Errors in Measurements of Colorectal Cancer Invasion in the Submucosal Layer Caused by Preparation of Pathological Specimens : An Analysis Using Simulation

The degree of cancer invasion into the submucosal (sm) layer has been investigated as a risk factor for lymph node metastasis in colorectal cancer. Depth, horizontal diameter, and area of cancer invasion were evaluated based on the assumption that the cross section was maximum. If the cross section does not represent the maximum section, errors may arise in these measurements. However, no analyses have been made regarding the magnitude of these errors. Using simulation, we first investigated the errors in the measurement values of shape of cancer invasion that can arise from preparation, that is, the errors resulting from the values obtained from the cross section and from the maximum section. Subsequently, we studied these errors in clinical cases. The shape of cancer invasion into the sm layer was postulated to be a half-round sphere, and measurement values were set to 100 in the maximum section. Depth, horizontal diameter, and area of the cancer invasion were calculated when the blade was placed off-center by one-half or one-quarter of the radius with five values of angle. When the specimens were sliced at one-half of the radius apart from the center, depth may be 67 to 138, horizontal diameter may be 87 to 96, and area may be 55 to 142. These errors may be 83 to 118, 97 to 99, and 77 to 122, respectively, when the specimens were sliced at one-quarter of the radius apart from the center. The errors were the least for horizontal diameter in either of the cases. In 66 clinical cases that, from the beginning, had undergone resection of the intestine, horizontal diameter of the sm invasive lesion alone was significantly correlated with the presence or absence of lymph node metastasis. From these results, when discussing measurement values of cancer invasion as risk factors for lymph node metastasis, we should take into consideration the fact that errors may arise from preparing the specimens, and that the error was minimum in horizontal diameter.