DDW HIGHLIGHTS
Colonoscopy Michal F. Kaminski, MD, PhD,1,2 Edyta Zagorowicz1,2 Warsaw, Poland
Several high-quality studies pertinent to colonoscopy were presented at Digestive Disease Week (DDW) 2014. The following review discusses studies that are considered to address some of the most burning questions in colonoscopy research: quality of colonoscopy and its measurement, surveillance intervals for adenomas, the treatment of diminutive adenomas, basic and advanced polypectomy techniques, and microbiome changes in relation to colonoscopy.
COLONOSCOPY QUALITY MEASUREMENT Professional societies have proposed the use of several quality measures for colonoscopy.1,2 However, the continual assessment of quality measures is challenging because it requires complete data capture, the merging of colonoscopy and pathology reports, and manual categorization. Natural language processing is a computer technology that identifies and extracts information from free-text reports in an automated fashion, and is therefore potentially suitable for colonoscopy quality measurement. At this year’s DDW, Raju et al3 reported on a crosssectional study that compared natural language processing with manual reporting of colonoscopy performance at one academic institution in the United States. The natural language processing program extracted the data from three data sources: the electronic medical record, the colonoscopy report, and the pathology report, processing them to produce colonoscopy quality metrics. The data were considered to be accurate when both methods agreed, otherwise the medical record was studied further to establish the correct information. Of the 12/478 colonoscopies studied, 2088 were first-time screening examinations, and Abbreviations: CRC, Colorectal cancers; DDW, Digestive Disease Week; PCR, polymerase chain reaction. DISCLOSURE: All authors disclosed no financial relationships relevant to this article. This report is published simultaneously in the journals Gastrointestinal Endoscopy and Endoscopy. Copyright ª 2014 by the American Society for Gastrointestinal Endoscopy and ª Georg Thieme Verlag KG 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2014.07.005
400 GASTROINTESTINAL ENDOSCOPY Volume 80, No. 3 : 2014
the natural language processing, which required no more than an hour, accurately identified 98% of them compared with 86% identified by manual reporting (R5 min/record). Both methods nearly perfectly identified adenomas (99% and 98% for the natural language processing and manual reporting, respectively) and sessile serrated adenomas (100% and 96%, respectively). This study is in line with a recent report showing very high accuracy of natural language processing in categorizing data necessary for colonoscopy quality measurement.4 One of the major drawbacks of natural language processing in daily practice is that it requires complete reporting of the data that are necessary for quality measurement. An alternative would be to use a structured colonoscopy reporting system, which forces endoscopists to report a minimum set of items, generates a free text reports, and provides automatic quality assessment.5 The latter, however, requires manual entry of histopathology results. In conclusion, this study provides evidence that natural language processing is a quick and accurate tool for colonoscopy quality measurement. Natural language processing combined with structured colonoscopy reporting systems may offer an optimal solution for continuous quality measurement.
INTERVAL COLORECTAL CANCERS Colorectal cancers (CRC) that occur before scheduled surveillance or between month 6 and 36 after a clearing colonoscopy are called interval cancers.6-8 Populationbased studies from the United States, Canada, and Europe have reported that interval cancers account for 3.4%–9.2% of all detected CRCs.6-8 At DDW 2014, Cheung et al9 reported a nationwide analysis of the rates, risk factors, and time trends in the occurrence of interval CRCs in England between the years 2001 and 2012. The authors used data from the National Health Service to identify patients who were diagnosed with a CRC 6–36 months following a negative (no cancer diagnosis) colonoscopy. Of 2/263/905 patients who underwent colonoscopy, 136/237 were diagnosed with CRC, 12/485 of whom (9.2%) were diagnosed with interval CRC. In a multivariate analysis, interval CRCs were directly associated with older age, female sex, and co-morbidities and, unexpectedly, were inversely associated with proximal colon location. www.giejournal.org
Kaminski & Zagorowicz
Yet the most interesting observation arising from this study, is that the annual rate of interval CRC declined, from 15.9% to 5.1%, over the study period. Suboptimal quality of colonoscopy is considered to be the key and most modifiable risk factor for interval CRC.6-8 Indeed, there has been considerable improvement in the quality of colonoscopy over the study period in the United Kingdom.10 Therefore, it is likely that the observed fall in annual rates of interval CRC follows previously reported improvement in the quality of colonoscopy. Although we still need to close the gap between improvement in quality of colonoscopy and its effect on interval CRC rates, this study shows promise for the future.
Colonoscopy
(10.7%) than in other groups, and this supports short-term surveillance as recommended by the guidelines. In conclusion, the natural history of diminutive adenomas shows that 5-year risk of advanced neoplasia is low in patients with 1–2 adenomas but five-fold higher in patients with R3 adenomas.
