3.2. Temporal Trends
When analyzed by year (
Figure 4), case numbers showed a bimodal distribution with two major peaks, 2017–2019, followed by a reduced incidence coinciding with the COVID-19 pandemic years (2020–2022), and a partial rebound in 2023. The highest annual incidence is recorded in 2019, with 41 cases (19.8%), while the lowest incidence occurred in 2022, with 12 cases (5.8%).
To explore the potential influence of the COVID-19 pandemic and associated restrictions on medical services, the cohort was divided into two periods. First period, pre-pandemic (2016–2019) has 148/207 cases (71.5%) and pandemic and post-pandemic recovery (2020–2023) has 59/207 cases (28.5%), as can be seen in
Figure 5.
This represents a 60.1% decrease in case volume between the two periods. The reduction was particularly evident in elective benign tumor resections, while the proportion of malignant tumors (discussed in
Section 3.4) did not decline proportionally, suggesting that urgent oncologic surgeries continued despite restrictions.
The decline in case numbers during 2020–2022 is likely multifactorial, but a major contributing factor was the COVID-19 pandemic-related restriction of non-urgent surgical activity, limited patient access to specialty clinics, and patient hesitancy to seek medical attention for non-life-threatening conditions. The partial rebound in 2023 may reflect a resumption of deferred surgical cases and increased patient presentation following the lifting of pandemic measures.
Figure 5a illustrates the annual incidence by age group, demonstrating that all age categories were affected by the decline, with the most pronounced drop in patients ≥60 years, potentially due to higher vulnerability and healthcare avoidance during the pandemic.
Table 2 presents the detailed comparison between the pre-pandemic and pandemic/post-pandemic periods, including absolute numbers and percentages for each age group and sex.
3.3. Histopathological Diagnosis
Histopathological evaluation of the 207 parotid nodular lesions identified 27 distinct diagnostic entities. These were grouped into benign epithelial tumors, malignant epithelial tumors, lymphoid neoplasms, non-neoplastic lesions, and other rare cases.
Table 3 summarizes the absolute counts for each diagnosis in the pre-pandemic (2016–2019) and pandemic/post-pandemic (2020–2023) periods while in
Figure 6 a bar chart representing the distribution of most common histopathologic diagnoses.
Benign epithelial tumors dominated the series, accounting for over half of all cases. Pleomorphic adenoma was the most common diagnosis (64/207; 30.9%), with 42 cases pre-pandemic and 22 cases during the pandemic/post-pandemic interval. Warthin tumor ranked second (59/207; 28.5%), with a similar temporal distribution (36 vs. 23 cases). Other benign tumors were rare, including lipoma (n = 2) and oncocytic cyst (n = 1). Malignant epithelial tumors accounted for a relevant minority of cases. The most frequent were squamous cell carcinoma (6/207; 2.9%), adenocarcinoma of salivary origin (5/207; 2.4%), mucoepidermoid carcinoma (6/207; 2.9%), and malignant epithelial carcinoma (NOS) (4/207; 1.9%).
Non-neoplastic lesions were less frequent but clinically relevant: chronic sialadenitis (n = 7), reactive lymph node (n = 4), branchial cyst (n = 4), salivary duct cyst (n = 2), acute suppurative sialadenitis (n = 1), benign lymphoepithelial lesion (n = 1), salivary cyst (n = 1), epidermoid cyst (n = 1), and normal adipose tissue (n = 1). Rare inflammatory lesions included desmoid-type fibromatosis (n = 1).
Of the 207 cases reviewed, 21 specimens were classified as indeterminate due to lack of viable tissue or inconclusive histopathological findings. These cases were excluded from further malignancy pattern analysis, leaving 186 analyzable cases. Therefore, the following sections (3.4 onward) report distributions and subgroup analyses based solely on this analyzable cohort.
The temporal distribution demonstrated that several uncommon entities (e.g., mucoepidermoid carcinoma, branchial cysts, non-Hodgkin lymphoma) were not observed in the 2020–2023 interval, reflecting reduced elective surgical activity and stricter case selection during the pandemic.
