Data were extracted from the SEER database (17 registries, Nov 2021 Sub [2000–2019]) between 2004 and 2015. The SEER database is a national program that publishes cancer incidence and survival data and covers approximately 48% of the U.S. population (https://seer.cancer.gov/). Data collected in the SEER database include demographics, tumor morphology, tumor site, stage, treatment, and vital status. Patients diagnosed with primary head and neck MEC were included. The excluded criteria were as follows: 1) <18 years of age at diagnosis; 2) Diagnosis of MEC from autopsy or death certificate; 3) Survival time < 1 month; 4) Patients with unclear surgical information; 5) Patients with a T-stage of T0; and 6) Patients with missing follow-up information. MEC was identified using the histological classification code 8430/3 of the International Classification of Disease for Oncology, Third Edition (ICD-O-3). Since the data in this study are de-identified publicly available data, this retrospective cohort study was exempted from the institutional ethical committee by our hospital.

Overall Survival (OS) and Cancer-Specific Survival (CSS) were used as outcomes. Death from any cause occurring between diagnosis and the end of follow-up was OS. Death from MEC between diagnosis and the end of follow-up was CSS. Survival time was the time from diagnosis to death, or to loss to follow-up, or to the end of administrative follow-up (December 31, 2019).

Data were extracted from the SEER database using SEER*STAT version 8.4.0. Data on patients were collected including age, sex, income, race (White, Black, others, unknown), marital status, primary site (parotid gland, other major salivary glands, palate, other oropharyngeal areas, and other head and neck areas), tumor size, TNM stage, tumor grade (I [well-differentiated], II [moderately differentiated], III [poorly differentiated], IV [undifferentiated], unknown), American Joint Committee on Cancer (AJCC) stage, chemotherapy (no/unknown, yes), surgery (no, partial organ excision, total organ excision, unknown type of surgery), radiation (no, yes), combination therapy (no surgery and no radiotherapy, surgery only, radiotherapy only, surgery combined with radiotherapy), and survival time. For combination therapy, only surgery and radiotherapy were considered in this study due to few patients received chemotherapy.

Survivors and non-survivors in head and neck MEC patients were characterized. Continuous data were presented as mean±Standard Deviation (SD) or median with quartiles (Q1, Q3) and categorical data were presented as numbers with percentages. Differences in continuous data were compared using the t-test or Wilcoxon rank-sum test, and differences in categorical data were compared using the χ2 test or Fisher’s exact test.

Kaplan-Meier (KM) curves were utilized to explore OS and CSS in patients with different age (<53 years, ≥53 years), sex (female, male), tumor primary site (parotid gland, other major salivary glands, palate, other oropharyngeal areas, and other head and neck areas), tumor grade (I, II, III, IV, unknown), and AJCC stage (I, II, III, IV, unknown), and the log-rank test was applied. The covariates of OS and CSS were screened using a univariable Cox proportional hazard model. Factors affecting OS and CSS were analyzed using a multivariable Cox proportional hazard model. Moreover, the effects of treatments on survival were analyzed. Subgroup analyses were further used to evaluate the impact of treatments on OS and CSS in different subgroup populations. Due to the heterogeneity of the MEC, patients were divided into two groups based on tumor grade (tumor grade I/II and tumor grade III/IV) to analyze the effect of treatment on survival in different tumor primary sites. Hazard Ratio (HR) with 95% Confidence Interval (95% CI) were estimated. Statistical analyses were completed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA) and R version 4.2.2 was used for forest plotting. Statistical significance was set as a p-value < 0.05.

We identified 2995 patients with primary head and neck MEC from the SEER database 2004‒2015. Finally, 2692 eligible patients were enrolled in the study (Fig. 1). Table 1 presents the characteristics of these 2692 patients. The median age was 53.00 (39.00, 65.00) years and 1444 (53.64%) patients were female. For the primary tumor site, 1397 (51.89%) patients were in the parotid gland, 260 (9.66%) patients in other major salivary glands, 396 (14.71%) patients in the palate, 556 (20.65%) patients in other oropharyngeal areas, and 83 (3.08%) patients in other head and neck areas. Only 162 (6.02%) patients received chemotherapy. For surgery, 1615 (59.99%) patients received partial organ excision, 803 (29.83%) patients received total organ excision or radical excision, 60 (2.23%) patients received an unknown type of surgery, and 214 (7.95%) patients did not receive surgery. For combination therapy, 1609 (59.77%) patients received surgery only, 79 (2.93%) patients received radiotherapy only, 869 (32.28%) patients received surgery combined with radiotherapy, and 135 (5.01%) patients received neither surgery nor radiotherapy. At the end of the current follow-up, 2123 (78.86%) patients were survivors, 569 (22.14%) patients died, and 341 (12.67%) patients died of MEC. The median follow-up time was 93.00 (57.00, 136.00) months.

Figure 1.

Flow chart for patients. MEC, Mucoepidermoid Carcinoma; SEER, Surveillance, Epidemiology, and End Results database.

(0.29MB).

The KM curves of OS and CSS according to age, sex, tumor primary site, tumor grade, and AJCC stage were analyzed. Patients older than 53 years were associated with poorer OS and CSS (p<0.0001) compared to patients younger than 53 years (Fig. 2A). Females had better OS and CSS (p<0.0001) than males (Fig. 2B). Patients with a tumor primary site located in the palate had the best OS and CSS (p<0.0001), while patients with a tumor primary site located in other head and neck areas had the worst OS and CSS (Fig. 2C). Patients with grade I or grade II were related to better OS and CSS (p<0.0001) compared to those with grade III or grade IV (Fig. 2D). Patients with AJCC stage I had optimal OS and CSS (p<0.0001), whereas patients with AJCC stage IV had poorer OS and CSS (Fig. 2E).

