Prediction of Diminutive/Small Polyp Histology: Didactic vs. Computer-based Training

A Randomised Controlled Trial of the Prediction of Diminutive/Small Polyp Histology: a Comparison Between Didactic Training Versus Self-directed Computer Based Training

Brigham and Women's Hospital, Mount Sinai Hospital, New York, Gunma University Graduate School of Medicine, Maebashi, Iwate Medical University, Istituto Europeo di Oncologia, Federico II University, University of Erlangen-Nürnberg

Bowel cancer is one of the most common cancers and the best method of diagnosing it is through endoscopic examination of the bowel (colonoscopy). Pre-cursors of bowel cancer are called polyps which can be detected and removed at the time of the colonoscopy. This reduces the chance of developing bowel cancer. There are different types of polyp ranging from completely harmless to those that may develop into cancer over time. Advances in technology mean more polyps are being detected and it is possible to predict the type of polyp. Therefore there is a new strategy in endoscopy whereby when a small polyp is detected, a prediction of polyp type is made, the polyp removed and then discarded rather than sending to the laboratory, thereby reducing costs to health services. In the hands of experts, accuracies in predicting polyp type is similar to when the polyp is removed and sent to the lab for analysis. Whilst experts can do this, non-experts cannot reach these standards and there is a need for effective training. The aim of the study is to compare the effectiveness of two training methods: Didactic face-to-face training and computer-based self-learning on the ability of trainees at predicting polyp type.

