Statement of the Problem
In oral squamous cell carcinoma (OSCC), the decision to perform an elective neck dissection (END) is straightforward when the primary tumor is >4cm, the depth of invasion is >4mm, or the surgical approach to the primary site provides access to the neck. With early stage disease, the management of the clinically negative neck (cN0) is less clear. Although recent evidence suggests that END is superior to watchful waiting, standardized guidelines do not exist for when to perform an END in early stage OSCC. Ultimately, any decision requires the best judgement of each surgeon, and therefore it can be assumed that neck observation is still recommended to a subset of patients. To date multiple studies have compared the survival between END and observation, however few studies have evaluated survival in a cohort of patients treated with neck observation alone. The primary purpose of this study was to describe how various factors influence survival among patients who receive neck observation with early stage cT1/2 N0 M0 OSCC. Our secondary aim was to compare the characteristics and survival rates between patients with cN0 and pN0 necks.
Materials and Methods
This is retrospective cohort study evaluating patients registered with the Surveillance, Epidemiology, and End Results (SEER) program. Between 2004 and 2015, patients were included if they had stage T1/2 N0 M0 OSCC. Stage was assigned using the American Joint Committee on Cancer (AJCC) 6th edition TNM grouping for oral cavity carcinomas. The final sample of T1/2 N0 M0 OSCC was divided between clinically (cN0) and pathologically (pN0) confirmed cases. Predictor variables were age at diagnosis, gender, race, tumor size, tumor location, and histologic grade. Outcome variables were survival time and cause of death.
Evaluating cN0 survival. For cN0 cases, the log-rank test was used to perform univariate analyses and test for differences in survival within predictor variables. Only the significant univariate predictors were included in the multivariate survival model which was performed using Cox proportional hazards regression models.
Comparing cN0 and pN0. Descriptive statistics were calculated and compared between cN0 and pN0 groups using chi-squared and Student’s t-tests. Kaplan-Meier analyses were used to estimate overall rates, and the log-rank test was used to compare survival between cN0 and pN0 groups.
Evaluating cN0 survival. From 2004 to 2015, there were 5,705 patients with T1/2 N0 M0 OSCC. Within these cases, the majority (n=3,433; 60.2%) underwent close observation without receiving an elective neck dissection. Among cN0 patients, overall estimated 1-, 2-, and 5-year survival rates were 89.6%, 79.5%, and 62.7%. Among the 1,610 patients who died, 358 (22.2%) had deaths attributable to their cancer. Disease-specific 1-, 2-, and 5-year survival rates were 96.6%, 92.2%, and 87.3%. Age, histologic grade (Figure 1), tumor size (Figure 2), and surgery were independently associated with improved OS in the univariate and multivariate analyses.
Comparing cN0 and pN0. Observation was more commonly performed in older females who had smaller and lower grade tumors. Observation was also more commonly performed for cancers of the hard palate and lip mucosa. Surgery to remove the primary tumor was less commonly performed when neck observation was employed. The cN0 patients had a marginal but significantly worse OS compared to pN0 patients (5-year OS: 62.7 vs 71.0%; p<0.01) (Figure 3).
Size, histologic grade, and patient age all significantly impact the prognosis of early stage cN0 OSCC to varying degrees. Even though cN0 lesions were smaller and of lower grade, the pathologically-confirmed N0 neck still had a better survival than the clinically observed N0 neck.