Figure 2.

Kaplan-Meier (KM) curves of OS and CSS in patients with head and neck MEC. (A) age subgroups; (B) sex subgroups; (C) tumor primary site subgroups; (D) tumor grade subgroups; (E) AJCC stage subgroups. OS, Overall Survival; CSS, Cancer-Specific Survival; MEC, Mucoepidermoid Carcinoma.

(0.53MB).

Factors correlated with OS and CSS in patients with head and neck MEC were presented in Table 2. For the multivariable analysis of OS, older age (≥53 years), males, unmarried, lower income (<75,000$), tumor site in other head and neck areas, higher tumor grade (III/IV), larger tumor size, and higher AJCC stage were linked to worse OS in patients with head and neck MEC, whereas others race and tumor site in the palate or other oropharyngeal areas were correlated with better OS. In the multivariable analysis of CSS, the factors affecting CSS were consistent with OS.

Table 3 shows the impact of treatment modality on survival in patients with head and neck MEC. Patients who received chemotherapy had poorer OS compared with patients who did not (multivariable: [HR=1.44, 95% CI 1.14–1.84]). For surgery, only patients who did not receive surgery (HR=3.20, 95% CI 2.45–4.18) had worse OS, while no significant difference was detected between partial and total organ excision on patients’ OS (p=0.729). For radiotherapy, no significant difference in OS was observed between patients with and without radiotherapy (p=0.459). For combination therapy, patients who received radiotherapy only (HR=3.21, 95% CI 2.27–4.53) or no surgery and no radiotherapy (HR=2.59, 95% CI 1.83–3.67) were linked to worse OS compared to patients who received surgery only, but no significant difference was detected between surgery only and surgery combined with radiotherapy on patients’ OS (p=0.218). In the multivariable analysis of CSS, patients who received chemotherapy were related to worse CSS compared with patients who did not (HR=1.41, 95% CI 1.08–1.85). For surgery, patients who did not receive surgery (HR=4.06, 95% CI 2.88–5.72) were correlated with poorer CSS compared with patients who received partial excision, but no significant difference was detected between partial excision and total excision on patients’ CSS (p=0.070). For radiotherapy, no significant difference was detected in CSS between radiotherapy and non-radiotherapy (p=0.124). For combination therapy, patients who received radiotherapy only (HR=4.43, 95% CI 2.85–6.86) or no surgery and no radiotherapy (HR=4.09, 95% CI 2.56–6.55) were related to poorer CSS compared to patients who received surgery only, but no significant difference was detected between surgery only and surgery combined with radiotherapy on patients’ CSS (p=0.069).

Fig. 3 shows the impact of treatment modality on survival based on age, sex, tumor primary site, tumor grade, and AJCC stage. Patients who received radiotherapy only or no surgery or no surgery and no radiotherapy had poorer OS and CSS at age < 53 years or ≥53 years, and patients who received chemotherapy also had worse OS at age < 53 years (p<0.05) (Fig. 3A). Females or males who received radiotherapy only or no surgery or no surgery and no radiotherapy had poorer OS and CSS, and females who received chemotherapy were also had worse OS (p<0.05) (Fig. 3A). Patients who received radiotherapy only or no surgery or no surgery and no radiotherapy were correlated with worse OS and CSS when the tumor was in the parotid gland and other major salivary glands (p<0.05) (Fig. 3B). Patients who received radiotherapy only were related to poorer OS and CSS (p<0.05). When the tumor was in other head and neck areas, patients who received chemotherapy were linked to worse OS and CSS (p<0.05). Patients who received chemotherapy or no surgery or no surgery and no radiotherapy had worse OS and CSS when the tumor grade was I/II or III/IV, while patients with tumor grade III/IV receiving surgery combined with radiotherapy may be related to better OS (p<0.05) (Fig. 3C). Patients who received radiotherapy only or no surgery had worse OS and CSS when the AJCC stage was I/II or III/IV (p<0.05) (Fig. 3C). In addition, OS and CSS were worse in stage I/II patients who received chemotherapy and in stage III/IV patients who did not receive surgery and radiotherapy (p<0.05).

Figure 3.

The impact of treatment modality on OS and CSS in different patients. (A) Age and sex subgroups; (B) Tumor primary site subgroups; (C) Tumor grade and AJCC stage subgroups. OS, Overall Survival; CSS, Cancer-Specific Survival; all analyses were multivariable analysis, adjusted for (OS and CSS): age, sex, race, marital status, income, primary site, tumor grade, tumor size, AJCC stage, chemotherapy (not adjusted in chemotherapy analysis), surgery (not adjusted in surgery or combination therapy analyses), and radiotherapy (not adjusted in radiation or combination therapy analyses).

(0.55MB).

Moreover, the effects of treatment modality on patients with the same tumor grade and tumor primary site were shown in Supplementary Table 1 and Supplementary Table 2. For patients with parotid gland MEC, patients with tumor grade III/IV who received radiotherapy only or no surgery and no radiotherapy had poorer CSS compared to patients who received surgery only, but patients with tumor grade I/III who received surgery combined with radiotherapy had poorer CSS (p<0.05). For patients with palate MEC, patients with tumor grade I/III who did not receive surgery and radiotherapy had worse CSS compared to patients who received surgery only (p<0.05). For other patients with the same tumor grade and tumor primary site, the stratification of tumor grade and tumor primary site resulted in too small a sample size for the corresponding group to draw conclusions.