BASIC POLYPECTOMY TECHNIQUE
In patients with one or two nonadvanced adenomas that are completely removed during a high-quality baseline colonoscopy, the European Society of Gastrointestinal Endoscopy recommends surveillance colonoscopy after 10 years.11 In contrast, detection of three or more adenomas, regardless of their size or dysplasia grade, is an indication to repeat the colonoscopy after 3 years. The quality of evidence supporting these recommendations was judged to be moderate. New evidence presented by Otake et al12 at DDW, came from a study that aimed to determine the incidence of advanced neoplasia in screenees with diminutive polyps that were judged to be adenomas on chromoendoscopy with magnification but were not removed at the index colonoscopy. The authors retrospectively studied the 5-year cumulative incidence of advanced neoplasia in 2070 patients who had no adenomas and 705 patients who were optically diagnosed with adenomas of %5 mm that were not removed at screening colonoscopy. The cumulative incidence of advanced neoplasia was 1.7% (35/2070) in the group without adenomas and 2.8% (20/705) in the group with diminutive adenomas (P Z 0.059). Notably, the cumulative incidence of advanced neoplasia was significantly higher in patients with R3 adenomas at the index colonoscopy (10.8%;7/65) than in patients with 1–2 adenomas (2.0%;13/640, P value not given). Two observations from this study are important for the timing of colonoscopy surveillance following adenoma removal. First, the 5-year risk of advanced neoplasia in patients with 1–2 untreated diminutive adenomas was comparably low, similar to that of patients without adenomas. Given that the adenomas should have been removed at index colonoscopy, this evidence strongly supports surveillance times being longer than 5 years in patients with 1–2 diminutive adenomas. In patients with at least three untreated diminutive adenomas, the risk of advanced adenoma was five-fold higher
Incomplete polypectomy may significantly contribute to the development of interval CRCs.13 Although available data indicate that most of these CRCs develop from incomplete resection of large adenomas,13 it is believed that some arise from remnants of smaller lesions. In order to achieve complete removal, it is recommended to resect small polyps (6–9 mm in size) using a polypectomy snare and diminutive ones (1–5 mm in size) using a cold biopsy or a cold snaring technique.14 Most experts suggest using cold biopsies only for polyps 1–3 mm in size and the cold snaring technique for polyps 4–6 mm or even 4–9 mm in size.15 At this year’s DDW, two interesting studies regarding basic polypectomy technique were presented.16,17 BrittoArias et al16 reported on a large cross-sectional analysis of polyp resection technique and associated completeness of adenoma removal during colonoscopy in the Austrian National Colorectal Cancer Screening Program. Database records of 115/356 colonoscopies, including 40/020 polypectomies, performed between November 2007 and July 2013 were studied. The authors found that, in defiance of recommendations, 46.4% of 15/128 polyps R5 mm in size were removed using biopsy forceps (52.7% when considering polyps 5–10 mm). The rate decreased over time by 17.8% in the hospitals, but increased by 7.5% in private practices, suggesting poor penetration of the existing guidelines in the community. Importantly, the inappropriate technique had significant consequences for the completeness of polyp removal. Of 18/387 adenomas with complete histopathological and resection details, incomplete resections occurred in 17.0% (95% confidence interval [CI] 15.41%–18.64%) of forceps and 6.2% (95%CI 5.37%–7.06%) of snare polypectomies. The relative risk (RR) of incomplete resection for forceps vs. snare polypectomy was 2.98 (95%CI 2.56–3.47) for adenomas R5 mm, whereas for adenomas !5 mm in size it was significant only for polypectomies performed in private practices (RR 1.68, 95%CI 1.42–2.00). In the DDW plenary session, Kim et al17 presented results of a single-center, randomized controlled trial comparing complete resection rates of cold snare polypectomy and hot snare polypectomy for 213 sessile and flat colorectal polyps 5–9 mm in size. The completeness of polyp resection was assessed by histological assessment of four-quadrant forceps biopsies taken from the edges of the polypectomy site. The complete resection rate was
www.giejournal.org
Volume 80, No. 3 : 2014 GASTROINTESTINAL ENDOSCOPY 401
NATURAL HISTORY OF DIMINUTIVE ADENOMAS
Colonoscopy
significantly higher in the hot snare polypectomy group than in the cold snare polypectomy group (92.2% vs. 79.1%;P Z 0.012). In a multivariate analysis, cold snare polypectomy was significantly associated with incomplete resection (odds ratio 2.94, 95%CI 1.21–7.13), whereas polyp size, morphology, and location were not. Both studies need to be interpreted in the context of results from the randomized controlled trial of cold snare polypectomy vs. cold forceps biopsy for diminutive colorectal polyps that was presented at last year’s DDW.18 In the latter study, cold forceps biopsy and polyp size 4–5 mm were significantly associated with incomplete removal. Data from these studies reinforce the concept that use of cold biopsy forceps should be limited to polyps 1–3 mm in size, whereas larger polyps need to be resected using a snare. The complete resection rate with cold snare polypectomy was 93.2% for polyps 1–5 mm in the study by Lee et al18 and 79.1% for polyps 5–9 mm in the study by Kim et al,17 respectively. It remains to be clarified whether the observed difference in complete resection rate using cold snaring reflects its suboptimal performance with larger polyps or difference in performance between endoscopists. It is worth repeating that the optimal cold snaring technique attempts to ensnare a 1–2 mm rim of normal mucosa around the polyp and avoids tenting before the resection. In conclusion, these studies provide important guidance on the optimal resection technique of small and diminutive polyps.
