Background: We examined the sizes of lymph nodes and metastatic foci within the lymph nodes that affect false-positive and false-negative lymph node staging by positron emission tomography in lung cancer.
Methods: Preoperative positron emission tomography and computed tomography scans were performed for 564 lymph node stations in 80 patients with peripheral-type lung cancer. The sizes of both the lymph nodes and the metastatic foci within the lymph nodes were measured, and these measurements were compared with those obtained with positron emission tomography scanning. To establish general sizes of metastatic foci within the lymph nodes, 277 metastatic lymph nodes in operative specimens previously resected from another 111 patients with lung cancer were examined as a control.
Results: The sensitivity was significantly higher for positron emission tomography than for computed tomographic scanning (P =.026). The sizes of metastatic foci within lymph nodes that showed false-negative (n = 8) and true-positive (n = 28) with positron emission tomography ranged from 0.5 to 9 mm (3 +/- 1 mm) and from 4 to 18 mm (10 +/- 3 mm), respectively (P <.001). None of the metastatic foci smaller than 4 mm could be detected with positron emission tomography scanning. The review of the 277 previously resected metastatic lymph nodes showed that 89 (32%) had metastatic foci smaller than 4 mm. The sizes of true-positive (n = 28) and false-positive (n = 10) lymph nodes ranged from 6 to 15 mm (10 +/- 2 mm) and from 9 to 16 mm (12 +/- 2 mm), respectively (P <.01). None of the false-positive lymph nodes was smaller than 9 mm.
Conclusions: Although positron emission tomography was superior to computed tomography scanning in lymph node staging in lung cancer, positron emission tomography was unable to distinguish metastatic foci smaller than 4 mm, which were not unusual sizes for lymph node metastases in lung cancer. Positive lymph nodes with positron emission tomography smaller than 9 mm are likely to be true-positive rather than false-positive.