Yuko Hiraga

Endoscopic treatment of submucosal invasive colorectal carcinoma with special reference to risk factors for lymph node metastasis

A clinicopathological analysis of the risk factors for lymph node metastasis was performed in 177 patients with submucosal invasive colorectal carcinoma (CRC). The submucosal deepest invasive portion was histologically subclassified as well (W), moderately (M), or poorly (Por) differentiated. M type was further subdivided into moderately-well (Mw) and moderatelypoorly (Mp) differentiated. The pattern of tumor growth was classified as polypoid growth (PG) and non-polypoid growth (NPG). Lymph node metastasis was detected in 21 (12%) of the 177 patients. Macroscopically, type IIc and IIa+IIc lesions showed a significantly higher incidence of lymph node metastasis (44% and 30%) than type IIa and I (4% and 8%). Regarding the histologic subclassification, Por and Mp lesions showed a significantly higher incidence of lymph node metastasis (67% and 37%) than W and Mw lesions (4% and 14%). NPG tumors showed a significantly higher incidence of lymph node metastasis (29%) than PG tumors (7%). The depth of submucosal invastion and lymphatic invasion (ly) were also significantly correlated with incidence of lymph node metastasis (submucosal scanty (sm-s) invasion 4%, massive invasion 20%; ly(+) 23%, ly(−) 5%). None of the lesions with both sm-s invasion and of W or Mw type showed lymph node metastasis. These results indicate that submucosal invasive CRC with both sm-s invasion and of W or Mw type, which shows no ly, is the appropriate indication for endoscopic curative treatment.

Immunoreactive MUC1 expression at the deepest invasive portion correlates with prognosis of colorectal cancer

This study sought to examine the relationship between MUC1 expression at the deepest invasive portion, invasive/metastatic potential, and prognosis of colorectal cancer in relation to cellular proliferation. MUC1 expression was detected immunohistochemically using KL-6 antibody (anti-MUC1 monoclonal antibody) in 100 surgically resected specimens of advanced colorectal cancer. Distinct staining of the luminal surfaces, defined as positive immunoreactive (IR)-MUC1 expression, was seen in more than 30% of the tumor cells at the deepest invasive portion. The proliferating cell nuclear antigen labeling index (PCNA-LI) was also examined in the same areas. IR-MUC1 expression was detected in 71 (71%) of 100 lesions. Lesions with lymphatic or venous invasion showed a significantly higher incidence of IR-MUC1 expression than those without lymphatic or venous invasion (80 vs. 42% and 82 vs. 61%, respectively). Lesions with lymph node metastasis showed a significantly higher incidence of IR-MUC1 expression than those without lymph node metastasis (88 vs. 53%). Lesions with liver metastasis showed a significantly higher incidence of IR-MUC1 expression than those without liver metastasis (92 vs. 59%). Dukes’ stage was also significantly correlated with IR-MUC1 expression. The incidence of IR-MUC1 expression did not significantly differ with regard to histologic subclassification and depth of invasion. There was no significant correlation between IR-MUC1 expression and the PCNA-LI. IR-MUC1 expression at the deepest invasive portion revealed a significant correlation with prognosis; furthermore, in patients with better differentiated lesions, in those with lesions confined to muscularis propria or subserosa (subadventitial) invasion, in those with Dukes’ B and C, or in those undergoing curative resection, IR-MUC1 expression significantly correlated with prognosis. Patients with high PCNA-LI lesions showed a significantly poorer prognosis than those with low PCNA-LI lesions. Only in patients undergoing curative resection, patients with IR-MUC1-positive and high PCNA-LI lesions showed a significantly poorer prognosis than those with IR-MUC1-negative and low PCNA-LI lesions. The significant risk factors in the order of poorer prognosis in patients undergoing curative resection by the multivariate analysis were the histologic grade (moderately–poorly, poorly or mucinous adenocarcinomas), IR-MUC1 expression, and lymph node metastasis. These results indicate that IR-MUC1 expression is an important predictor of the metastatic potential and the prognosis of colorectal cancer, independent of histologic grade, depth of invasion or cellular proliferative activity. Combined analysis of IR-MUC1 and histologic grade, and combined expression of IR-MUC1 and PCNA at the deepest invasive portion are especially useful in predicting colorectal cancer prognosis.