3.4. Malignancy Patterns
Histopathological evaluation was performed for all 207 parotid nodular lesions. Of these, 21 cases remained indeterminate and were excluded from the classification analysis. The final evaluable cohort therefore comprised 186 cases. Histopathological classification of the 186 parotid lesions showed that the majority were benign tumors (126 cases; 67.74%), followed by malignant tumors (31 cases; 16.66%), non-neoplastic lesions (22 cases; 11.82%), lymphoid neoplasms (6 cases; 3.22%), and other rare stromal lesions (1 case; 0.53%) (
Figure 7).
Benign tumors were dominated by pleomorphic adenoma and Warthin tumor, which together accounted for nearly 60% of the series.
Malignant cases included both primary salivary gland carcinomas (e.g., squamous cell carcinoma, mucoepidermoid carcinoma, adenocarcinoma, acinic cell carcinoma, myoepithelial carcinoma) and rare variants such as oncocytic carcinoma, basal cell carcinoma, sarcoma, and cribriform adenocarcinoma. A single case of metastatic melanoma to the parotid lymph node was also identified.
The overall incidence of malignancy (15.0%) is in line with epidemiological studies from tertiary referral centers, reinforcing the predominance of benign pathology in surgically excised parotid nodular lesions.
3.4.1. Malignancy by Period
When divided into pre-pandemic (2016–2019) and pandemic/post-pandemic (2020–2023) intervals, a clear decline in malignant parotid tumors was observed. In 2016–2019, 26 malignancies were identified among 130 analyzable cases, corresponding to a malignancy rate of 20.0%. By contrast, in 2020–2023, only 5 malignancies were recorded among 56 analyzable cases, yielding a malignancy rate of 8.9%. This represents a relative reduction of approximately 55% and an absolute difference of 11.1 percentage points between the two periods (
Table 4).
Benign tumors dominated both intervals (80 cases in 2016–2019 vs. 46 cases in 2020–2023). Lymphoid neoplasms were less frequent but present in both intervals (5 vs. 1), whereas non-neoplastic lesions showed a similar decline (18 vs. 4). Only a single case from the “other” category (desmoid-type fibromatosis) was observed in 2016–2019, with none in 2020–2023 (
Figure 8).
These findings underline that, despite prioritization of oncologic and symptomatic cases during COVID-19 restrictions, the absolute number and proportion of malignant tumors decreased in the pandemic/post-pandemic interval. This likely reflects reduced patient access to healthcare, delays in diagnostic workup, and deferral of non-urgent surgical interventions, with benign tumors being most affected by postponement.
3.4.2. Malignancy by Age Group
An age-dependent gradient in malignancy incidence was observed and it is presented in
Table 5 and
Figure 9. This pattern suggests a clear age-associated risk, consistent with previous epidemiological studies indicating that the probability of parotid malignancy increases with age, likely due to cumulative genetic alterations and prolonged exposure to environmental risk factors.
Benign tumors remained the most frequent category across all age groups, particularly in patients younger than 60 years. Non-neoplastic lesions were relatively more common in the <40-year group, while “Other” diagnoses, including rare entities such as sarcoma and desmoid-type fibromatosis, were almost exclusively encountered in the ≥60-year group.
3.4.3. Malignancy by Sex
When stratified by sex, male patients had a higher proportion of malignant tumors than females (17/94; 18.1% vs. 14/92; 15.2%). Benign tumors were predominant in both groups—54/94 (57.4%) in males and 72/92 (78.3%) in females.
Figure 10 illustrates these distributions graphically, highlighting the predominance of benign pathology in both sexes, the higher relative burden of malignant and non-neoplastic lesions in males, and the clear female predominance for benign tumors.
3.4.4. Histological Spectrum of Malignant Cases
From the total cases analyzed, 31 (15.0%) were confirmed as malignant tumors, comprising both primary salivary gland cancers and metastatic lesions.
