Summary Statistics
A total of 2624 primary HNACC cases were included in the study. The demographic and clinicopathological characteristics of the cases are summarized in Table 1. There is a significant difference in gender distribution; 1052 of the cases are male, and 1572 are female. The mean age at diagnosis was 50 years for the total cohort and ranged from 5 to 96 years. According to this dataset, HNACC occurred in the lip, nose and nasal cavity, middle ear, eye and orbit, gum, floor of the mouth, tonsil, tongue, thyroid gland, oropharynx, nasopharynx, hypopharynx, larynx, pharynx, submandibular gland, sublingual gland and parotid gland. Among the total cohort, 2416 cases originated from salivary glands, including 2325 parotid gland ACC cases. White people accounted for 81% of the total population (2131/2624). With respect to pathological differentiation, definite pathological information was available for only 938 cases. Most of the cases were grade I and grade II; grade III plus IV carcinoma accounted for 12.3%. The percentages of cases with lymph node metastasis and distant metastasis were 7% (96/1351) and 6% (76/1183), respectively. Surgical resection was the primary treatment modality. Surgery was performed in 2516 patients, including 905 patients in which simultaneous neck dissection was performed, and 839 cases received surgery plus radiotherapy.
Survival analysis
Kaplan-Meier analysis was performed for time-to-event analysis for overall survival (OS) and disease-specific survival (DSS). Regardless of confounding factors, 5-year, 10-year and 20-year OS was 90%, 80% and 64%, respectively. Significant differences in OS were found depending on age range (P < 0.0001), mean age (P < 0.0001), marital status at diagnosis (P < 0.0001), gender (P = 0.01), pathological differentiation (P < 0.0001), race (P = 0.007), AJCC stage (P < 0.0001), AJCC T stage (P < 0.0001), AJCC N stage (P < 0.0001), AJCC M stage (P < 0.0001), surgery, (P < 0.0001), radiotherapy (P < 0.0001), chemotherapy (P < 0.0001) and combined treatment modality (P < 0.0001) (Fig 1). Of the total cohort, 2203 patients were available for DSS analysis. The median follow-up time for these cases was 123 months (range, 1-503 months). Regardless of all other factors, 10-year and 20-year DSS was 93.6% and 90%, respectively, for patients who received surgery alone and 84.3% and 75.8%, respectively, for patients who received surgery plus radiotherapy (P < 0.0001). In addition, statistically significant differences in DSS were found to be associated with age (P < 0.0001), race (P = 0.007), mean age (P < 0.0001), marital status at diagnosis (P < 0.0001), gender (P = 0.007), AJCC stage (P < 0.0001), AJCC T stage (P < 0.0001), AJCC N stage (P < 0.0001), AJCC M stage (P < 0.0001), neck dissection (P = 0.024) and pathological differentiation (P < 0.0001) (Fig 2).
Multivariate survival analysis was performed using the Cox proportional hazards regression model and the significant variables listed above. Surgical treatment was favourably associated with better DSS and OS [HR (95% CI) = 0.13 (0.03-0.6), P = 0.0092; HR (95% CI) = 0.23 (0.07-0.79), P = 0.0203]. Gender was the only demographic independent prognostic factor for both DSS and OS [Male vs female, HR (95% CI) = 3.3 (1.51-7.22), P = 0.0028 for DSS; HR (95% CI) = 2.44 (1.05-5.64), P = 0.0376 for OS]. Higher pathological grade was adversely associated with DSS and OS [Grade II HR (95% CI) = 4.03 (1.04 - 15.7), P = 0.0444; Grade III + IV, HR (95% CI) = 35.64 (10.9 - 125.94), P = 0.0000 for DSS; Grade III + IV, HR (95% CI) = 4.49 (2.3 - 8.77), P = 0.0000 for OS, Grade I as reference]. Details of the multivariate Cox regression analysis are presented in Fig 3.