Colonoscopy is the gold standard for screening for bowel cancers and detection of pre-cursors to colorectal cancer (polyps). Early detection of polyps, allows endoscopic removal and therefore reduction in colorectal cancer. With improvements in technology endoscopists are detecting more lesions within the bowel with the majority small/diminutive <5mm (80%), however the clinical relevance of these lesions is minimal as the risk of advanced histology or cancer is <1%. The current practice involves removing these lesions and sending for histopathological assessment, incurring a significant risk to the patient, cost and is time-consuming, with very little benefit. Novel imaging techniques including Narrow-band imaging (NBI-Olympus, Japan), i-Scan Optical enhancement (OE-Pentax, Japan) and Blue-light laser imaging (BLI- Fujifilm, Japan) can help endoscopists characterise these small lesions between being neoplastic and non-neoplastic (hyperplastic). NBI involves the narrowing of bandwidths of light using a light filter. The light at this end of the spectrum is absorbed by haemoglobin (protein found within blood) therefore making blood vessels more pronounced. During the process whereby a polyp develops and later becomes neoplastic, there is an increase in blood vessels compared with normal tissue or hyperplastic polyps (benign), therefore NBI can be used to detect such lesions. I-Scan OE is an alternative imaging technique which enhances the pattern of the surface of polyps as well as the blood vessels, by manipulating dark-light borders and red, blue and green components of light. Blue laser imaging (BLI) is also new system for image-enhanced endoscopy using laser light. Blue laser imaging utilizes two monochromatic lasers (410 and 450 nm) instead of xenon light. A 410 nm laser visualizes vascular microarchitecture, similar to narrow band imaging, and a 450 nm laser provides white light by excitation. These novel technologies have been demonstrated to be superior over standard white light endoscopy with NBI the most extensively investigated. A systematic analysis of 6 studies >500 polyps, resulted in a pooled sensitivity of 92%, spec 86%, accuracy of 89% at differentiating neoplastic from non-neoplastic lesions when using NBI. Head to head studies of NBI versus white light endoscopy (WLE) have shown NBI is better at differentiating between neoplastic and non-neoplastic lesions. Similar results have been found with i-Scan, with performances better than WLE and like NBI are similar to chromoendoscopy (a technique that involves spraying dye over bowel mucosa which is time-consuming and costly). BLI is a newer imaging platform, with the current evidence suggesting it is effective at differentiating polyps (neoplastic versus non-neoplastic) with accuracies of 95.2%, and when comparing with white light endoscopy the miss rate of adenoma was significantly lower with BLI (1.6% versus 10.0% p=0.001). In order to characterise between neoplastic and non-neoplastic lesions, endoscopic scoring systems have been developed to assist endoscopists. Examples include NICE (NBI International Colorectal Endoscopic). Recently Iacucci et al have developed a simplified classification system (SIMPLE- Simplified Identification Method for Polyp Labelling during the Endoscopy) for optical diagnosis of small and diminutive adenomas, SSA/Ps and hyperplastic polyps using the newly introduced OE-iSCAN system which achieved a high degree of diagnostic accuracy for small/diminutive polyp diagnosis. Furthermore, they have showed that a training module on SIMPLE classification resulted in an overall NPV of 91.3%. This user-friendly classification system can be used by experienced and non-experienced gastroenterologists on multiple endoscopy imaging platforms to differentiate neoplastic from non-neoplastic polyps. A classification system developed by Bisschops R et al recently using BLI called BASIC (BLI Adenoma Serrated International Classification). This takes into account the polyp surface, pit appearance and vessels, which has shown to have a high concordance amongst experts. In the hands of experts using NBI-NICE classification system accuracies of 98.9%, sensitivity 98%, specificity 100%, NPV 97.7% and PPV100% were demonstrated when diagnosis was made with high confidence. Essential to the adopted use of these classifications is training for endoscopists, both experienced and those in training. There is good evidence that there is a short learning curve involved when using NBI. One study using a self-administered computer based training module, community based gastroenterologists (non-expert) were able to reach excellent NPV of >90% but fell short of other requirements (prediction of surveillance intervals). Much like NBI, the learning curve at acquiring the skills in order to differentiate between hyperplastic and adenomatous lesions using i-Scan has been investigated. An early study by Neumann et al demonstrated a rapid learning curve with 4 endoscopists without previous experience with i-scan reached an accuracy of at least 85% after reviewing 67-110 lesions (with individualised feedback) following a 1 hour teaching session on pit pattern analysis. There have attempts at identifying the most effective training tool and method at teaching non-experts how to characterise lesions effectively. Studies have used still images of lesions, however this is limited as it does not reflect real-life practice as it does not allow views from different angles. It is thought videos simulate real-life practice as close as possible. A study using videos has demonstrated trainees were able to achieve accuracies of 90%. More recently Rastogi's group sought to identify which training method was more effective in prediction of diminutive polyp histology amongst trainees: didactic face to face training versus computer-based self-learning. The participants were randomised to either receive didactic training in the form of a classroom training session or self-learning via computer-based material on characterisation of polyps using NBI. Trainees reviewed 40 videos of diminutive polyps with the histology being revealed and explained. Both groups were given a further 40 videos for testing. This study found those taught in the didactic group characterised polyps with higher confidence, but the overall performance was similar in the two groups. The accuracy and sensitivity were slightly better in the self-learning group (93.9% vs 85.7% p 0.01 and 95.0% vs 86.9%; p0.03 respectively) in those polyps assessed with high confidence. This study demonstrates that a computer-based training module can be as effective in didactic training, perhaps a reflection on the amount of online self-learning trainees are exposed to. The investigators aim to recruit participants to receive either didactic face-to-face training or self-directed computer based learning, whereby participants will be taught how to characterise lesions using the NICE, BASIC and SIMPLE classification. The investigators aim to recruit trainees, novice endoscopists and experienced endoscopists to compare the different groups. Pre- and post-training assessments will be completed allowing us to examine the impact of training, which will consist of 40-60 videos (equal proportion of NBI, iScan OE and BLI) in the pre-training assessment and 40-60 videos (different set of videos but also equal proportion of NBI, iScan OE and BLI) in the post-training assessment. A follow up assessment will be completed at 6 months to assess the retention of skills and sustainability of colonic polyp characterisation using the optical diagnosis techniques. An existing library of NBI and OE-iScan videos will be used and further videos will be collected during routine colonoscopies with patients consenting for images to be used for teaching purposes. The investigators hypothesise that following the training module there will be an improvement in performance between the pre-training and post-training assessments. The investigators also hypothesise that there will be no difference between the didactic face-to-face group and the self-training group. This is an important study as better characterisation of small polyps may eventually lead to a 'resect and discard' strategy in the future. This involves characterising small or diminutive polyps (<10mm) as either non-neoplastic or neoplastic, resecting the lesion but not sending for histopathological analysis, which has significant cost savings. In order to do this training is essential. Whilst didactic training is attractive, it is costly and resource heavy. The option of self-directed learning is an attractive one as it can be delivered at times that suit the user, at their pace and can be delivered in greater volumes. This study is unique as it is examining the impact of the training module on different groups of participants (novice, training and experienced endoscopists), using multiple endoscopic platforms(NBI, i-Scan OE and BLI) at a multicentre, international level. It will enable us to assess whether the training module improves performance using different imaging modalities.