The most common histological type was squamous cell carcinoma (n = 6; 19.4%), followed by mucoepidermoid carcinoma (n = 6; 19.4%), and adenocarcinoma of salivary origin (n = 5; 16.1%). Other frequent diagnoses included lymphoma (n = 6; 19.4%) and malignant epithelial carcinoma, NOS (n = 4; 12.9%). Less common entities such as acinic cell carcinoma (n = 2), myoepithelial carcinoma (n = 2), basal cell adenocarcinoma (n = 2), oncocytic carcinoma (n = 1), sarcoma (pleomorphic) (n = 1), cribriform adenocarcinoma of the salivary gland (n = 1), and metastatic melanoma to the parotid lymph node (n = 1) were rare but clinically relevant (
Table 6).
A slight predominance in males was observed (19/31; 61.3%) compared with females (12/31; 38.7%).
When stratified by age, malignant tumors were uncommon in patients <40 years (2/32; 6.3%), increased in those aged 40–59 years (12/70; 17.1%), and were most frequent in patients ≥60 years (17/84; 20.2%).
This distribution reinforces the association between increasing age and the risk of malignant transformation in parotid tumors, aligning with trends reported in previous epidemiological studies.
3.4.5. Histological Spectrum of Benign Cases
Among the analyzed parotid nodular lesions, 126 (60.9%) were classified as benign tumors. The distribution was dominated by two major entities: pleomorphic adenoma (n = 64; 50.8%) and Warthin tumor (n = 59; 46.8%), which together accounted for nearly all benign cases. Other rare benign lesions included lipoma (n = 2; 1.6%) and oncocytic cyst (n = 1; 0.8%).
A modest female predominance was observed (72/126; 57.1% females vs. 54/126; 42.9% males). In terms of age distribution, benign tumors were most frequent in patients <60 years, with 72/102 (70.6%) cases in this group. Pleomorphic adenoma was more common in younger adults, particularly in the <40 and 40–59 age groups, while Warthin tumor predominated in older adults (≥60 years).
This age-related divergence reflects known biological patterns: pleomorphic adenoma typically affects younger and middle-aged adults, while Warthin tumor is strongly associated with older age and smoking habits. Rare benign entities (lipoma, oncocytic cyst) were confined to patients aged ≥60 years (
Table 7).
3.5. Statistical Analysis
All statistical analyses were conducted using a two-sided significance threshold of α = 0.05.
For categorical variables (e.g., sex, laterality, diagnostic category, period), we evaluated associations using the chi-square (χ2) test of independence, which is designed to test whether two categorical variables are statistically related. The chi-square test compares the observed frequencies in each category with the frequencies expected if the two variables were independent.
When expected cell counts were <5 in more than 20% of cells, a situation where the chi-square approximation is unreliable, we instead applied Fisher’s exact test, which computes the exact probability of observing the distribution under the null hypothesis.
For binary comparisons (e.g., malignant vs. benign by sex, or by period), we calculated odds ratios (ORs) with 95% confidence intervals (CIs) to provide an effect size estimate. Odds ratios are particularly useful for clinical interpretation because they quantify how much more (or less) likely an outcome is in one group compared with another.
For categorical variables with more than two levels (e.g., age group stratified into <40, 40–59, and ≥60 years; or laterality with right, left, bilateral), results are reported using χ2 test statistics, supplemented with narrative interpretation of which groups contributed most to the observed pattern.
For continuous variables, age at diagnosis was analyzed in more detail. First, we tested whether age followed a normal distribution using the Shapiro–Wilk test for normality. The test returned a significant result (p < 0.05), indicating that the age distribution deviated from Gaussian form and was right-skewed, with more patients concentrated in older age brackets. Because the assumptions of parametric tests (e.g., Student’s t-test) were not met, we used the non-parametric Mann–Whitney U test for two-group comparisons of age distributions (e.g., malignant vs. benign cases, or pre-pandemic vs. pandemic/post-pandemic). This test is based on ranking values and does not assume normality, making it appropriate for skewed clinical data.
All analyses were performed on the analyzable cohort (n = 186), which excluded 21 indeterminate cases due to lack of viable tissue or inconclusive histopathology.
