Introduction It is recognised that following rectal cancer resection, a threatened or involved circumferential resection margin (CRM) is associated with increased local recurrence rates. However the significance of a pathological positive lymph node (LN) threatening or involving the CRM remains unclear. The aim of this study was to investigate whether a CRM threatened with a positive lymph node was as significant as for a threatened margin with tumour. The effect of a positive margin on local recurrence, distant recurrence, overall survival and disease-free survival is examined.

Method From April 2005 to July 2013, a prospectively collected database was used, from a single pathologist with a particular interest in rectal cancer. Data was collected on patients with a threatened or involved CRM, defined as cancer within 1mm of the margin. Data was collected on demographics, type of surgery, lymph node yield, the type of CRM involvement (LN/tumour), local and distant recurrence (LR and DR), overall survival (OS) and disease-free survival (DFS).

Results From a total of 303 patients, complete data was available for 280 patients: 195male, 86 female; median age 69 (range 23–89). 90 patients had APR and 191 anterior resection. Median LN yield was 41 (range 9–122) 33/281 patients (11.7%) had a positive margin: 16 patients had tumour at the CRM, 13 patients had involved LN and 4 patients involved nerve or blood vessels. Median follow-up was 30 months (range 0–80). In patients with a positive CRM, LR occurred in 6/33 (18.1%) (1 from LN and 5 from tumour) and DR in 12/33 (36.3%) (5 from LN and 7 from tumour). In patients with negative CRM, LR was 17/248 (6.9%) and DR was 26/248 (10.5%). Kaplan Meier predicted 3 and 5-year OS for negative CRM was 84.6 and 75.1% vs. 74.1% and 49.4% for LN at CRM versus 37.5% and 20.0% for tumour at CRM. OS was significantly lower for tumour at CRM than negative CRM (p = 0.001) but no significant difference for LN at CRM (p = 0.209). Kaplan Meier predicted 3 and 5-year DFS for negative CRM was 76.4% and 65.3% vs. 53.8% and 43.1% for LN at CRM versus 25% and 18.8%for tumour at CRM. DFS was significantly lower for tumour at CRM than negative CRM (p = 0.001) but no significant difference for LN at CRM (p = 0.075).

Conclusion A positive LN at the CRM is not associated with high rates of local recurrence (8%) but did result in distant recurrence in 38% patients. OS and DFS were worse for patients with a positive CRM due to tumour at the margin compared with patients with a negative CRM. However, despite a trend towards lower survival, having a positive LN at the CRM is not associated with significantly reduced OS or DFS.

Disclosure of interest None Declared.