Kaminski & Zagorowicz
converted to open laparotomy). The mean operating time and time to discharge were shorter in the polypectomy group than in the hemicolectomy group (95 vs. 179 minutes, P Z 0.001 and 2.63 vs. 4.94 days, P ! 0.001, respectively). Postoperative complications, readmissions, and reoperation rates were similar in both groups. In order to fully understand the potential advantage of laparoscopic-assisted polypectomy, the rate of early local polyp recurrence should have been measured and reported. For conventional endoscopic piecemeal resection of large polyps, it ranges from 14% to 55%.15 Although most recurrences are successfully managed endoscopically, high rates could have lowered the potential benefit of the method. In conclusion, this study provides evidence that laparoscopic-assisted polypectomy for a resection of large, benign, right-sided polyps considered not amenable to standard endoscopic resection, may result in shorter operating times and shorter hospital stays compared with laparoscopic hemicolectomy.
BOWEL PREPARATION AND MICROBIOME SHIFT
Some benign-looking colonic polyps are not amenable to endoscopic resection (extending across several haustral folds, overlying the appendiceal orifice, or otherwise inaccessible), and therefore are treated by segmental colectomy.15 Endoscopic removal of the polyp with laparoscopic assistance may help to avoid morbidity associated with bowel resection. During the procedure, the colon is mobilized if required, the proximal bowel is crossclamped, and the polyp is colonoscopically resected. The serosal surface of the colon is monitored laparoscopically, and repaired if needed. If an immediate pathologic evaluation of a frozen section indicates malignancy, or technical difficulties occur, then a segmental resection may be performed.19 Buscaglia et al20 reported the first randomized controlled trial comparing laparoscopic-assisted polypectomy with laparoscopic hemicolectomy for right-sided polyps not amenable to endoscopic resection. Between 2009 and 2013, 28 patients were enrolled with benignlooking lesions, which, in the opinion of an expert colonoscopist, were endoscopically unresectable. All of the polyps were completely removed (one patient needed conversion to laparoscopic hemicolectomy and four patients were
Diversity of and changes in the human endogenous gastrointestinal microflora have been explored.21 Cultureindependent phylogenetic analyses performed in patients with inflammatory bowel disease indicate significant differences between the microbiomes of Crohn’s disease and ulcerative colitis patients and healthy controls. In addition, shifts in the microbial composition in parallel with changing inflammatory bowel disease activity have been observed. As reported by patients, and confirmed in the literature, colonoscopy may induce exacerbation of inflammatory bowel disease.22 Changes in the gut microbiome in relation to bowel preparation for colonoscopy are one of the potential culprits. At DDW 2014, Kim et al23 reported on a study of the effects of polyethylene glycol, which is used for colonoscopy bowel preparation, on the composition of the gut microbiome. First, they collected pre-colonoscopy stool samples from seven patients without inflammatory bowel disease who were scheduled for screening examination. Next, stool was collected during colonoscopy following bowel preparation with polyethylene glycol, and then at weekly intervals for 5 weeks after the examination. Following the extraction of stool DNA and quantitative polymerase chain reaction (PCR) of total bacterial 16S rRNA, the authors sequenced the V4 hypervariable regions of the 16S rRNA and performed bioinformatic analyses of each sample. They observed that total 16S bacterial counts, as measured by quantitative PCR, were significantly decreased after bowel preparation and returned to baseline within 2 weeks after the colonoscopy. Analysis of the total 2/510/597 sequences obtained, demonstrated significant shifts in the
402 GASTROINTESTINAL ENDOSCOPY Volume 80, No. 3 : 2014
www.giejournal.org
LAPAROSCOPIC–ASSISTED POLYPECTOMY
Kaminski & Zagorowicz
composition of the individual microflora between the samples taken before bowel preparation and the samples taken during colonoscopy. The baseline microbiome composition was regained by all seven patients within the first week after colonoscopy and no significant changes were seen in the following 5 weeks. In conclusion, this was the first study to show large shifts in the human microbiome induced by the bowel preparation used for colonoscopy, with complete and rapid recovery. Further studies of microbiome changes induced by colonoscopy in patients with inflammatory bowel disease would be very interesting in the context of colonoscopy-induced colitis.
