MINUTE SUPERFICIAL‐TYPE DE NOVO COLORECTAL CARCINOMA WITH SUBMUCOSAL INVASION

A 59‐year‐old male was evaluated at the Foundation for Detection of Early Gastric Carcinoma after 4 years, following partial sigmoidectomy for early invasive cancer. Double‐contrast barium enema study and endoscopy revealed a new minute flat‐elevated lesion with an irregular central depression in the proximal sigmoid colon. Endoscopic mucosal resection was carried out with suspicious diagnosis of intramucosal colonic cancer. Macroscopically, the specimen constituted a flat‐elevated lesion measuring 4 mm in its largest diameter with a shallow irregular depression. Histologically, various‐sized vesicular nuclei and cribriform glands without adenomatous component were confirmed. A very small part of the lesion invaded into the submucosal layer. The diagnosis of carcinoma was made based on those histological findings, and this case was regarded as a minute superficial‐type de novo carcinoma. Radiographic and endoscopic findings were compared with the histological features in detail. Carcinoma was found at the depressed area on the surface of the lesion. Furthermore, the invasion of high‐grade carcinoma into the submucosal layer was located just beneath the deeper peripheral part of the depression which showed by colonoscopy. Consequently, a distinct depression on a flat‐elevated lesion may represent not only the presence of de novo cancer but also possible involvement of the deeper invasion.     

EFFICACY OF MAGNIFYING COLONOSCOPY FOR THE DIAGNOSIS OF COLORECTAL NEOPLASIA: COMPARISON WITH HISTOPATHOLOGICAL FINDINGS

Background: Some authors have reported a good correlation between the shape and arrangement in the ori?ce of mucosal crypt (pit pattern) in the diagnosis of colorectal lesions and histopathological ?ndings. However, there remains no good consensus on the de?nition of irregularity in con?gurations. Methods: We studied 110 colorectal neoplasia of 110 cases that were examined using a magnifying colonoscope and resected endoscopically or surgically between 2001 and 2003. We divided the lesions into two groups according to Fujii's classi?cation as follows: type Non‐V showed type III or IV by Kudo's classi?cation and irregular pits not occurring in a demarcated area; and type V showed irregular pits, which occurred in a demarcated area in addition to type V_N by Kudo. The diagnosis by this classi?cation correlated with the histopathological ?ndings. We also addressed the causes of misdiagnosis in comparison with histopathological ?ndings. Results: The diagnostic accuracy for Non‐V and V was 95.1% and 82.1%, respectively. Lesions classi?ed as V showed a signi?cant correlation. Histologically, carcinoma consisted of high‐grade atypia (P < 0.0001), an erosive change of surface epithelium (P < 0.0001) and the appearance of desmoplastic reactions (P = 0.002) in comparison with Non‐V. Misdiagnosis was likely due to differences in the grade of atypia between super?cial and deeper glands and explainable erosive changes on the surface, as well as simple misreading of the pit pattern. Conclusion: Diagnosis of colorectal neoplasia by magnifying colonoscope using our classi?cation was useful for evaluating the depth of invasion and correlated well with histopathological ?ndings.     

MID‐TERM PROGNOSIS AFTER ENDOSCOPIC RESECTION FOR SUBMUCOSAL COLORECTAL CARCINOMA: SUMMARY OF A MULTICENTER QUESTIONNAIRE SURVEY CONDUCTED BY THE COLORECTAL ENDOSCOPIC RESECTION STANDARDIZATION IMPLEMENTATION WORKING GROUP IN JAPANESE SOCIETY FOR CANCER OF THE COLON AND RECTUM