3.5.1. Period (2016–2019 vs. 2020–2023) Paired with Malignancy
To test whether the proportion of malignant diagnoses changed between the pre-pandemic (2016–2019) and pandemic/post-pandemic (2020–2023) periods, we applied a chi-square test of independence.
The chi-square statistic (χ2(1) ≈ 2.1–2.4, p ≈ 0.12–0.15) showed no statistically significant association between period and malignancy proportion.
We also computed the odds ratio (OR ≈ 2.1, 95% CI 0.79–5.66), which indicated that patients in the pre-pandemic years had approximately twice the odds of a malignant diagnosis compared to those in the pandemic/post-pandemic years. However, the confidence interval crossed 1.0, and the p-value exceeded the 0.05 threshold, confirming statistical non-significance.
Although the absolute number of parotid surgeries declined dramatically during the pandemic (
Section 3.2), the relative proportion of malignancies remained stable. This suggests that oncologic and high-priority cases were maintained as surgical priorities, while benign and elective cases were disproportionately deferred. The non-significant statistical result emphasizes that, in this cohort, pandemic-related disruptions primarily affected surgical volume rather than altering the malignant-to-benign case ratio.
3.5.2. Sex Paired with Malignancy
We next examined the relationship between sex and diagnosis type (benign vs. malignant tumors). Among the 157 analyzable cases, benign tumors were more frequent in females (72/126; 57.1%), while malignant tumors were relatively more common in males (19/31; 61.3%). A chi-square (χ2) test of independence did not show a statistically significant association between sex and diagnosis type (χ2(1) = 2.91, p = 0.088). The odds of malignancy in males were approximately twice those in females (OR = 2.11, 95% CI 0.89–4.98), although this result did not achieve statistical significance. Similarly, the relative risk (RR) indicated that males were about 82% more likely to present with malignancy compared with females.
From a clinical perspective, this distribution highlights a trend toward sex-based differences in tumor behavior: benign tumors predominated in females, while malignancies were more frequent in males. Although the current cohort size limited the power to detect statistical significance, this pattern is consistent with prior epidemiological studies reporting a slight male predominance in parotid malignancies and a higher frequency of benign lesions in females.
3.5.3. Age Group (<40, 40–59, ≥60) Paired with Malignancy
Patients were stratified into three ordered age groups: <40 years (n = 32), 40–59 years (n = 70), and ≥60 years (n = 84). Across these groups, benign tumors predominated, but the proportion of malignancies increased progressively with age.
Specifically:
In the <40 group, 22 cases were benign (68.8%) and 2 were malignant (6.3%).
In the 40–59 group, 50 cases were benign (71.4%) and 12 were malignant (17.1%).
In the ≥60 group, 54 cases were benign (64.3%) and 17 were malignant (20.2%).
A chi-square (χ2) test of independence comparing benign vs. malignant distributions across the three age groups did not reach statistical significance (χ2(2) ≈ 1.91, p ≈ 0.39), indicating that differences in malignancy rates across age strata could have occurred by chance.
To formally test for a linear age-related trend, we performed a Cochran–Armitage trend test, assigning scores 1 (<40), 2 (40–59), and 3 (≥60). The test yielded a Z ≈ 0.58 (p ≈ 0.56), confirming no statistically significant monotonic increase in malignancy with advancing age in this cohort.
For clinical interpretability, we collapsed age into two categories: <60 years (n = 102) vs. ≥60 years (n = 84). The odds ratio (OR) for malignancy in the ≥60 group compared with younger patients was 1.20 (95% CI: 0.55–2.63), suggesting only a modest and non-significant increased risk in older patients.
While raw numbers suggest that malignancies accumulate with age, particularly in the ≥60 years group, the statistical analysis indicates that the difference is not robust. Benign tumors remained the majority diagnosis in all age strata, though the type varied: pleomorphic adenoma dominated in younger patients, while Warthin tumor was the most common benign lesion in older adults.