REFERENCES 1. Rex DK, Petrini JL, Baron TH, et al. Quality indicators for colonoscopy. Am J Gastroenterol 2006;101:873-85. 2. Valori R, Rey JF, Atkin WS, et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis. First editiond quality assurance in endoscopy in colorectal cancer screening and diagnosis. Endoscopy 2012;44(suppl 3):SE88-105. 3. Raju GS, Ross WA, Lum P, et al. Natural language processing as an alternative to manual reporting of colonoscopy quality metrics [abstract]. Gastrointest Endosc 2014;79:AB116. 4. Imler TD, Morea J, Kahi C, et al. Natural language processing accurately categorizes findings from colonoscopy and pathology reports. Clin Gastroenterol Hepatol 2013;11:689-94. 5. van Doorn SC, van Vliet J, Fockens P, et al. A novel colonoscopy reporting system enabling quality assurance. Endoscopy 2014;46:181-7. 6. Kaminski MF, Regula J, Kraszewska E, et al. Quality indicators for colonoscopy and the risk of interval cancer. N Engl J Med 2010;362: 1795-803. 7. Bressler B, Paszat LF, Chen Z, et al. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology 2007;132:96-102. 8. Corley DA, Jensen CD, Marks AR, et al. Adenoma detection rate and risk of colorectal cancer and death. N Engl J Med 2014;370:1298-306. 9. Cheung K, Evison F, Patel P, et al. How commonly is colorectal cancer later diagnosed following a colonoscopy that does not report colorectal cancer (an analysis of 11 years of national data in England) [abstract]? Gastrointest Endosc 2014;79:AB117-8. 10. Gavin DR, Valori RM, Anderson JT, et al. The national colonoscopy audit: a nationwide assessment of the quality and safety of colonoscopy in the UK. Gut 2013;62:242-9. 11. Hassan C, Quintero E, Dumonceau JM, et al. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy 2013;45:842-64.
www.giejournal.org
Colonoscopy 12. Otake Y, Kakugawa Y, Matsumoto M, et al. Incidence of advanced neoplasia in individuals with untreated diminutive adenomas: a longitudinal study [abstract]. Gastrointest Endosc 2014;79:AB126. 13. Robertson DJ, Lieberman DA, Winawer SJ, et al. Colorectal cancers soon after colonoscopy: a pooled multicohort analysis. Gut 2014;63: 949-56. 14. Steele RJ, Pox C, Kuipers EJ, et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis. First editiond management of lesions detected in colorectal cancer screening. Endoscopy 2012;44(suppl 3):SE140-50. 15. Tolliver KA, Rex DK. Colonoscopic polypectomy. Gastroenterol Clin North Am 2008;37:229-51; ix. 16. Britto-Arias M, Waldmann E, Bannert C, et al. Forceps versus snare polypectomiesdlow adherence to guideline results in incomplete resection [abstract]. Gastrointest Endosc 2014;79:AB129. 17. Kim HS, Hwang HW, Park HJ, et al. Cold snare versus hot snare polypectomy for the complete resection of 5-9 mm sized colorectal polyps: a randomized controlled trial [abstract]. Gastrointest Endosc 2014;79:AB401-2. 18. Lee CK, Shim JJ, Jang JY. Cold snare polypectomy vs. cold forceps polypectomy using double-biopsy technique for removal of diminutive colorectal polyps: a prospective randomized study. Am J Gastroenterol 2013;108:1593-600. 19. Franklin ME Jr, Diaz EJ, Abrego D, et al. Laparoscopic-assisted colonoscopic polypectomy: the Texas Endosurgery Institute experience. Dis Colon Rectum 2000;43:1246-9. 20. Buscaglia J, Lascarides CE, Nagula S, et al. Laparoscopic-assisted colonoscopy with polypectomy (LACP) versus laparoscopic hemicolectomy for endoscopically unresectable right-sided colon polyps: results of a randomized controlled trial [abstract]. Gastrointest Endosc 2014;79: AB160. 21. Eckburg PB, Bik EM, Bernstein CN, et al. Diversity of the human intestinal microbial flora. Science 2005;308:1635-8. 22. Menees S, Higgins P, Korsnes S, et al. Does colonoscopy cause increased ulcerative colitis symptoms? Inflamm Bowel Dis 2007;13: 12-8. 23. Kim SB, Chuang P, Walsh E, et al. Significant shifts of the microbiome after bowel preparation for colonoscopy are transient as demonstrated by longitudinal stool sampling and high-throughput sequencing. Gastroenterology 2014;146(5 suppl 1):S835.
Received July 1, 2014. Accepted July 1, 2014. Current affiliations: Department of Gastroenterology, Hepatology, and Oncology, Medical Center for Postgraduate Education (1), Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Warsaw, Poland (2). Reprint requests: Michal F. Kaminski, MD, PhD, Department of Gastroenterology, Hepatology, and Oncology, Center of Oncology, Roentgen Street 5, Warsaw 02-781 Poland.
