Cumulative incidence of neck recurrence with increasing depth of invasion

Objective: To determine if there is a critical depth of invasion that predicts micrometastasis in early oral tongue cancer. Methods: Retrospective series identifying patients undergoing primary surgical resection of T1 or T2 oral tongue cancer who elected against neck treatment between 2000 and 2015. Cox proportional-hazard model compared the relative hazard and cumulative incidence of recurrence to depth of invasion. The model used a 2 parameter quadratic effect for depth that was chosen based on Akaike's information criterion. Results: Ninety-three patients were identified with T1 or T2 oral tongue squamous cell carcinoma and clinically N0 neck undergoing glossectomy without elective neck treatment. 61% were male and median age was 60 years. Median follow up was 45 months, and 76 patients had at least two years of follow up. Thirty-six of 76 patients recurred (47.4%), with 15 recurring in the oral cavity (19.7%) and 21 developing nodal metastasis (27.6%). Cox proportional-hazards quadratic polynomial showed increasing hazard of recurrence with depth of invasion and the cumulative incidence increased sharply within the range of data from 2 to 6 mm depth of invasion. Conclusions: Depth of invasion is significantly associated with nodal metastasis and has been added to the 8th AJCC staging guidelines. Variable depths of invasion have been associated with regional metastasis; however, there is likely not a critical depth that predicts neck recurrence due to progressive hazards and cumulative risk of occult metastasis. The risk of regional metastasis is likely much greater than previously believed and increases progressively with increasing depth.