Proliferating cell nuclear antigen expression correlates with the metastatic potential of submucosal invasive colorectal carcinoma

To examine the malignant potential of submucosal invasive colorectal carcinoma, the relationship between proliferating cell nuclear antigen (PCNA) expression and clinicopathologic risk factors for lymph node metastasis was studied in 149 patients with submucosal invasive colorectal carcinoma. The depth of submucosal invasion was classified as scanty or massive. Histologic subclassification at the submucosal deepest invasive portion was done as follows: well differentiated (W), moderately well differentiated (Mw), moderately poorly differentiated (Mp) or poorly differentiated (Por). Tumor growth was divided into polypoid growth and nonpolypoid growth. The PCNA expression (labeling index, LI) was examined at the submucosal deepest invasive portion. The PCNA-LI of tumors showing lymph node metastasis (mean, 56.5 +/- 19.0%) was significantly higher than that of tumors without lymph node metastasis (mean, 41.5 +/- 19.3%; p < 0.01). The PCNA-LI of Mp tumors (mean, 57.7 +/- 16.5%) was significantly higher than that of W (mean, 38.5 +/- 19.0%; p < 0.05) and Mw (mean, 43.7 +/- 19.1%; p < 0.05) tumors. The PCNA-LI of tumors without adenomatous features (mean, 47.9 +/- 20.5%) was significantly higher than that of tumors with such features (mean, 37.1 +/- 17.1%; p < 0.05). The PCNA-LI was not correlated with other risk factors for lymph node metastasis, such as lymphatic invasion, depth of submucosal invasion, macroscopic type, and growth pattern. These results indicate that the PCNA-LI may be useful marker for predicting the potential metastases to lymph nodes in submucosal invasive colorectal carcinoma, while the proliferative activity of cancer cells correlates with the degree of the differentiation in the area of deepest invasion.

Sa1585 The Corresponding Rate of Magnification Observation Findings of Narrow Band Imaging (NBI) and Flexible Spectral Imaging Color Enhancement (FICE) in Gastric Tumors and in Colorectal Lesions

The Corresponding Rate of Magnification Observation Findings of Narrow Band Imaging (NBI) and Flexible Spectral Imaging Color Enhancement (FICE) in Gastric Tumors and in Colorectal Lesions Yuko Hiraga*, Masaaki Sumioka, Tomoki Hiramoto, Morihisa Akagi, Daisuke Komichi, Yukio Kuwada, Mikiya Kitamoto, Chiyuki Watanabe, Takashi Nishisaka Department of endoscopy, Hiroshima prefectural Hospital, Hirosima, Japan; Department of gastroenterology, Hiroshima prefectural Hospital, Hirosima, Japan; Department of pathology, Hiroshima prefectural Hospital, Hirosima, Japan Introduction/Objectives: There is a difference in how to see the microvessels by NBI and FICE for the reason that the method of the image enhancement is different, so that it has not been clear whether NBI magnification classification can apply to FICE also. We have been able to use both the instruments in our hospital, so that we were able to do the comparative study of NBI findings and FICE ones at the same lesions for the purpose of examining whether the classification of NBI magnification fits FICE magnification findings also. Methods: We examined a total of 161 lesions (22 gastric tumors, 139 colorectal lesions) that observed by both NBI magnification and FICE magnification between April 2009 and August 2011, treated by endoscopic or surgical resection and diagnosed clinicopathologically. The interval of NBI and FICE examinations was 62.3days (shortest one day, longest 475 days). We examined on the comparison with the NBI findings and the FICE ones and on the relationship between them and histological diagnosis. We classified NBI/FICE magnifying findings according to whether surface pattern was visible or not and whether that was regular or not, and whether microvessels were regular or not and whether avascular area appeared or not. We followed VS classification system (Vascular pattern; regular, irregular or absent, Surface pattern; regular, irregular or absent) in gastric tumors, and followed Hiroshima classification (A, B, C1, C2 or C3) in colorectal lesions. Results: 1) Gastric tumors: In surface pattern diagnosis, FICE findings corresponded with NBI ones at all 22 lesions. But in microvessels pattern diagnosis, evaluations differed in two lesions (regular-irregular, regular-absent). The corresponding rate of NBI and FICE magnification findings was overall 90.9% in gastric tumors, in each of histological diagnosis, it was 91.7% in adenomas, and 90.0% in cancers. 2) Colorectal lesions: The diagnoses of FICE findings differed from NBI ones in 16 lesions, concretely A-B 1 lesion, B-C1 13 lesions and C1-C2 2 lesions. The diagnoses of C3-type, with invisible surface pattern and fragmental microvessels or avascular area, corresponded in all 6 lesions. The corresponding rate of NBI and FICE magnification findings was overall 88.5% in colorectal lesions, in each of histological diagnosis, it was 100% in non-neoplasm, 89.6% in adenomas, 90.5% in intramucosal cancers, and 66.7% in invasive cancers. Conclusions: In gastric tumors and colorectal lesions, the diagnoses between NBI magnification and FICE ones almost corresponded. We were able to apply the classification of NBI magnification to also FICE magnification by using the classification that paid attention to a visibility of surface pattern and to an irregularity of microvessels diameter and distribution.