Volume 80, No. 3 : 2014 GASTROINTESTINAL ENDOSCOPY 403
Colonoscopy Michal F. Kaminski, MD, PhD,1,2 Edyta Zagorowicz1,2 Warsaw, Poland
Several high-quality studies pertinent to colonoscopy were presented at Digestive Disease Week (DDW) 2014. The following review discusses studies that are considered to address some of the most burning questions in colonoscopy research: quality of colonoscopy and its measurement, surveillance intervals for adenomas, the treatment of diminutive adenomas, basic and advanced polypectomy techniques, and microbiome changes in relation to colonoscopy.
COLONOSCOPY QUALITY MEASUREMENT Professional societies have proposed the use of several quality measures for colonoscopy.1,2 However, the continual assessment of quality measures is challenging because it requires complete data capture, the merging of colonoscopy and pathology reports, and manual categorization. Natural language processing is a computer technology that identifies and extracts information from free-text reports in an automated fashion, and is therefore potentially suitable for colonoscopy quality measurement. At this year’s DDW, Raju et al3 reported on a crosssectional study that compared natural language processing with manual reporting of colonoscopy performance at one academic institution in the United States. The natural language processing program extracted the data from three data sources: the electronic medical record, the colonoscopy report, and the pathology report, processing them to produce colonoscopy quality metrics. The data were considered to be accurate when both methods agreed, otherwise the medical record was studied further to establish the correct information. Of the 12/478 colonoscopies studied, 2088 were first-time screening examinations, and Abbreviations: CRC, Colorectal cancers; DDW, Digestive Disease Week; PCR, polymerase chain reaction. DISCLOSURE: All authors disclosed no financial relationships relevant to this article. This report is published simultaneously in the journals Gastrointestinal Endoscopy and Endoscopy. Copyright ª 2014 by the American Society for Gastrointestinal Endoscopy and ª Georg Thieme Verlag KG 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2014.07.005
400 GASTROINTESTINAL ENDOSCOPY Volume 80, No. 3 : 2014
the natural language processing, which required no more than an hour, accurately identified 98% of them compared with 86% identified by manual reporting (R5 min/record). Both methods nearly perfectly identified adenomas (99% and 98% for the natural language processing and manual reporting, respectively) and sessile serrated adenomas (100% and 96%, respectively). This study is in line with a recent report showing very high accuracy of natural language processing in categorizing data necessary for colonoscopy quality measurement.4 One of the major drawbacks of natural language processing in daily practice is that it requires complete reporting of the data that are necessary for quality measurement. An alternative would be to use a structured colonoscopy reporting system, which forces endoscopists to report a minimum set of items, generates a free text reports, and provides automatic quality assessment.5 The latter, however, requires manual entry of histopathology results. In conclusion, this study provides evidence that natural language processing is a quick and accurate tool for colonoscopy quality measurement. Natural language processing combined with structured colonoscopy reporting systems may offer an optimal solution for continuous quality measurement.
INTERVAL COLORECTAL CANCERS Colorectal cancers (CRC) that occur before scheduled surveillance or between month 6 and 36 after a clearing colonoscopy are called interval cancers.6-8 Populationbased studies from the United States, Canada, and Europe have reported that interval cancers account for 3.4%–9.2% of all detected CRCs.6-8 At DDW 2014, Cheung et al9 reported a nationwide analysis of the rates, risk factors, and time trends in the occurrence of interval CRCs in England between the years 2001 and 2012. The authors used data from the National Health Service to identify patients who were diagnosed with a CRC 6–36 months following a negative (no cancer diagnosis) colonoscopy. Of 2/263/905 patients who underwent colonoscopy, 136/237 were diagnosed with CRC, 12/485 of whom (9.2%) were diagnosed with interval CRC. In a multivariate analysis, interval CRCs were directly associated with older age, female sex, and co-morbidities and, unexpectedly, were inversely associated with proximal colon location. www.giejournal.org
Kaminski & Zagorowicz
Yet the most interesting observation arising from this study, is that the annual rate of interval CRC declined, from 15.9% to 5.1%, over the study period. Suboptimal quality of colonoscopy is considered to be the key and most modifiable risk factor for interval CRC.6-8 Indeed, there has been considerable improvement in the quality of colonoscopy over the study period in the United Kingdom.10 Therefore, it is likely that the observed fall in annual rates of interval CRC follows previously reported improvement in the quality of colonoscopy. Although we still need to close the gap between improvement in quality of colonoscopy and its effect on interval CRC rates, this study shows promise for the future.
Colonoscopy
(10.7%) than in other groups, and this supports short-term surveillance as recommended by the guidelines. In conclusion, the natural history of diminutive adenomas shows that 5-year risk of advanced neoplasia is low in patients with 1–2 adenomas but five-fold higher in patients with R3 adenomas.
