Elective Neck Dissection, but Not Adjuvant Radiation Therapy, Improves Survival in Stage I and II Oral Tongue Cancer with Depth of Invasion >4 mm

Purpose/objective(s) In early-stage, node negative oral tongue cancer, there is limited data supporting tumor depth of invasion (DOI) as an indication for post-operative radiotherapy (PORT) to the primary site. The primary aim of this study is to examine the effect of tumor DOI and PORT on overall survival (OS). Materials and methods The National Cancer Database (NCDB) was used to query patients with AJCC stage I and II oral tongue cancer (2006-2013). Patients were stratified by receipt of PORT, elective neck dissection (ND), and DOI (≤4 mm or >4 mm). Kaplan-Meier analysis was performed to compare OS (using the log-rank test) between PORT versus no-PORT. Multivariable Cox proportional hazards regression model performed to evaluate the independent effect of PORT and neck dissection on OS. Results Among 939 patients, 69.3% were clinical stage I, 67.4% received ND, 23.4% had DOI >4 mm, and 10.4% received PORT. The addition of PORT did not improve OS with tumor DOI ≤4 mm (p = 0.634) or >4 mm (p = 0.816). The addition of elective neck dissection improved OS for DOI >4 mm (p = 0.010), but not for ≤4 mm (p = 0.128). On multivariable analysis, ND improved OS if DOI >4 mm (HR, 0.37; 95% CI, 0.17-0.81 [p = .012]), when also controlling for age, sex, PORT status, clinical stage, and pathological stage. Conclusion Tumor DOI should not be used as a sole indication for PORT in early stage oral tongue cancers. Elective neck dissection at the time of excision of the primary tumor results in higher OS for tumors with DOI >4 mm.


Introduction
Depth of invasion (DOI) is defined as the length measured from the tumor surface to the deepest point of invasive tumor in a paraffin embedded section [1]. The cut-off commonly used to stratify patients into low and high risk is 4 mm [2]. DOI is an important prognostic factor for nodal metastasis in oral tongue cancer, with increasing DOI associated with nodal involvement and worse prognosis [3][4][5][6][7].
Recent studies demonstrated no benefit to adding radiation therapy (RT) for deeper tumors [5][6][7][8][9]. O'steen et al. retrospectively evaluated the outcomes of 32 patients with stage N0-2b oral tongue or floor of mouth cancers with the primary tumor not crossing the midline who underwent PORT. The DOI in >75% patients was >4 mm, >75% had positive or close (<5 mm) margins and 38% had perineural invasion (PNI). RT to contralateral (CL) neck was omitted. At a median follow-up of 5.5 years among patients alive at the end of the study, there were no isolated nodal recurrences despite the majority of tumors possessing of DOI >4 mm. The authors concluded that the risk of nodal recurrence when omitting CL neck RT was very low if the primary tumor did not cross the midline, irrespective of other risk factors [5].
Rajappa et al. evaluated 375 pT1-2N0 oral tongue cancer patients. The cohort's median age was 49, and 93% had squamous cell carcinomas, with 37.6% and 5.87% possessing PNI and lymphovascular invasion (LVI), respectively. PORT was delivered in 37.6% of the cohort for PNI/LVI in the majority of cases, and for close margins in the remaining patients. At a mean follow-up of 40.9 months, there was a 18.4% local recurrence rate, with a 12.7-month mean duration of recurrence. Forty-four percent of the recurrences were salvaged while the remainder developed distant metastasis (DM) or unresectable disease. The two-and five-year overall survival (OS) were 94.5% and 93.9%, respectively. The patients were further divided into three groups: DOI <5 mm, 6-10 mm and >10 mm, each of which were categorized into RT vs no-RT groups. Adding RT did not improve OS or disease-free survival (DFS) in any group [8].
A National Cancer Database (NCDB) analysis of 934 patients with pathological T2N0 oral tongue cancers from 2004-2013 was performed to determine whether lesions with >5 mm DOI benefitted from receiving PORT [9]. Six hundred and seventy-seven (72.5%) patients had surgery alone and 257 (27.5%) received surgery plus PORT. Thirty-four (13.4%) received chemotherapy in addition to surgery and PORT. With a median follow-up of 28.4 months +/-10.4 months, the three-year OS was 81.3%. In multivariate analysis (MVA), adding PORT did not improve OS, even for patients with >5 mm DOI (p = 0.769).
This study evaluates the potential benefit of PORT in pT1-2N0 (stage I and II) oral tongue cancers with a DOI >4 mm.

Materials And Methods
The NCDB is a national oncology database and the data represents >70% newly diagnosed cancer cases and >34 million historical records [10]. This study was deemed to be exempt as per our Institutional review board. Patients with AJCC stage I and II oral tongue cancer diagnosed between 2006 and 2013 were queried. Inclusion criteria entailed oral cavity tumors (tongue) with wide excision, stage I and II, histology codes 8052, 8070-8078, and 8083. Patients receiving chemotherapy, had an OS less than six months, underwent any RT other than EBRT, or any residual tumor after surgical resection, were excluded from analysis.
Patients were stratified by receipt of PORT, elective neck dissection, and extent of tumor DOI (≤4 mm or >4 mm). Kaplan-Meier analysis was performed to compare OS (using the log-rank test) between patients receiving and not receiving PORT. Multivariable Cox proportional hazards regression model was used to evaluate the independent effect of PORT on OS, while controlling for tumor DOI and other clinical characteristics. The patient inclusion flow diagram is shown in Figure 1.

Results
Patient characteristics are summarized in Table 1 and comparisons of patients in the <4 mm DOI and >4 mm DOI groups are shown in Tables 2, 3. Ninety-eight (10.5%) patients received RT while 841 (89.5%) did not. Overall, a greater percentage of patients did not receive PORT. This trend persevered among each tumor stage. Additionally, African Americans were less likely to undergo PORT. In addition, the patients not receiving PORT were more likely to have LVI in the <4 mm category.     (Figure 3), and pathological stage (p = 0.108) (Figure 4) among patients with a tumor DOI >4 mm.

Discussion
This population-based study of stage I and II oral tongue cancers showed a survival benefit of elective neck dissection in patients with DOI >4 mm, but no benefit of adding adjuvant RT in regard of less of DOI.
Numerous studies identify DOI as a poor prognostic factor. A retrospective Japanese study of 337 stage I-II tongue cancer patients undergoing surgical resection revealed that T stage, DOI (cut-off was 4 mm), tumor budding (the presence of a single cancer cell or cluster of less than five cancer cells at the invasive front) and adjacent tissue at the invasive front are predictive of delayed neck metastasis [11].
Although 4 mm is commonly considered the DOI cut off for significance, a retrospective study of DOI cut-off points in previously untreated early stage oral tongue cancers showed 7.25 mm to be most predictive of occult nodal metastasis, 8 mm for OS and DFS [3]. In another  Limitations of our study include its retrospective nature, relatively small number in the total group receiving RT and lack of data on details of treatment such as technique of RT, use of image guidance and dose, local control and toxicity. Select studies evaluating DOI as a prognostic factor are listed in Table 6. In clinical practice, DOI does dictate neck dissection based on risk of neck metastasis although we show survival benefit with END in DOI > 4 mm. RT is associated with significant side effects including mucositis, pain, dysphagia, necrosis, dry mouth and loss of taste, and can be avoided for early stage tongue cancers.

Conclusions
Our study is the first large population-based study of both stage I and II oral cavity cancers to show addition of elective neck irradiation for tumors >4 mm does not improve survival. However, elective neck dissection in oral tongue cancers with DOI >4 mm confers a positive