We carried out a retrospective questionnaire survey of 792 submucosal colorectal carcinoma (CRC) cases from 15 institutions affiliated with the Colorectal Endoscopic Resection Standardization Implementation Working Group in Japanese Society for Cancer of the Colon and Rectum. In these cases, endoscopic resection (ER) and surveillance was carried out without additional surgical resection. Local recurrence or metastasis was observed in 18 cases. Local submucosal recurrence was observed in 11 cases, and metastatic recurrence was observed in 13 cases. Among the 15 cases in which the depth of submucosal invasion was measured, two cases showed depth less than 1000 µm, which has other risk factors for metastasis. Metastatic recurrence was observed in the lung, liver, lymph node, bone, adrenal glands, and the brain; in some cases, metastatic recurrence was observed in multiple organs. Death due to primary disease was observed in six cases. The average interval between ER and recurrence was 19.7 ± 9.2 months. In 16 cases, recurrence was observed within 3 years after ER. Thus, validity of ER without additional surgical resection for cases with the conditions that the depth of submucosal invasion is less than 1000 µm and the histological grade is well or moderately differentiated adenocarcinoma with no lymphatic and venous involvement was proven.     

COMPARISON OF MAGNIFYING CHROMOENDOSCOPY AND NARROW‐BAND IMAGING IN ESTIMATION OF EARLY COLORECTAL CANCER INVASION DEPTH: A PILOT STUDY

Background: Several previous studies have identified narrow‐band imaging (NBI) with magnification as being useful in evaluating early colorectal cancer invasion depth, but comparative diagnostic accuracy of invasion depth between pit pattern analysis using magnifying chromoendoscopy and NBI remains unclear. The aim of this retrospective study was to compare NBI and pit pattern analysis using magnifying chromoendoscopy in estimating early colorectal cancer invasion depth and to assess interobserver agreement. Patients and Methods: We analyzed a total of 72 early colorectal cancers in 72 patients fulfilling the inclusion criteria. Each lesion image was subsequently reviewed by two experienced colonoscopists (A, B) and then classified clinically based on invasive/non‐invasive pattern and Sano's capillary pattern classification with a five‐point scale of confidence. Results: In terms of diagnostic accuracy with confidence for A and B, the areas under the receiver operating characteristics curve were 0.84 and 0.81 for pit pattern analysis and 0.82 and 0.79 for NBI, respectively. Interobserver agreement for the diagnosis of submucosal deep (>1000 µm) invasion was evaluated for both modalities and indicated substantial agreement with pit pattern analysis (κ = 0.63) and moderate agreement with NBI (κ = 0.44). Conclusion: Estimating invasion depth of early colorectal cancer using NBI appeared to have been comparable to pit pattern analysis, but there was greater interobserver variability using NBI.     

ENDOSCOPIC SUBMUCOSAL DISSECTION FOR THE RELIABLE EN BLOC RESECTION OF COLORECTAL MUCOSAL TUMORS

Scheduled piecemeal resection has been actively conducted for granular type laterally spreading tumor (LST‐G) in Japan, as long as a definite preoperative diagnosis is made. However, en bloc resection is desirable for depressed lesions (e.g. IIc lesion) as well as non‐granular type laterally spreading tumor (LST‐NG) since they have considerable high risk for submucoasl invasion and require precise histopathological evaluation. Endoscopic submucosal dissection (ESD) has been developed for the en bloc resection of mucosal tumors of gastrointestinal tract and widely applied especially in gastric lesions. Although the large intestine involves structural and technical difficulties, we conducted en bloc resection by ESD while exercising sorts of ingenuity for preparation; endoscopes, instruments, local injections, and others. ESD is a reliable technique that allows en bloc resection of gastrointestinal mucosal lesions, and even has a splendid possibility for the treatment of early stage colorectal cancer.     

