We studied 42 patients, consisting of 11 stage I/II patients who were node-negative (LN−) and 31 stage II/III patients who were node-positive (LN+). In the LN− patients, detection of occult neoplastic cells (ONCs) in lymph nodes had a sensitivity of 0.0% (0/5), a false-positive (FP) rate of 33.3% (2/6), a specificity of 66.7% (4/6), a false-negative (FN) rate of 100% (5/5), a positive predictive value (PPV) of 0.0% (0/2), and a negative predictive value (NPV) of 44.4% (4/9). In the LN+ patients, the sensitivity of ONCs was 25.0% (5/20), FP rate was 36.4% (4/11), specificity was 63.6% (7/11), FN rate was 75.0% (15/20), PPV was 55.6% (5/9), and NPV was 31.8% (7/22). In LN− patients, positivity for at least 2 of the 3 high-risk criteria had a sensitivity of 20.0% (1/5), FP rate of 16.7% (1/6), specificity of 83.3% (5/6), FN rate of 80.0% (4/5), PPV of 50.0% (1/2), and NPV of 55.6% (5/9). In LN+ patients, these criteria had a sensitivity of 75.0% (15/20), FP rate of 9.1% (1/11), specificity of 90.9% (10/11), FN rate of 25.0% (5/20), PPV of 93.8% (15/16), and NPV of 66.7% (10/15). These results suggest that the high-risk criteria may be useful for predicting recurrence or metastasis in stage II/III lymph node-positive patients with esophageal cancer.

Related Articles