Invasion In Early Colorectal Cancer Research Articles

Diagnosis of Depth of Submucosal Invasion in Colorectal Cancer with AI Using Deep Learning.

The submucosal invasion depth predicts prognosis in early colorectal cancer. Although colorectal cancer with shallow submucosal invasion can be treated via endoscopic resection, colorectal cancer with deep submucosal invasion requires surgical colectomy. However, accurately diagnosing the depth of submucosal invasion via endoscopy is difficult. We developed a tool to diagnose the depth of submucosal invasion in early colorectal cancer using artificial intelligence. We reviewed data from 196 patients who had undergone a preoperative colonoscopy at the Osaka University Hospital and Osaka International Cancer Institute between 2011 and 2018 and were diagnosed pathologically as having shallow submucosal invasion or deep submucosal invasion colorectal cancer. A convolutional neural network for predicting invasion depth was constructed using 706 images from 91 patients between 2011 and 2015 as the training dataset. The diagnostic accuracy of the constructed convolutional neural network was evaluated using 394 images from 49 patients between 2016 and 2017 as the validation dataset. We also prospectively tested the tool from 56 patients in 2018 with suspected early-stage colorectal cancer. The sensitivity, specificity, accuracy, and area under the curve of the convolutional neural network for diagnosing deep submucosal invasion colorectal cancer were 87.2% (258/296), 35.7% (35/98), 74.4% (293/394), and 0.758, respectively. The positive predictive value was 84.4% (356/422) and the sensitivity was 75.7% (356/470) in the test set. The diagnostic accuracy of the constructed convolutional neural network seemed to be as high as that of a skilled endoscopist. Thus, endoscopic image recognition by deep learning may be able to predict the submucosal invasion depth in early-stage colorectal cancer in clinical practice.

Estimation of invasion depth of early colorectal cancer using EUS and NBI-ME: a meta-analysis

Endoscopic ultrasonography (EUS) and narrow band imaging-magnifying endoscopy (NBI-ME) are often used as diagnostic tools to estimate the depth of invasion in early colorectal cancer (CRC). The aim of this study was to compare NBI-ME with EUS in distinguishing between slight submucosal invasion (invasion depth < 1000μm) and massive submucosal invasion in patients with early CRC, since slight submucosal invasion is currently considered as an indication for endoscopic resection. For this meta-analysis, relevant studies were identified from PubMed, Embase, Web of Science, Scopus and the Cochrane Library databases between January 1997 and September 2016. Data on the yield of tumors were extracted, pooled, and analyzed by stata12.0 software. The sensitivity, specificity, positive likelihood ratio and negative likelihood ratio in differentiating slight submucosal invasion from massive submucosal invasion were calculated for both diagnostic modalities. Sixteen studies involving 2197 lesions were included: nine were studies on EUS and 7 were studies on NBI-ME. The pooled sensitivity of EUS was 0.902 (95% CI 0.863-0.930), the specificity was 0.877 (95% CI 0.810-0.922), the positive likelihood ratio was 7.314 (95% CI 4.551-11.755) and the negative likelihood ratio was 0.112 (95% CI 0.076-0.164). The pooled sensitivity and specificity of NBI-ME were 0.981 (95% CI 0.949-0.993) and 0.651 (95% CI 0.600-0.699), respectively, the positive likelihood ratio was 2.815 (95% CI 2.432-3.258) and the negative likelihood ratio was 0.029 (95% CI 0.010-0.080). The sensitivity tended to be higher in ME-NBI than EUS for early CRC with slight submucosal invasion, whereas the specificity was significantly lower in NBI-ME than in EUS.

Endoscopic ultrasonography combined with magnifying endoscopy with narrow band imaging for the diagnostic value of early colorectal cancer