Training will be conducted in a classroom for those participants randomised to receive didactic training, with training provided via a PowerPoint (Microsoft Inc., Redmond, Washington, USA) presentation by an expert endoscopist. An endoscopist with extensive experience in optical characterisation using virtual chromoendoscopy reviewed all teaching material.

Participants randomised to the computer-based self-learning group will be given the same PowerPoint presentation as the didactic group and completed the training in a separate room. Participants completed training without feedback interaction.

Included in the training material is an overview of "Resect and Discard", endoscopic platforms (NBI, BLI and i-Scan), NICE classification, SIMPLE classification, BASIC classification and example still images and videos of both classifications in use. Still images will be used to ensure participants have the best opportunity to observe and learn Kudo Pit Patterns and other polyp features without movement artefact before observing videos, which are more challenging to interpret. The training will last approximately 1 hour.

Accuracy of polyp prediction, sensitivity, specificity, positive predictive value and negative predictive value

Inclusion Criteria: Training colonoscopists: gastroenterology trainee in the process of training in colonoscopy and have had some experience of colonoscopy. Exclusion Criteria: Inability to consent to take part in the study Gastroenterology trainees without experience of colonoscopy

Lieberman D, Moravec M, Holub J, Michaels L, Eisen G. Polyp size and advanced histology in patients undergoing colonoscopy screening: implications for CT colonography. Gastroenterology. 2008 Oct;135(4):1100-5. doi: 10.1053/j.gastro.2008.06.083. Epub 2008 Jul 3.

Rex DK, Overhiser AJ, Chen SC, Cummings OW, Ulbright TM. Estimation of impact of American College of Radiology recommendations on CT colonography reporting for resection of high-risk adenoma findings. Am J Gastroenterol. 2009 Jan;104(1):149-53. doi: 10.1038/ajg.2008.35.

Hewett DG, Kaltenbach T, Sano Y, Tanaka S, Saunders BP, Ponchon T, Soetikno R, Rex DK. Validation of a simple classification system for endoscopic diagnosis of small colorectal polyps using narrow-band imaging. Gastroenterology. 2012 Sep;143(3):599-607.e1. doi: 10.1053/j.gastro.2012.05.006. Epub 2012 May 15.

Hassan C, Pickhardt PJ, Rex DK. A resect and discard strategy would improve cost-effectiveness of colorectal cancer screening. Clin Gastroenterol Hepatol. 2010 Oct;8(10):865-9, 869.e1-3. doi: 10.1016/j.cgh.2010.05.018. Epub 2010 Jun 1.

Ignjatovic A, East JE, Suzuki N, Vance M, Guenther T, Saunders BP. Optical diagnosis of small colorectal polyps at routine colonoscopy (Detect InSpect ChAracterise Resect and Discard; DISCARD trial): a prospective cohort study. Lancet Oncol. 2009 Dec;10(12):1171-8. doi: 10.1016/S1470-2045(09)70329-8. Epub 2009 Nov 10.

McGill SK, Evangelou E, Ioannidis JP, Soetikno RM, Kaltenbach T. Narrow band imaging to differentiate neoplastic and non-neoplastic colorectal polyps in real time: a meta-analysis of diagnostic operating characteristics. Gut. 2013 Dec;62(12):1704-13. doi: 10.1136/gutjnl-2012-303965. Epub 2013 Jan 7.

