Su1629 Endoscopic Removal of Early Stage Colorectal Cancer in a ...

13 downloads 6180 Views 255KB Size Report
1Gastroenterology, University of Wisconsin School of Medicine and. Public Health ... FIT-Based Screening Population: Progress Still to Be Made. Els Wieten* ...
Abstracts

Figure 2.

Su1628 High Colorectal Cancer Screening Rates Can Be Achieved With Colonoscopy As The Dominant Screening Modality Jennifer M. Weiss*1, Mark E. Benson1, Deepak V. Gopal1, Perry Pickhardt2, David H. Kim2, Maureen A. Smith3,4, Patrick Pfau1 1 Gastroenterology, University of Wisconsin School of Medicine and Public Health, Madison, WI; 2Radiology, University of Wisconsin School of Medicine and Public Health, Madison, WI; 3Population Health Sciences, University of Wisconsin School of Medicine and Public Health, Madison, WI; 4Family Medicine, University of Wisconsin School of Medicine and Public Health, Madison, WI Background: Colorectal cancer (CRC) screening rates continue to be suboptimal with a recent plateau in national screening rates. Patients’ reluctance to get screened, particularly by colonoscopy, fear of invasive screening tests, and lack of access to screening have been suggested as reasons for low screening rates. Aim: We sought to determine CRC screening rates in a large unified health system where colonoscopy is the dominant screening modality. Methods: Billing and electronic medical record (EMR) data was prospectively collected over a 12 month period (January 2014 – December 2014) for adults eligible for CRC screening in a large health system. Data collection was limited to age appropriate patients (50-75 years) who were medically-managed by the physician group; determined by at least two primary care office visits in an outpatient, non-urgent care setting within the previous 36 months, with at least one of those visits in the prior 24 months. Patients were excluded if they had a history of a total colectomy based on ICD-9 codes and CPT codes. CRC screening completion was determined via EMR data by (a) fecal occult blood test (FIT/gFOBT) in the prior 12 months, (b) flexible sigmoidoscopy, double contrast barium enema, or computed tomographic colonography (CTC) in the past 5 years, or (c) colonoscopy in the prior 10 years. Results: 62,614 patients were eligible for CRC screening with 50,988 patients documented as meeting the above criteria for CRC screening completion, yielding an overall screening rate of 81.4%. Twenty-four percent of these patients had documented completed CRC screening outside of our health system meeting the above requirements. Of the patients screened within our system 91.5% (NZ35,555) were screened by colonoscopy, 4.5% (NZ1,751) completed a CTC, 3.4% (NZ1,341) completed a stool test (FIT/gFOBT), 0.5% (NZ192) completed a flexible sigmoidoscopy, and 2 patients were screened with a double contrast barium enema. Conclusions: 1) High CRC screening rates of greater than 80% can be achieved with colonoscopy as the dominant screening modality. 2) Contrary to previous belief, in a screening program the vast majority (> 90%) of patients accept and complete colonoscopy as a CRC screening modality. Grant Support: Supported by MRSG-13-144-01 – CPHPS from the American Cancer Society, grant UL1TR000427 through the National Center for Advancing Translational Sciences (NCATS), and grant 1R01CA144835-01 from the National Cancer Institute.

Su1629 Endoscopic Removal of Early Stage Colorectal Cancer in a FIT-Based Screening Population: Progress Still to Be Made Els Wieten*, Esmée J. Grobbee, Ernst J. Kuipers, Paul Didden, Marco J. Bruno, Manon Spaander Gastroenterology and Hepatology, Erasmus University Medical Center, Rotterdam, Netherlands

participate in a FIT-based screening program, between November 2006 and October 2014. Participants were referred for colonoscopy in case of a positive FIT (cut-off 10mg hemoglobin per gram feces). From all detected CRCs, we included endoscopically resectable cancers, defined as a CRC confined to the submucosa, without invasion of the muscularis propria or deeper wall (T1N0M0). Endoscopic resections were considered sufficient, in case of an R0 resection and resection margin > 1mm. Dutch guidelines were followed for treatment and surveillance recommendations, which are comparable to American Society for Gastrointestinal Endoscopy guidelines. Results: Twenty patients were diagnosed with a histologically proven T1N0M0 malignant colorectal polyp. These cancers had a median size of 16 mm (range 5-40 mm) and were located in cecum (nZ1), ascending colon (nZ3), descending colon (nZ1), sigmoid (nZ13) and rectum (nZ2). Seven (35%) polyps were identified as potentially malignant before resection based on its endoscopic appearance (figure 1, table 1). In total, sixteen (80%) primary endoscopic resections were performed and four (20%) primary surgical resections. Reasons for primary surgical resections were (incorrect) suspicion of stadium T2 or more by the endoscopist (nZ3) and familial adenomatous polyposis (nZ1). R0 resection rate with resection margin >1 mm, was 50% in the identified CRCs at colonoscopy and in 33% of the non-identified CRCs. Five (71%) patients from the endoscopically identified CRCs underwent surgery and six (46%) from the non-identified CRC (table 1). Conclusion: In this FIT-based CRC screening population, more than half of the endoscopically resectable CRCs were not identified as such during initial colonoscopy. Additional surgery was still required in a substantial part of these patients often due to incomplete and undetermined resection margins. Our findings implicate that improvement in the recognition of early CRC and use of endoscopic resection techniques are mandatory to ensure completeness for proper pathological evaluation and staging. Table 1. Endoscopy results and additional surgery after diagnosis of

endoscopic resectable colorectal cancers detected by FIT in a CRC screening population. Identified Non-identified CRC CRC n[7 n [ 13 Endoscopy Endoscopic resection (n,%) Yes No Endoscopic resection method (n,%) ESD Liscoagulation Piecemeal Not described En bloc resection (n,%) Yes No Endoscopic R0 resection and margin > 1 mm (n,%) Yes No Surgery Surgical resection performed (n,%) Yes No Reason surgical resection (n) Suspicion of stadium T2 or more Endoscopic resection margin not free or 40 mm and shorten procedure times. Moreover, the retroflexion approach shortened procedure times and may contribute to increased rates of en bloc resection.

