising incidences of Warthin’s tumors may be linked to obesity: a single-institutional experience. Eur Arch Otorhinolaryngol. (2019) 276:1191–6. doi: 10.1007/s00405-019-05319-6

PubMed Abstract | Crossref Full Text | Google Scholar

10. Lee SK, Rho BH, and Won KS. Parotid incidentaloma identified by combined 18F-fluorodeoxyglucose whole-body positron emission tomography and computed tomography: findings at grayscale and power Doppler ultrasonography and ultrasound-guided fine-needle aspiration biopsy or core-needle biopsy. Eur Radiol. (2009) 19:2268–74. doi: 10.1007/s00330-009-1407-5

PubMed Abstract | Crossref Full Text | Google Scholar

11. Wang HC, Zuo CT, Hua FC, Huang ZM, Tan HB, Zhao J, et al. Efficacy of conventional whole-body ¹⁸F-FDG PET/CT in the incidental findings of parotid masses. Ann Nucl Med. (2010) 24:571–7. doi: 10.1007/s12149-010-0394-6

PubMed Abstract | Crossref Full Text | Google Scholar

12. Borsetto D, Fussey JM, Cazzador D, Smith J, Ciorba A, Pelucchi S, et al. The diagnostic value of cytology in parotid Warthin’s tumors: international multicenter series. Head Neck. (2020) 42:522–9. doi: 10.1002/hed.26032

PubMed Abstract | Crossref Full Text | Google Scholar

13. Reerds S, Van Engen-Van Grunsven A, van den Hoogen F, Takes RP, Marres H, and Honings J. Accuracy of parotid gland FNA cytology and reliability of the Milan System for Reporting Salivary Gland Cytopathology in clinical practice. Cancer Cytopathol. (2021) 129:719–28. doi: 10.1002/cncy.22435

PubMed Abstract | Crossref Full Text | Google Scholar

14. Alkan U, Shkedy Y, Mizrachi A, Shpitzer T, Popovtzer A, and Bachar G. Inflammation following invasive procedures for Warthin’s tumour: A retrospective case series. Clin Otolaryngol. (2017) 42:1241–6. doi: 10.1111/coa.12857

PubMed Abstract | Crossref Full Text | Google Scholar

15. Shah KS and Ethunandan M. Tumour seeding after fine-needle aspiration and core biopsy of the head and neck–a systematic review. Br J Oral Maxillofac Surg. (2016) 54:260–5. doi: 10.1016/j.bjoms.2016.01.004

PubMed Abstract | Crossref Full Text | Google Scholar

16. Jung YJ, Han M, Ha EJ, and Choi JW. Differentiation of salivary gland tumors through tumor heterogeneity: a comparison between pleomorphic adenoma and Warthin tumor using CT texture analysis. Neuroradiology. (2020) 62:1451–8. doi: 10.1007/s00234-020-02485-x

PubMed Abstract | Crossref Full Text | Google Scholar

17. Matsuda E, Fukuhara T, Donishi R, Kawamoto K, Hirooka Y, and Takeuchi H. Usefulness of a novel ultrasonographic classification based on anechoic area patterns for differentiating Warthin tumors from pleomorphic adenomas of the parotid gland. Yonago Acta Med. (2017) 60:220–6. doi: 10.24563/yam.2017.12.002

PubMed Abstract | Crossref Full Text | Google Scholar

18. Yamamoto T, Kimura H, Hayashi K, Imamura Y, and Mori M. Pseudo-continuous arterial spin labeling MR images in Warthin tumors and pleomorphic adenomas of the parotid gland: qualitative and quantitative analyses and their correlation with histopathologic and DWI and dynamic contrast enhanced MRI findings. Neuroradiology. (2018) 60:803–12. doi: 10.1007/s00234-018-2046-9

PubMed Abstract | Crossref Full Text | Google Scholar

19. Lee YH, Yu IK, Han MH, Lee BH, Kim MS, and Song CJ. Warthin’s tumor of the parotid gland: CT and MR features. J Korean Soc Radiol. (2009) 61(1):17–22. doi: 10.3348/jksr.2009.61.1.17

Crossref Full Text | Google Scholar

20. Ikeda M, Motoori K, Hanazawa T, Nagai Y, Yamamoto S, Ueda T, et al. Warthin tumor of the parotid gland: diagnostic value of MR imaging with histopathologic correlation. Am J Neuroradiol. (2004) 25:1256–62.