Wu L, Li Y, Li Z, Cao Y, Gao F. Diagnostic accuracy of narrow-band imaging for the differentiation of neoplastic from non-neoplastic colorectal polyps: a meta-analysis. Colorectal Dis. 2013 Jan;15(1):3-11. doi: 10.1111/j.1463-1318.2012.02947.x.

Sakamoto T, Matsuda T, Aoki T, Nakajima T, Saito Y. Time saving with narrow-band imaging for distinguishing between neoplastic and non-neoplastic small colorectal lesions. J Gastroenterol Hepatol. 2012 Feb;27(2):351-5. doi: 10.1111/j.1440-1746.2011.06854.x.

Chiu HM, Chang CY, Chen CC, Lee YC, Wu MS, Lin JT, Shun CT, Wang HP. A prospective comparative study of narrow-band imaging, chromoendoscopy, and conventional colonoscopy in the diagnosis of colorectal neoplasia. Gut. 2007 Mar;56(3):373-9. doi: 10.1136/gut.2006.099614. Epub 2006 Sep 27.

Machida H, Sano Y, Hamamoto Y, Muto M, Kozu T, Tajiri H, Yoshida S. Narrow-band imaging in the diagnosis of colorectal mucosal lesions: a pilot study. Endoscopy. 2004 Dec;36(12):1094-8. doi: 10.1055/s-2004-826040.

Gono K, Obi T, Yamaguchi M, Ohyama N, Machida H, Sano Y, Yoshida S, Hamamoto Y, Endo T. Appearance of enhanced tissue features in narrow-band endoscopic imaging. J Biomed Opt. 2004 May-Jun;9(3):568-77. doi: 10.1117/1.1695563.

Curvers WL, van den Broek FJ, Reitsma JB, Dekker E, Bergman JJ. Systematic review of narrow-band imaging for the detection and differentiation of abnormalities in the esophagus and stomach (with video). Gastrointest Endosc. 2009 Feb;69(2):307-17. doi: 10.1016/j.gie.2008.09.048. No abstract available.

Chang CC, Hsieh CR, Lou HY, Fang CL, Tiong C, Wang JJ, Wei IV, Wu SC, Chen JN, Wang YH. Comparative study of conventional colonoscopy, magnifying chromoendoscopy, and magnifying narrow-band imaging systems in the differential diagnosis of small colonic polyps between trainee and experienced endoscopist. Int J Colorectal Dis. 2009 Dec;24(12):1413-9. doi: 10.1007/s00384-009-0760-9. Epub 2009 Jul 15.

Su MY, Hsu CM, Ho YP, Chen PC, Lin CJ, Chiu CT. Comparative study of conventional colonoscopy, chromoendoscopy, and narrow-band imaging systems in differential diagnosis of neoplastic and nonneoplastic colonic polyps. Am J Gastroenterol. 2006 Dec;101(12):2711-6. doi: 10.1111/j.1572-0241.2006.00932.x.

Hoffman A, Kagel C, Goetz M, Tresch A, Mudter J, Biesterfeld S, Galle PR, Neurath MF, Kiesslich R. Recognition and characterization of small colonic neoplasia with high-definition colonoscopy using i-Scan is as precise as chromoendoscopy. Dig Liver Dis. 2010 Jan;42(1):45-50. doi: 10.1016/j.dld.2009.04.005. Epub 2009 May 26.

Lee CK, Lee SH, Hwangbo Y. Narrow-band imaging versus I-Scan for the real-time histological prediction of diminutive colonic polyps: a prospective comparative study by using the simple unified endoscopic classification. Gastrointest Endosc. 2011 Sep;74(3):603-9. doi: 10.1016/j.gie.2011.04.049. Epub 2011 Jul 18.

Yoshida N, Yagi N, Inada Y, Kugai M, Okayama T, Kamada K, Katada K, Uchiyama K, Ishikawa T, Handa O, Takagi T, Konishi H, Kokura S, Yanagisawa A, Naito Y. Ability of a novel blue laser imaging system for the diagnosis of colorectal polyps. Dig Endosc. 2014 Mar;26(2):250-8. doi: 10.1111/den.12127. Epub 2013 Jun 3.