ENDOSCOPIC ULTRASOUND‐GUIDED FINE‐NEEDLE ASPIRATION BIOPSY FOR THE DIAGNOSIS OF GASTROINTESTINAL SUBMUCOSAL TUMORS

Gastrointestinal submucosal tumors (SMT) detected by barium meal study or endoscopy include various kinds of diseases and various degrees of malignancy. Endoscopic ultrasonography (EUS) can provide useful information about the differentiation of intra‐ and extra‐wall lesions, location and originating layer, presumption of their histological nature, measurement of the actual size of the lesion, and the possibility of differentiating between a benign and a malignant lesion. However, EUS alone does not reveal the complete pathology. EUS fine‐needle aspiration biopsy (EUS‐FNAB) has been reported to be a useful tissue sampling method for pancreatic mass lesions, lymph nodes swelling, posterior mediastinal masses and also gastrointestinal submucosal tumors. The EUS‐FNAB procedure is effective not only for the differential diagnosis of benignancy and malignancy, but also for the specific histopathological nature of gastrointestinal SMT using immunohistochemical staining. When used with MIB‐1 (Ki‐67) staining, and gene analysis in case of gastrointestinal stromal tumor, EUS‐FNAB may indicate its prognosis and influence decisions regarding therapeutic strategy. Thus, EUS‐FNAB is an indispensable procedure in the diagnosis of SMT.     

ADVANTAGES OF USING THIN ENDOSCOPE‐ASSISTED ENDOSCOPIC SUBMUCOSAL DISSECTION TECHNIQUE FOR LARGE COLORECTAL TUMORS

Background: Our purpose was to evaluate the effectiveness of a newly developed non‐invasive traction technique known as thin endoscope‐assisted endoscopic submucosal dissection (TEA‐ESD) procedure for the removal of colorectal laterally spreading tumors (LST). Patients and Methods: A total of 37 LST located in the rectum and distal sigmoid colons of 37 patients were eligible for outcome analysis. Twenty‐one LST were treated with TEA‐ESD and were then retrospectively compared to 16 LST that had previously been treated with standard ESD. Tumor size, en bloc resection rate, procedure time, combined number of different electrical surgical knives used during each procedure and associated complications were evaluated in this case–control study. Results: There was no statistically significant difference in tumor size between the TEA‐ESD group and the ESD control group (43.6 ± 16 mm and 42.4 ± 14 mm, respectively). All LST were successfully resected en bloc in both groups. Procedure duration was shorter for the TEA‐ESD group than the ESD control group, although the difference was not statistically significant (96 ± 53 minutes vs 116 ± 74 minutes; P = 0.18). The percentage of cases in which only one electrical surgical knife was used during the entire procedure was significantly higher in the TEA‐ESD group compared to the ESD control group (85.7% vs 31.3%; P = 0.0005). There were no perforations in the TEA‐ESD group while the ESD control group experienced one perforation. At the present time, TEA‐ESD is limited to the rectum and distal sigmoid colon. Conclusion: It was technically easier, safer and more cost‐effective to perform ESD for LST in the rectum and the distal sigmoid colon using the newly developed TEA‐ESD traction technique.     

EVALUATION OF THE RISK FACTORS OF LYMPH NODE METASTASIS IN PT1 STAGE COLORECTAL CARCINOMA: INDICATION FOR AN ENDOSCOPIC MUCOSAL RESECTION

Aim: In the present study, we aimed to clarify the parameters that can be used for clinically relevant treatment decisions. Patients and methods: During the period from July 1985 to April 2005, 283 pT1 cancers were selected for this study. Risk factors for lymph node (LN) metastasis were evaluated as follows: endoscopic appearance, tumor size, location, lymphatic permeation, venous invasion, patterns of cancer invasion into the submucosal layer, and depth of vertical invasion in the submucosal layer. Results: Results of the logistic regression analysis from these significant parameters were as follows: infiltrating growth pattern (odds ratio: 12.63); lymphatic permeation positive (odds ratio: 5.726); sigmoid colon (odds ratio: 4.585); tumor size (odds ratio: 1.718). However, the leading edge of only one cancer with LN metastasis in the expanding growth group was also cribriform pattern. Conclusion: The invasive growth patterns of infiltrating growth, lymphatic permeation, tumor location of the sigmoid colon, and maximum tumor size were independent and significant risk factors for LN metastasis in our logistic regression analysis. In particular, the former three factors (infiltrating growth, lymphatic permeation and sigmoid colon) revealed high odds ratio and covered all cases of LN metastasis. From the results of our study, the indication for operative intervention after EMR in pT1 cancer is those lesions which possess at least one factor among the three (infiltrating growth, lymphatic permeation and sigmoid colon). Also, assessment of the leading edge with a cribriform pattern should be dealt with carefully.     

