Improved survival for patients with clinically T1/T2, N0 tongue tumors undergoing a prophylactic neck dissection

BACKGROUND: Prophylactic surgical treatment of the neck in "early tongue tumors" is a controversial issue. METHODS: From a database of 226 patients with squamous cell carcinoma of the tongue treated at Canniesburn Hospital, Glasgow, U.K., between 1980 and 1996, a total of 137 patients with a minimum follow up of 24 months or until death were clinically identified as being T1/T2, N0 (UICC) when first seen. These patients were divided into three groups according to the management of the neck; 53 patients did not have a neck dissection at any time (NKD0), 47 patients underwent a synchronous neck dissection at the time of treatment of the primary (NKDS), and 37 patients subsequently required a metachronous neck dissection when lymph node metastasis became clinically apparent (NKDM). These three groups were compared with respect to age, sex, site, duration of symptoms, previous treatment (if any), initial treatment protocol, resection margin, type of neck dissection (if any), loco-regional recurrence, systemic escape, number of positive lymph nodes, and presence of extracapsular spread. Disease-related survival was calculated using Kaplan-Meier survival curves with logrank test and chi-square statistical analysis. RESULTS: The pT stage was upgraded to T3/4 in 3/53 patients (6%) of the NKD0 group, 11/47 patients (23%) of the NKDS group, and 2/37 patients (5%) of the NKDM group (p < 0.001). The 5-year determinate survival rates for the three groups were: NKD0 59.7%, NKDS 80.5%, NKDM 44.8%, and (NKD0 + NKDM) 53.6% with a statistically significant improvement in survival for NKDS vs NKDM (logrank 10.58, p = 0.001) and for NKDS vs (NKD0 + NKDM) (logrank 6.06, p = 0.014). The incidences of positive nodes in the NKDS and NKDM groups were 18/47 patients (38%) and 32/37 patients (86%) respectively. Neck positive patients in the NKDM group had a significantly greater number of positive nodes in comparison with N positive patients in the NKDS group (chi trend, p = 0.001), a higher incidence of extracapsular spread, 30/32 vs 9/18 (chi test, p < 0. 0001), and decreased survival. The incidence of occult cervical metastasis for the whole group was 41%. CONCLUSION: Patients with clinical T1/2, N0 tongue tumors who underwent a synchronous neck dissection had an improved survival outcome even though as a group they had a higher incidence of occult metastasis, relatively more T2 lesions, a worse pT stage, and had more posterior third lesions requiring more difficult initial surgery. Tongue tumors have a high incidence of subclinical nodal disease, which is less curable when it presents clinically. The information gleaned from the nodal status allows a more informed plan of adjuvant therapy.