Tumor size, mm, meanSD Resection size, mm, meanSD En bloc resection, n (%) Immediate perforation, n (%) Procedure time, meanSD (min) in ESD group

www.giejournal.org

Forward group (n[120)

32.015.0 38.614.1 106/108 (98%) 4/108 (3.7%) (nZ105)

25.812.2 31.412.9 116/120 (96%) 1/120 (0.8%) (nZ106)

p value

93.249.8 76.135.2 (61)

92.752.3 77.236.0 (80)

N.S. (0.9518) N.S. (0.8594)

116.8057.3 (44)

140.665.2 (26)

N.S. (0.1161)

Objective: Universal narrow band imaging (NBI) classification using magnifying colonoscopy was developed by Japan NBI expert team (JNET). We aim to clarify the diagnostic accuracy of JNET classification in a real-time clinical practice. Methods: There were a total of 978 histologically proven consecutive lesions that were removed by endoscopy during April and September 2015. Using prospectively stored database of recorded in real-time procedures JNET classification, Sano’s classification and pit pattern diagnosis were collected. Forty-two lesions were excluded due to lack of endoscopic data, ulcerative colitis, squamous cell carcinoma, inflammatory polyp and others. In total 936 lesions were included. Definition of the JNET classification is listed in the Table1. Definition of the Sano’s classification and pit pattern diagnosis are in accordance with the original report. Expert endoscopist was defined as performed 5000 colonoscopies or more, and non-expert endoscopist was less than 5000 colonoscopies. Results: The median tumor size was 6mm (IQR 410). When JNET classification was used, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy for differentiating neoplastic lesion from non-neoplastic lesion were 95% (761/800), 64% (87/136), 94% (761/810), 69% (87/126), 91% (848/936), respectively. In a subgroup analysis using expert group data, the sensitivity, specificity, PPV, NPV and accuracy were 94% (467/ 497), 68% (63/93), 94% (467/497), 67% (63/94) and 90% (530/591), respectively. In contrast, in the non-expert group, the specificity and NPV were decreased to 56% (24/43) and 56% (24/32). Regarding depth diagnosis, sensitivity, specificity, PPV, NPV and accuracy for submucosal deep invasive cancer (pT1b) among lesions diagnosed as JNET type 2A/2B and 3 were 75% (58/77), 100% (684/684), 100% (58/58), 97% (684/703) and 98% (742/761), respectively. Correlation between JNET classification and histopathological diagnosis was summarize in figure 1. Only JNET type 2B included heterogeneous lesions, from low grade intramucosal neoplasia to pT1b. Among these lesions classified as JNET type 2B, sensitivity, specificity, PPV, NPV and accuracy for pT1b in pit pattern diagnosis were 90% (9/10), 100% (61/61), 100% (9/ 9), 98% (61/62), 99% (70/71), respectively. Using Sano’s classification among same sample, the specificity and NPV were lower at 59.1% (81/137) and 79.4% (81/102) for the differential diagnosis and equal at 99.9% (711/712) and 99.3% (711/716) for depth diagnosis compared to JNET classification. Conclusion: The present study clearly demonstrates usefulness of the JNET classification in both differential diagnosis and depth diagnosis. If the lesion classified JNET type 2B, pit pattern diagnosis should be required for the depth diagnosis.

JNET classification NBI

Type 1

Type 2A

Type 2B

Type 3

Vessel pattern

Invisible *1

Variable caliber Irregular distribution

Loose vessel area Interruption of thick vessels

Surface pattern

Regular dark or white spots surrounding normal mucosa histology

Regular caliber Regular distribution (meshed/spiral pattern) *2 Regular (tubular/ branched/ papillary) Hyperplastic polyp/Sessile serrated polyp

Irregular or obscure

Amorphous areas

Low grade intramucosal neoplasia

High grade intramucosal neoplasia/Shallow submucosal invasive cancer *3

Most likely

p value 0.0008 0.0001 N.S. N.S.

Forward group (n[120)

Su1631 Usefulness of a New Classification “Japan NBI Expert Team (J-NET) Classification” for Endoscopic Diagnosis of Superficial Colorectal Neoplasms, Data From Real-Time Colonoscop Shunsuke Kobayashi*1, Masayoshi Yamada1, Taku Sakamoto1, Takeshi Nakajima1, Takahisa Matsuda1, Shigeki Sekine2, Yutaka Saito1 1 Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan; 2 Molecular Pathology Division, National Cancer Center Research Institute, Tokyo, Japan

Summary of result Retroflexion group (n[108)

Retroflexion group (n[108)

Deep

submucosal invasive cancer

*1. If visible, the caliber in the lesion is similar to surrounding normal mucosa. *2. Microvessels are often distributed in a punctate pattern and well-ordered reticular or spiral vessels may not be observed in depressed lesions. *3. Deep submucosal invasive cancer may be included.

Volume 83, No. 5S : 2016 GASTROINTESTINAL ENDOSCOPY AB371