BASIC POLYPECTOMY TECHNIQUE
In patients with one or two nonadvanced adenomas that are completely removed during a high-quality baseline colonoscopy, the European Society of Gastrointestinal Endoscopy recommends surveillance colonoscopy after 10 years.11 In contrast, detection of three or more adenomas, regardless of their size or dysplasia grade, is an indication to repeat the colonoscopy after 3 years. The quality of evidence supporting these recommendations was judged to be moderate. New evidence presented by Otake et al12 at DDW, came from a study that aimed to determine the incidence of advanced neoplasia in screenees with diminutive polyps that were judged to be adenomas on chromoendoscopy with magnification but were not removed at the index colonoscopy. The authors retrospectively studied the 5-year cumulative incidence of advanced neoplasia in 2070 patients who had no adenomas and 705 patients who were optically diagnosed with adenomas of %5 mm that were not removed at screening colonoscopy. The cumulative incidence of advanced neoplasia was 1.7% (35/2070) in the group without adenomas and 2.8% (20/705) in the group with diminutive adenomas (P Z 0.059). Notably, the cumulative incidence of advanced neoplasia was significantly higher in patients with R3 adenomas at the index colonoscopy (10.8%;7/65) than in patients with 1–2 adenomas (2.0%;13/640, P value not given). Two observations from this study are important for the timing of colonoscopy surveillance following adenoma removal. First, the 5-year risk of advanced neoplasia in patients with 1–2 untreated diminutive adenomas was comparably low, similar to that of patients without adenomas. Given that the adenomas should have been removed at index colonoscopy, this evidence strongly supports surveillance times being longer than 5 years in patients with 1–2 diminutive adenomas. In patients with at least three untreated diminutive adenomas, the risk of advanced adenoma was five-fold higher
Incomplete polypectomy may significantly contribute to the development of interval CRCs.13 Although available data indicate that most of these CRCs develop from incomplete resection of large adenomas,13 it is believed that some arise from remnants of smaller lesions. In order to achieve complete removal, it is recommended to resect small polyps (6–9 mm in size) using a polypectomy snare and diminutive ones (1–5 mm in size) using a cold biopsy or a cold snaring technique.14 Most experts suggest using cold biopsies only for polyps 1–3 mm in size and the cold snaring technique for polyps 4–6 mm or even 4–9 mm in size.15 At this year’s DDW, two interesting studies regarding basic polypectomy technique were presented.16,17 BrittoArias et al16 reported on a large cross-sectional analysis of polyp resection technique and associated completeness of adenoma removal during colonoscopy in the Austrian National Colorectal Cancer Screening Program. Database records of 115/356 colonoscopies, including 40/020 polypectomies, performed between November 2007 and July 2013 were studied. The authors found that, in defiance of recommendations, 46.4% of 15/128 polyps R5 mm in size were removed using biopsy forceps (52.7% when considering polyps 5–10 mm). The rate decreased over time by 17.8% in the hospitals, but increased by 7.5% in private practices, suggesting poor penetration of the existing guidelines in the community. Importantly, the inappropriate technique had significant consequences for the completeness of polyp removal. Of 18/387 adenomas with complete histopathological and resection details, incomplete resections occurred in 17.0% (95% confidence interval [CI] 15.41%–18.64%) of forceps and 6.2% (95%CI 5.37%–7.06%) of snare polypectomies. The relative risk (RR) of incomplete resection for forceps vs. snare polypectomy was 2.98 (95%CI 2.56–3.47) for adenomas R5 mm, whereas for adenomas !5 mm in size it was significant only for polypectomies performed in private practices (RR 1.68, 95%CI 1.42–2.00). In the DDW plenary session, Kim et al17 presented results of a single-center, randomized controlled trial comparing complete resection rates of cold snare polypectomy and hot snare polypectomy for 213 sessile and flat colorectal polyps 5–9 mm in size. The completeness of polyp resection was assessed by histological assessment of four-quadrant forceps biopsies taken from the edges of the polypectomy site. The complete resection rate was
www.giejournal.org
Volume 80, No. 3 : 2014 GASTROINTESTINAL ENDOSCOPY 401
NATURAL HISTORY OF DIMINUTIVE ADENOMAS
Colonoscopy
significantly higher in the hot snare polypectomy group than in the cold snare polypectomy group (92.2% vs. 79.1%;P Z 0.012). In a multivariate analysis, cold snare polypectomy was significantly associated with incomplete resection (odds ratio 2.94, 95%CI 1.21–7.13), whereas polyp size, morphology, and location were not. Both studies need to be interpreted in the context of results from the randomized controlled trial of cold snare polypectomy vs. cold forceps biopsy for diminutive colorectal polyps that was presented at last year’s DDW.18 In the latter study, cold forceps biopsy and polyp size 4–5 mm were significantly associated with incomplete removal. Data from these studies reinforce the concept that use of cold biopsy forceps should be limited to polyps 1–3 mm in size, whereas larger polyps need to be resected using a snare. The complete resection rate with cold snare polypectomy was 93.2% for polyps 1–5 mm in the study by Lee et al18 and 79.1% for polyps 5–9 mm in the study by Kim et al,17 respectively. It remains to be clarified whether the observed difference in complete resection rate using cold snaring reflects its suboptimal performance with larger polyps or difference in performance between endoscopists. It is worth repeating that the optimal cold snaring technique attempts to ensnare a 1–2 mm rim of normal mucosa around the polyp and avoids tenting before the resection. In conclusion, these studies provide important guidance on the optimal resection technique of small and diminutive polyps.