PubMed Abstract | Google Scholar

21. Rao Y, Ma Y, Wang J, Xiao W, Wu J, Shi L, et al. Performance of radiomics in the differential diagnosis of parotid tumors: a systematic review. Front Oncol. (2024) 14:1383323. doi: 10.3389/fonc.2024.1383323

PubMed Abstract | Crossref Full Text | Google Scholar

22. Psychogios G, Vlastos I, Thölken R, and Zenk J. Warthin’s tumour seems to be the most common benign neoplasm of the parotid gland in Germany. Eur Arch Otorhinolaryngol. (2020) 277:2081–4. doi: 10.1007/s00405-020-05894-z

PubMed Abstract | Crossref Full Text | Google Scholar

23. Kuzucu İ, Güler İ, Kum RO, Baklacı D, and Özcan M. Increased neutrophil lymphocyte ratio and platelet lymphocyte ratio in Malignant parotid tumors. Braz J Otorhinolaryngol. (2020) 86:105–10. doi: 10.1016/j.bjorl.2019.02.009

PubMed Abstract | Crossref Full Text | Google Scholar

24. Sahin A, Kars A, Kilic K, Ucar HB, and Sakat MS. Efficacy of the systemic immune inflammation index in Malignant and benign parotid neoplasms. Cureus. (2022) 14:e31878. doi: 10.7759/cureus.31878

PubMed Abstract | Crossref Full Text | Google Scholar

25. Collins GS, Moons K, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ. (2024) 385:e078378. doi: 10.1136/bmj-2023-078378

PubMed Abstract | Crossref Full Text | Google Scholar

26. Yanes-Diaz J, Riestra-Ayora J, Rodriguez-Rivero A, Yebra-Gonzalez L, Chaure-Cordero M, Vaduva C, et al. Trend changes in the incidence of benign parotid tumours in the last 30 years in a Spanish population. Eur Arch Otorhinolaryngol. (2023) 280:855–60. doi: 10.1007/s00405-022-07644-9

PubMed Abstract | Crossref Full Text | Google Scholar

27. Freedman LS, Oberman B, and Sadetzki S. Using time-dependent covariate analysis to elucidate the relation of smoking history to Warthin’s tumor risk. Am J Epidemiol. (2009) 170:1178–85. doi: 10.1093/aje/kwp244

PubMed Abstract | Crossref Full Text | Google Scholar

28. Patel DK and Morton RP. Demographics of benign parotid tumours: Warthin’s tumour versus other benign salivary tumours. Acta Otolaryngol. (2016) 136:83–6. doi: 10.3109/00016489.2015.1081276

PubMed Abstract | Crossref Full Text | Google Scholar

29. Tunç O, Gönüldaş B, Arslanhan Y, and Kanlıkama M. Change in Warthin’s tumor incidence: a 20-year joinpoint trend analysis. Eur Arch Otorhinolaryngol. (2020) 277:3431–4. doi: 10.1007/s00405-020-06081-w

PubMed Abstract | Crossref Full Text | Google Scholar

30. Feng B, Wang Z, Cui J, Li J, Xu H, Yu D, et al. Distinguishing parotid polymorphic adenoma and Warthin tumor based on the CT radiomics nomogram: A multicenter study. Acad Radiol. (2023) 30:717–26. doi: 10.1016/j.acra.2022.06.017

PubMed Abstract | Crossref Full Text | Google Scholar

31. Wang CW, Chu YH, Chiu DY, Shin N, Hsu HH, Lee JC, et al. JOURNAL CLUB: the Warthin tumor score: A simple and reliable method to distinguish Warthin tumors from pleomorphic adenomas and carcinomas. AJR Am J Roentgenol. (2018) 210:1330–7. doi: 10.2214/AJR.17.18492

PubMed Abstract | Crossref Full Text | Google Scholar

32. Teymoortash A, Krasnewicz Y, and Werner JA. Clinical features of cystadenolymphoma (Warthin’s tumor) of the parotid gland: a retrospective comparative study of 96 cases. Oral Oncol. (2006) 42:569–73. doi: 10.1016/j.oraloncology.2005.10.017

PubMed Abstract | Crossref Full Text | Google Scholar

33. Santucci M, Gallo O, Calzolari A, and Bondi R. Detection of Epstein-Barr viral genome in tumor cells of Warthin’s tumor of parotid gland. Am J Clin Pathol. (1993) 100:662–5. doi: 10.1093/ajcp/100.6.662

PubMed Abstract | Crossref Full Text | Google Scholar

34. Scully C. Viruses and salivary gland disease: are there associations. Oral Surg Oral Med Oral Pathol. (1988) 66:179–83. doi: 10.1016/0030-4220(88)90090-4

