Background: Multidetector-row computed tomography (MDCT) is widely used to predict pathological nodal status. However, an appropriate nodal size cutoff value to predict pathological nodal status has not been determined, and the impact of preoperative lymph node size on long-term outcomes is unclear.
Methods: This study included 137 gastric cancer patients with nodal involvement who underwent R0 gastrectomy between September 2002 and December 2006. Lymph nodes with a short-axis diameter of 10 mm or more as measured by MDCT were regarded as metastasized. An appropriate cutoff value with a high positive predictive value (PPV) and high specificity also was identified, and the subsequent clinicopathological characteristics and long-term outcomes were investigated.
Results: A cutoff value of 15 mm was found to be appropriate for grouping patients into large (≥15 mm) and small (<15 mm) lymph node metastasis (LLNM and SLNM) groups, with a high PPV (98.6 %) and specificity (99.8 %). There were no differences in clinicopathological characteristics between the groups except for pathological nodal status. In the LLNM group, the 5-year survival rate was 55 %, which was significantly lower than in the SLNM group (73.2 %; P = 0.008). After stratification by tumor depth, the same trend was observed in patients with pT3 disease (46.8 % vs. 72.7 %; P = 0.015) and those with pT4 disease (14.3 % vs. 64.8 %; P = 0.035).
Conclusions: Gastric cancer patients with lymph nodes measuring 15 mm or more preoperatively have worse long-term outcomes. These patients would therefore be suitable candidates for future clinical trials investigating the efficacy of neoadjuvant chemotherapies.