Kaminski & Zagorowicz
converted to open laparotomy). The mean operating time and time to discharge were shorter in the polypectomy group than in the hemicolectomy group (95 vs. 179 minutes, P Z 0.001 and 2.63 vs. 4.94 days, P ! 0.001, respectively). Postoperative complications, readmissions, and reoperation rates were similar in both groups. In order to fully understand the potential advantage of laparoscopic-assisted polypectomy, the rate of early local polyp recurrence should have been measured and reported. For conventional endoscopic piecemeal resection of large polyps, it ranges from 14% to 55%.15 Although most recurrences are successfully managed endoscopically, high rates could have lowered the potential benefit of the method. In conclusion, this study provides evidence that laparoscopic-assisted polypectomy for a resection of large, benign, right-sided polyps considered not amenable to standard endoscopic resection, may result in shorter operating times and shorter hospital stays compared with laparoscopic hemicolectomy.
BOWEL PREPARATION AND MICROBIOME SHIFT
Some benign-looking colonic polyps are not amenable to endoscopic resection (extending across several haustral folds, overlying the appendiceal orifice, or otherwise inaccessible), and therefore are treated by segmental colectomy.15 Endoscopic removal of the polyp with laparoscopic assistance may help to avoid morbidity associated with bowel resection. During the procedure, the colon is mobilized if required, the proximal bowel is crossclamped, and the polyp is colonoscopically resected. The serosal surface of the colon is monitored laparoscopically, and repaired if needed. If an immediate pathologic evaluation of a frozen section indicates malignancy, or technical difficulties occur, then a segmental resection may be performed.19 Buscaglia et al20 reported the first randomized controlled trial comparing laparoscopic-assisted polypectomy with laparoscopic hemicolectomy for right-sided polyps not amenable to endoscopic resection. Between 2009 and 2013, 28 patients were enrolled with benignlooking lesions, which, in the opinion of an expert colonoscopist, were endoscopically unresectable. All of the polyps were completely removed (one patient needed conversion to laparoscopic hemicolectomy and four patients were
Diversity of and changes in the human endogenous gastrointestinal microflora have been explored.21 Cultureindependent phylogenetic analyses performed in patients with inflammatory bowel disease indicate significant differences between the microbiomes of Crohn’s disease and ulcerative colitis patients and healthy controls. In addition, shifts in the microbial composition in parallel with changing inflammatory bowel disease activity have been observed. As reported by patients, and confirmed in the literature, colonoscopy may induce exacerbation of inflammatory bowel disease.22 Changes in the gut microbiome in relation to bowel preparation for colonoscopy are one of the potential culprits. At DDW 2014, Kim et al23 reported on a study of the effects of polyethylene glycol, which is used for colonoscopy bowel preparation, on the composition of the gut microbiome. First, they collected pre-colonoscopy stool samples from seven patients without inflammatory bowel disease who were scheduled for screening examination. Next, stool was collected during colonoscopy following bowel preparation with polyethylene glycol, and then at weekly intervals for 5 weeks after the examination. Following the extraction of stool DNA and quantitative polymerase chain reaction (PCR) of total bacterial 16S rRNA, the authors sequenced the V4 hypervariable regions of the 16S rRNA and performed bioinformatic analyses of each sample. They observed that total 16S bacterial counts, as measured by quantitative PCR, were significantly decreased after bowel preparation and returned to baseline within 2 weeks after the colonoscopy. Analysis of the total 2/510/597 sequences obtained, demonstrated significant shifts in the
402 GASTROINTESTINAL ENDOSCOPY Volume 80, No. 3 : 2014
www.giejournal.org
LAPAROSCOPIC–ASSISTED POLYPECTOMY
Kaminski & Zagorowicz
composition of the individual microflora between the samples taken before bowel preparation and the samples taken during colonoscopy. The baseline microbiome composition was regained by all seven patients within the first week after colonoscopy and no significant changes were seen in the following 5 weeks. In conclusion, this was the first study to show large shifts in the human microbiome induced by the bowel preparation used for colonoscopy, with complete and rapid recovery. Further studies of microbiome changes induced by colonoscopy in patients with inflammatory bowel disease would be very interesting in the context of colonoscopy-induced colitis.