Objective To investigate the diagnostic value of endoscopic ultrasonography combined with magnifying endoscopy with narrow band imaging (NBI-ME) in the diagnosis of early colorectal cancer. Methods A total of 187 patients with early colorectal cancer who were treated in Cancer Hospital Chinese Academy Medical Sciences and Peking Union Medical College from January 2015 to December 2017 were retrospectively analyzed. The patients were examined with NBI-ME along with endoscopic ultrasonography before endoscopic or surgical treatment, and the depth of invasion was judged. Finally, postoperative pathology was taken as the gold standard to evaluate the value of endoscopic ultrasonography combined with NBI-ME in the diagnosis of early colorectal cancer. Results The histopathological results were treated as the golden standard. The diagnostic accuracy rate of endoscopic ultrasonography and endoscopic ultrasonography combined with NBI-ME in 187 patients with early colorectal cancer was 84.0 % (157/187), 92.5 % (173/187) respectively. Overall accuracy rate of the invasive depth of NBI-ME in the patients with early colorectal cancer was 87.2 % (166/187). Sensitivity of endoscopic ultrasonography combined with NBI-ME to Tm, Tsm1 and ≥Tsm2 in early colorectal cancer pathological tissues was 92.4 %, 90.4 %, 93.3 % respectively, and the corresponding specificity was 93.9 %, 90.1 %, 90.9 % respectively. The diagnostic accuracy rate of endoscopic ultrasonography combined with NBI-ME was higher than that of endoscopic ultrasonography, and there was a significant difference (χ 2 = 6.594, P = 0.010 2). Conclusion Endoscopic ultrasonography combined with NBI-ME has a high accuracy rate for evaluating the depth of invasion in early colorectal cancer, which contributes to a good application value for preoperative evaluation of early colorectal cancer. Key words: Colorectal neoplasms; Early diagnosis; Endoscopic ultrasonography; Magnifying endoscopy; Narrow band imaging

Lifting versus Non-Lifting Sign for Prediction of Depth of Invasion of Early Colorectal Cancer

Submucosal (sm) invasion by colorectal cancer can be classified as slight (depth of invasion < 1000 µm) or massive (depth of invasion > 1000 µm). Lesions classified as sm1 and sm2 are elevated in response to sm injection, whereas lesions classified as sm3 are not elevated. When sm invasion is slight, there is negligible risk of lymphatic metastasis, and cure can be achieved by endoscopic resection alone. However, when sm invasion is massive, the risk of lymphatic metastasis is substantial, increasing with the depth of invasion. The authors previously reported that a test based on lifting indicating sm1 or sm2 lesions would have a sensitivity of 83% and a specificity of 100%, whereas a test based on nonlifting indicating sm3 lesions would have a sensitivity of 100% and a specificity of 83%. Here the evaluation of these two types of tumor, classified as lifting and nonlifting lesions, is demonstrated. This article is part of an expert video encyclopedia.

Are we accurately measuring the depth of the submucosal invasion in early colorectal cancer by equating the Kikuchi submucosa levels with distances measured in fractions of a millimetre?

Are we accurately measuring the depth of the submucosal invasion in early colorectal cancer by equating the <scp>K</scp>ikuchi submucosa levels with distances measured in fractions of a millimetre?

Clinical Usefulness of Magnifying Chromoendoscopy and Magnifying Narrow Band Imaging Endoscopy for Predicting the Submucosal Invasion of Early Colorectal Cancers

Clinical Usefulness of Magnifying Chromoendoscopy and Magnifying Narrow Band Imaging Endoscopy for Predicting the Submucosal Invasion of Early Colorectal Cancers

Preoperative Evaluation of Early Colorectal Cancer Using an Ultrasound Mini Probe

We investigated whether endoscopic ultrasonography (EUS) can assess the depth of invasion of early colorectal cancer exhibiting the V pit pattern on magnifying endoscopy with submucosal invasion of 1,000 μm or deeper. Among 38 colorectal tumors exhibiting the V pit pattern on magnifying endoscopy, the findings on EUS with a mini-probe (15 MHz) were compared with histopathological findings. The diagnostic accuracy, sensitivity and specificity of EUS were examined separately in tumors exhibiting the Vi or V5 pit pattern. Diagnostic accuracy of EUS for cancers exhibiting the Vi pit pattern on magnifying endoscopy with submucosal invasion of 1,000 μm or deeper was 9/15 (60%). Sensitivity was 90%, specificity 14.3%, positive predictive value 31.7% and negative predictive value 76.3%. Diagnostic accuracy of EUS for cancers exhibiting the VN pit pattern on magnify-ing endoscopy with submucosal invasion of 1,000 μm or deeper was 13/18 (72%). The sensitivity and specificity of EUS were 100% and 37.5%, respectively. EUS tended to diagnose the invasion depth of cancer with submucosal invasion exhibiting the V pit pattern as deeper than it actually was. EUS accurately diagnosed early colorectal cancer with shallow invasion exhibiting the VN pit pattern and surgery was avoided.