OPTIMUM NUMBER OF PERCUTANEOUS TRANSHEPATIC CHOLANGIOSCOPY‐GUIDED BIOPSIES FOR HISTOLOGIC DIAGNOSIS OF BILE DUCT CARCINOMA

Background: Percutaneous transhepatic cholangioscopy (PTCS)‐guided biopsy is used for the diagnosis of bile duct carcinoma, but the number of biopsy specimens required for diagnosis is unclear. The aim of this study was to clarify whether multiple PTCS‐guided biopsies are needed for accurate histologic diagnosis. Methods: We examined the relationships between size of the first biopsy specimen, endoscopic, cholangiographic, and pathologic features, and the presence of carcinoma in the first biopsy specimen of the primary lesion in 27 bile duct carcinomas. Results: Twenty‐six of 27 carcinomas (96%) were histologically diagnosed by PTCS‐guided biopsy; 20 (74%) were detected in the first biopsy specimen, six in the second or third biopsy specimen, and one was not detected in four biopsy specimens. Carcinomas with papillogranular mucosa by endoscopy, convex margins by cholangiography, or macroscopic types (except for sclerosing type) were detected on the first biopsy specimen more frequently than were others (15/15 vs 5/12, P <0.001; 13/13 vs 7/14, P <0.01; and 15/16 vs 5/11, P <0.01, respectively). There was no relationship between positivity for carcinoma and size of the first biopsy specimen, vascular dilatation by endoscopy, or histologic type. With the combination of preoperative endoscopy and cholangiography, main lesions with papillogranular mucosa and/or convex margins were proven to be carcinoma on the first biopsy specimen significantly more frequently than were others (17/17 vs 3/10, P <0.0001). Conclusion: If the main lesion contains neither papillogranular mucosa nor a convex margin, multiple PTCS‐guided biopsies should be performed in order to increase the sensitivity for diagnosing bile duct carcinoma.     

SELF‐EXPANDABLE METALLIC STENT FOR PATIENTS WITH NON‐RESECTABLE MALIGNANT COLORECTAL STRICTURE: REVIEW OF 102 CASES IN THE JAPANESE LITERATURE

Alternative procedures using endoscopy have been developed, one of which is treatment with self‐expandable metallic stents (SEMS). In Japan, as SEMS for colorectal stricture has not been approved by the public insurance system, esophageal stent is used for colon and rectum exceptionally as a colonic SEMS after obtaining informed consent from the patient. This situation is very different to other countries. In the present study, we review the Japanese medical literature to determine the current status, feasibility, and challenges remaining for SEMS to show the current status of SEMS usage for colonic strictures in Japan. We investigated SEMS for patients with non‐resectable malignant colorectal stricture in 102 Japanese case reports. Primary colorectal cancer comprised half of the cases. The insertion success rate was 100% and the clinical effectiveness rate was 93%. Restricture occurred in 12 cases (12%), and half of those cases were treated by stent in stent. Stent migration occurred in eight cases (8%) and perforation in two cases (2%). The range of SEMS insertion duration was 1 to 576 days (mean: 132 days, median 142 days). There were no deaths related to the procedure. This procedure allows patients to forgo colostomy and is cheap, safe and effective, with a short treatment time. This procedure is a viable palliative alternative to colostomy for patients with inoperative malignant colorectal stricture. Widespread application of the procedure has been hampered.