Variation in lymph node evaluation in rectal cancer: a Dutch nationwide population-based study

Abstract

BACKGROUND: For adequate staging and subsequent accurate estimation of prognosis, a sufficient number of lymph nodes (LNs) has to be evaluated. This study aimed to identify factors associated with adequate nodal evaluation and to determine its relationship with survival. METHODS: Data from all patients with stage I to III rectal carcinoma who underwent surgical treatment and who were diagnosed in the period 2000 to 2006 were retrieved from the Netherlands Cancer Registry. Multilevel logistic analysis was performed to examine the influence of relevant factors on the number of evaluated LNs. Kaplan-Meier and Cox regression analyses were used to analyze the association with overall survival. RESULTS: The number of evaluated LNs was determined for 10,788 (91%) of 11,818 tumors. Median number of evaluated LNs was 7, ranging from 4 to 11 between pathology laboratories. The proportion of patients with positive LNs increased with increasing number of evaluated LNs. Men, younger patients, tumors with deeper invasion and nodal involvement, patients without preoperative radiotherapy who underwent a low anterior resection, and patients whose LNs were evaluated in an academic pathology laboratory were more likely to have >/=12 LNs evaluated. After adding these factors to the model, unexplained variation between pathology laboratories and between hospitals remained. The overall survival increased with increasing number of evaluated LNs. CONCLUSIONS: A large variation in LN evaluation among patients with rectal cancer was revealed. Improvement in LN evaluation by both hospitals and pathology laboratories could improve staging, leading to more reliable estimation of prognosis.