3.5.4. Laterality (Right, Left, Bilateral) Paired with Malignancy
Laterality distribution was examined for the analyzable cohort (n = 186). Overall, right-sided lesions were slightly more common (n = 96; 51.6%) than left-sided lesions (n = 87; 46.8%), with bilateral involvement rare (n = 3; 1.6%).
When stratified by diagnosis type:
Right-sided lesions: 78 benign (81.3%) and 18 malignant (18.7%).
Left-sided lesions: 74 benign (85.1%) and 13 malignant (14.9%).
Bilateral lesions: 2 benign (66.7%) and 1 malignant (33.3%).
A chi-square (χ2) test of independence was applied to evaluate the relationship between laterality (right, left, bilateral) and malignancy status. The test result was χ2(2) ≈ 0.06, p = 0.97, indicating no statistically significant association between side of involvement and likelihood of malignancy.
For binary comparison of right vs. left lesions only, the odds ratio (OR) for malignancy in right-sided lesions compared with left-sided lesions was 1.31 (95% CI: 0.59–2.95). This suggests a slightly higher risk for right-sided malignancies, but the wide confidence interval and non-significant p-value confirm that this difference is likely due to chance.
Malignant tumors were nearly evenly distributed across right and left parotid glands, with one exceptional bilateral malignant case. Benign tumors predominated across all laterality categories. These findings, consistent with prior reports, confirm that laterality is not a clinically relevant predictor of tumor behavior in parotid neoplasms.
3.5.5. Age as a Continuous Variable
Age was analyzed as a continuous variable to explore potential differences between clinical subgroups. The Shapiro–Wilk test for normality indicated that the age distribution deviated significantly from a Gaussian pattern (p < 0.05), with a right-skewed shape and a median value lower than the mean. This skewness was consistent with a subset of very elderly patients, resulting in a longer upper tail of the distribution. Given the non-normal distribution, non-parametric methods were selected for group comparisons.
Mann–Whitney U tests were applied to compare median ages between (i) malignant and benign cases, and (ii) cases diagnosed in the pre-pandemic period (2016–2019) versus the pandemic/post-pandemic period (2020–2023). In both comparisons, no statistically significant differences were detected (p > 0.05). The overlap of interquartile ranges and the absence of a meaningful shift in medians suggest that age was not a major determinant of malignancy status in this cohort, nor did the age profile of patients change appreciably between study periods.
These findings imply that the demographic structure of patients with parotid lesions remained stable over time, and that the higher malignant case counts observed in older age groups (
Section 3.5.3) reflected categorical trends rather than large shifts in the underlying continuous age distribution.
3.5.6. Temporal Trends in Annual Case Volume (2016–2023)
To assess how the total number of parotid tumor cases evolved across the study period, we analyzed annual case counts from 2016 through 2023. The data showed a plateau during 2017–2019 (40–41 cases per year), followed by a sharp decline starting in 2020. The lowest point occurred in 2021, with only 11 recorded cases, after which partial recovery was noted in 2023 (21 cases). Despite this rebound, volumes remained well below pre-pandemic levels.
A simple linear regression was applied, with year as the independent variable and annual case count as the dependent variable, to test whether there was a systematic temporal trend. The model produced a slope of –3.51 cases/year (SE = 1.83), with R2 = 0.44 and p ≈ 0.074. Although the p-value did not reach the conventional 0.05 threshold for statistical significance, the negative slope reflects a clinically relevant average decline of 3–4 cases per year.
Visual inspection of the regression line (
Figure 11) highlights that the decline was not gradual but rather characterized by an abrupt step decrease coinciding with the COVID-19 pandemic years (2020–2022). The shaded interval in the plot corresponds to this period of restricted elective surgical activity and reduced diagnostic throughput, which disproportionately affected benign and indeterminate lesions.
From a healthcare systems perspective, these findings suggest that the fall in surgical case volume was primarily driven by external constraints (pandemic-related service reductions) rather than a true epidemiological change in disease incidence. The partial rebound observed in 2023 may represent deferred surgical cases or improved patient access after restrictions were lifted. Ongoing surveillance will be necessary to determine whether volumes eventually return to pre-2020 levels or stabilize at a new baseline.