PubMed Abstract | Crossref Full Text | Google Scholar

35. Kärjä V, Syrjänen K, and Syrjänen S. No Epstein Barr and cytomegalovirus DNA found in salivary gland tumours. ORL J Otorhinolaryngol Relat Spec. (1997) 59:97–9. doi: 10.1159/000276917

PubMed Abstract | Crossref Full Text | Google Scholar

36. Laane CJ, Murr AH, Mhatre AN, Jones KD, and Lalwani AK. Role of Epstein-Barr virus and cytomegalovirus in the etiology of benign parotid tumors. Head Neck. (2002) 24:443–50. doi: 10.1002/hed.10065

PubMed Abstract | Crossref Full Text | Google Scholar

37. Aoki R and Tanaka T. Pathogenesis of Warthin’s tumor: neoplastic or non-neoplastic. Cancers (Basel). (2024) 16:912. doi: 10.3390/cancers16050912

PubMed Abstract | Crossref Full Text | Google Scholar

38. Mochizuki K, Oishi N, Kawai M, Odate T, Tahara I, Inoue T, et al. Expressions of CXCL12, CXCL10 and CCL18 in Warthin tumors characterized pathologically by having a lymphoid stroma with germinal centers. Histol Histopathol. (2021) 36:931–8. doi: 10.14670/HH-18-352

PubMed Abstract | Crossref Full Text | Google Scholar

39. Gontarz M, Bargiel J, Gąsiorowski K, Marecik T, Szczurowski P, Hramyka A, et al. Could obesity be related to the increasing incidence of Warthin tumors. J Clin Med. (2024) 13:4935. doi: 10.3390/jcm13164935

PubMed Abstract | Crossref Full Text | Google Scholar

40. Divella R, Gadaleta Caldarola G, and Mazzocca A. Chronic inflammation in obesity and cancer cachexia. J Clin Med. (2022) 11:2191. doi: 10.3390/jcm11082191

PubMed Abstract | Crossref Full Text | Google Scholar

41. Zahedi AS, Daneshpour MS, Akbarzadeh M, Hedayati M, Azizi F, and Zarkesh M. Association of baseline and changes in adiponectin, homocysteine, high-sensitivity C-reactive protein, interleukin-6, and interleukin-10 levels and metabolic syndrome incidence: Tehran lipid and glucose study. Heliyon. (2023) 9:e19911. doi: 10.1016/j.heliyon.2023.e19911

PubMed Abstract | Crossref Full Text | Google Scholar

42. Cobos-Palacios L, Ruiz-Moreno MI, Vilches-Perez A, Vargas-Candela A, Muñoz-Úbeda M, Benítez Porres J, et al. Metabolically healthy obesity: Inflammatory biomarkers and adipokines in elderly population. PloS One. (2022) 17:e0265362. doi: 10.1371/journal.pone.0265362

PubMed Abstract | Crossref Full Text | Google Scholar

43. Mouton AJ, Li X, Hall ME, and Hall JE. Obesity, hypertension, and cardiac dysfunction: novel roles of immunometabolism in macrophage activation and inflammation. Circ Res. (2020) 126:789–806. doi: 10.1161/CIRCRESAHA.119.312321

PubMed Abstract | Crossref Full Text | Google Scholar

44. Zalewska A, Kossakowska A, Taranta-Janusz K, Zięba S, Fejfer K, Salamonowicz M, et al. Dysfunction of salivary glands, disturbances in salivary antioxidants and increased oxidative damage in saliva of overweight and obese adolescents. J Clin Med. (2020) 9(2):548. doi: 10.3390/jcm9020548

PubMed Abstract | Crossref Full Text | Google Scholar

45. Lamers ML, Gimenes FA, Nogueira FN, Nicolau J, Gama P, and Santos MF. Chronic hyperglycaemia increases TGFbeta2 signaling and the expression of extracellular matrix proteins in the rat parotid gland. Matrix Biol. (2007) 26:572–82. doi: 10.1016/j.matbio.2007.05.001

PubMed Abstract | Crossref Full Text | Google Scholar

46. Sato T, Mito K, and Ishii H. Relationship between impaired parasympathetic vasodilation and hyposalivation in parotid glands associated with type 2 diabetes mellitus. Am J Physiol Regul Integr Comp Physiol. (2020) 318:R940–940R949. doi: 10.1152/ajpregu.00016.2019

PubMed Abstract | Crossref Full Text | Google Scholar