REFERENCES 1. Rex DK, Petrini JL, Baron TH, et al. Quality indicators for colonoscopy. Am J Gastroenterol 2006;101:873-85. 2. Valori R, Rey JF, Atkin WS, et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis. First editiond quality assurance in endoscopy in colorectal cancer screening and diagnosis. Endoscopy 2012;44(suppl 3):SE88-105. 3. Raju GS, Ross WA, Lum P, et al. Natural language processing as an alternative to manual reporting of colonoscopy quality metrics [abstract]. Gastrointest Endosc 2014;79:AB116. 4. Imler TD, Morea J, Kahi C, et al. Natural language processing accurately categorizes findings from colonoscopy and pathology reports. Clin Gastroenterol Hepatol 2013;11:689-94. 5. van Doorn SC, van Vliet J, Fockens P, et al. A novel colonoscopy reporting system enabling quality assurance. Endoscopy 2014;46:181-7. 6. Kaminski MF, Regula J, Kraszewska E, et al. Quality indicators for colonoscopy and the risk of interval cancer. N Engl J Med 2010;362: 1795-803. 7. Bressler B, Paszat LF, Chen Z, et al. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology 2007;132:96-102. 8. Corley DA, Jensen CD, Marks AR, et al. Adenoma detection rate and risk of colorectal cancer and death. N Engl J Med 2014;370:1298-306. 9. Cheung K, Evison F, Patel P, et al. How commonly is colorectal cancer later diagnosed following a colonoscopy that does not report colorectal cancer (an analysis of 11 years of national data in England) [abstract]? Gastrointest Endosc 2014;79:AB117-8. 10. Gavin DR, Valori RM, Anderson JT, et al. The national colonoscopy audit: a nationwide assessment of the quality and safety of colonoscopy in the UK. Gut 2013;62:242-9. 11. Hassan C, Quintero E, Dumonceau JM, et al. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy 2013;45:842-64.
www.giejournal.org
Colonoscopy 12. Otake Y, Kakugawa Y, Matsumoto M, et al. Incidence of advanced neoplasia in individuals with untreated diminutive adenomas: a longitudinal study [abstract]. Gastrointest Endosc 2014;79:AB126. 13. Robertson DJ, Lieberman DA, Winawer SJ, et al. Colorectal cancers soon after colonoscopy: a pooled multicohort analysis. Gut 2014;63: 949-56. 14. Steele RJ, Pox C, Kuipers EJ, et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis. First editiond management of lesions detected in colorectal cancer screening. Endoscopy 2012;44(suppl 3):SE140-50. 15. Tolliver KA, Rex DK. Colonoscopic polypectomy. Gastroenterol Clin North Am 2008;37:229-51; ix. 16. Britto-Arias M, Waldmann E, Bannert C, et al. Forceps versus snare polypectomiesdlow adherence to guideline results in incomplete resection [abstract]. Gastrointest Endosc 2014;79:AB129. 17. Kim HS, Hwang HW, Park HJ, et al. Cold snare versus hot snare polypectomy for the complete resection of 5-9 mm sized colorectal polyps: a randomized controlled trial [abstract]. Gastrointest Endosc 2014;79:AB401-2. 18. Lee CK, Shim JJ, Jang JY. Cold snare polypectomy vs. cold forceps polypectomy using double-biopsy technique for removal of diminutive colorectal polyps: a prospective randomized study. Am J Gastroenterol 2013;108:1593-600. 19. Franklin ME Jr, Diaz EJ, Abrego D, et al. Laparoscopic-assisted colonoscopic polypectomy: the Texas Endosurgery Institute experience. Dis Colon Rectum 2000;43:1246-9. 20. Buscaglia J, Lascarides CE, Nagula S, et al. Laparoscopic-assisted colonoscopy with polypectomy (LACP) versus laparoscopic hemicolectomy for endoscopically unresectable right-sided colon polyps: results of a randomized controlled trial [abstract]. Gastrointest Endosc 2014;79: AB160. 21. Eckburg PB, Bik EM, Bernstein CN, et al. Diversity of the human intestinal microbial flora. Science 2005;308:1635-8. 22. Menees S, Higgins P, Korsnes S, et al. Does colonoscopy cause increased ulcerative colitis symptoms? Inflamm Bowel Dis 2007;13: 12-8. 23. Kim SB, Chuang P, Walsh E, et al. Significant shifts of the microbiome after bowel preparation for colonoscopy are transient as demonstrated by longitudinal stool sampling and high-throughput sequencing. Gastroenterology 2014;146(5 suppl 1):S835.
Received July 1, 2014. Accepted July 1, 2014. Current affiliations: Department of Gastroenterology, Hepatology, and Oncology, Medical Center for Postgraduate Education (1), Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Warsaw, Poland (2). Reprint requests: Michal F. Kaminski, MD, PhD, Department of Gastroenterology, Hepatology, and Oncology, Center of Oncology, Roentgen Street 5, Warsaw 02-781 Poland.
Volume 80, No. 3 : 2014 GASTROINTESTINAL ENDOSCOPY 403