Shimoda R, Sakata Y, Fujise T, Yamanouchi K, Tsuruoka N, Hara M, Nakayama A, Yamaguchi D, Akutagawa T, Fujimoto K, Iwakiri R. The adenoma miss rate of blue-laser imaging vs. white-light imaging during colonoscopy: a randomized tandem trial. Endoscopy. 2017 Feb;49(2):186-190. doi: 10.1055/s-0042-118450. Epub 2016 Nov 14.

Iacucci M, Trovato C, Daperno M, Akinola O, Greenwald D, Gross SA, Hoffman A, Lee J, Lethebe BC, Lowerison M, Nayor J, Neumann H, Rath T, Sanduleanu S, Sharma P, Kiesslich R, Ghosh S, Saltzman JR; SIMPLE classification investigator team. Development and validation of the SIMPLE endoscopic classification of diminutive and small colorectal polyps. Endoscopy. 2018 Aug;50(8):779-789. doi: 10.1055/s-0044-100791. Epub 2018 Mar 23. Erratum In: Endoscopy. 2018 Aug;50(8):C8.

Bisschops R, Hassan C, Bhandari P, Coron E, Neumann H, Pech O, Correale L, Repici A. BASIC (BLI Adenoma Serrated International Classification) classification for colorectal polyp characterization with blue light imaging. Endoscopy. 2018 Mar;50(3):211-220. doi: 10.1055/s-0043-121570. Epub 2017 Oct 24.

Rogart JN, Jain D, Siddiqui UD, Oren T, Lim J, Jamidar P, Aslanian H. Narrow-band imaging without high magnification to differentiate polyps during real-time colonoscopy: improvement with experience. Gastrointest Endosc. 2008 Dec;68(6):1136-45. doi: 10.1016/j.gie.2008.04.035. Epub 2008 Aug 8.

Ladabaum U, Fioritto A, Mitani A, Desai M, Kim JP, Rex DK, Imperiale T, Gunaratnam N. Real-time optical biopsy of colon polyps with narrow band imaging in community practice does not yet meet key thresholds for clinical decisions. Gastroenterology. 2013 Jan;144(1):81-91. doi: 10.1053/j.gastro.2012.09.054. Epub 2012 Oct 3.

Neumann H, Vieth M, Fry LC, Gunther C, Atreya R, Neurath MF, Monkemuller K. Learning curve of virtual chromoendoscopy for the prediction of hyperplastic and adenomatous colorectal lesions: a prospective 2-center study. Gastrointest Endosc. 2013 Jul;78(1):115-20. doi: 10.1016/j.gie.2013.02.001. Epub 2013 Mar 23.

Higashi R, Uraoka T, Kato J, Kuwaki K, Ishikawa S, Saito Y, Matsuda T, Ikematsu H, Sano Y, Suzuki S, Murakami Y, Yamamoto K. Diagnostic accuracy of narrow-band imaging and pit pattern analysis significantly improved for less-experienced endoscopists after an expanded training program. Gastrointest Endosc. 2010 Jul;72(1):127-35. doi: 10.1016/j.gie.2010.01.054. Epub 2010 May 20.

Ignjatovic A, Thomas-Gibson S, East JE, Haycock A, Bassett P, Bhandari P, Man R, Suzuki N, Saunders BP. Development and validation of a training module on the use of narrow-band imaging in differentiation of small adenomas from hyperplastic colorectal polyps. Gastrointest Endosc. 2011 Jan;73(1):128-33. doi: 10.1016/j.gie.2010.09.021.

Khan T, Cinnor B, Gupta N, Hosford L, Bansal A, Olyaee MS, Wani S, Rastogi A. Didactic training vs. computer-based self-learning in the prediction of diminutive colon polyp histology by trainees: a randomized controlled study. Endoscopy. 2017 Dec;49(12):1243-1250. doi: 10.1055/s-0043-116015. Epub 2017 Aug 14.