Detection of desmoplastic reaction in biopsy specimens is useful for predicting the depth of invasion of early colorectal cancer: a Japanese collaborative study

We have previously demonstrated a relationship between the depth of submucosal invasion (SM depth) and the frequency of lymph node metastasis in resected submucosal invasive colorectal cancers (SICRCs). Here, we assessed the desmoplastic reaction (DR) in pretreatment biopsy specimens of SICRC to predict the SM depth. A total of 359 patients with SICRCs, who had undergone surgical or endoscopic mucosal resection, were enrolled. The SM depth of the SICRC lesions was evaluated according to the procedure established by the Japanese Society for Cancer of the Colon and Rectum, and the patients' corresponding pretreatment biopsy specimens were examined histologically to evaluate the prevalence of DR. For pedunculated SICRCs, the prevalence of DR in pretreatment biopsy specimens was significantly higher in moderately differentiated than in well-differentiated adenocarcinomas, but was not significantly related to SM depth. For nonpedunculated SICRCs, the prevalence of DR in pretreatment biopsy specimens was significantly related to histological type, tumor size, and SM depth. When non-pedunculated SICRCs were further divided using a specific cutoff value of 1000 μm for SM depth, the DR positivity rate in pretreatment biopsy specimens was significantly higher in SICRCs with an SM depth of ≥1000 μm (termed "SM massive CRCs") than in cases where the SM depth was <1000 μm (termed "SM slight CRCs"). Detection of DR in pretreatment biopsy specimens is useful for the prediction of SM depth in nonpedunculated SICRCs, and may be useful for the selection of such cases that would be treatable by endoscopic mucosal resection and endoscopic submucosal dissection (EMR/ESD).

6958 Daignosis of submucosal carcinomatous infiltration by microscanner in early colorectal cancer.

AIM: To evaluate the efficacy of the microscanner (MS) in diagnosing the depth of submucosal invasion in early colorectal cancer. PATIENTS: One hundred seventy-four patients with colorectal cancers including 80 protruding lesions (mucosal cancer (m) 41, submucosal cancer (sm) 39) and 94 superficial lesions (m, 41; sm, 52). METHODS: The diagnostic accuracy of depth of invasion in early colorectal cancer by MS was analyzed. The MS diagnosis was made based on the extent of a tumor echo into the third hyperechoic layer of the bowel wall which is considered to be the submucosa. Carcinomas which showed involvement of the submucosal layer were divided into two groups according to the histological degree of invasion: sm1, a group in which the depth of invasion was 1/3 of the submucosal layer or less; sm2-3, a group in which the depth of invasion of the submucosa was over 1/3. The MS diagnosis was compared with the histological diagnosis of the depth of invasion in each lesion. RESULTS: 1) Visualization of the lesion by MS was possible in 87% (152/174) of the lesions, i.e., 81% in protruding lesions and 92% in superficial lesions. 2) The accuracy of MS in diagnosing the depth of invasion in all lesions, in protruding lesions, and in superficial lesions was 93%, 88%, and 95%, respectively. 3) Diagnostic accuracy of MS as to the degree of submucosal invasion was 91% in m and sm1 cancer and 96% in sm2-3 cancer. CONCLUSIONS: MS is useful for the diagnosis of the depth of invasion in early colorectal cancers. Improvement of the accuracy in diagnosing protruding lesions is required.

Correlation of lifting versus non-lifting and microscopic depth of invasion in early colorectal cancer

Background: The non-lifting sign is considered a contraindication to endoscopic resection. Our objective was to investigate whether lifting or nonlifting of a lesion is determined by the volume of normal submucosal tissue. Methods: We measured the thickness of the submucosa and examined the relation between submucosal invasion and lesion elevation induced by submucosal injection in 60 patients with colorectal cancer with evidence of submucosal invasion. Extent of tumor elevation was classified into two groups: A, lifting; B, non-lifting. Submucosal invasion was classified as sm1, sm2, or sm3. The distance between the carcinoma and the line of resection and that between the carcinoma and the muscularis propria were measured. Results: Of 31 sm1 lesions, 29 (93.5%) were group A. All 6 sm3 lesions were group B. All lesions in group A had a value for the distance between carcinoma and muscularis propria of more than 1000 μm. Group B lesions with sm3 invasion had distances of only 105 to 750 μm. Conclusion: Lesions classified as sm3 do not elevate in response to submucosal injection, and lesions that become elevated on injection can be resected endoscopically because they are sm1 or sm2 and have a thickness of normal submucosa of more than 1000 μm. (Gastrointest Endosc 1999;50:329-33.)