- Email: [email protected]
Optical diagnosis of small colorectal polyps during colonoscopy: When to resect and discard?
Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
Contents lists available at ScienceDirect
Best Practice & Research Clinical Gastroenterology
10
Optical diagnosis of small colorectal polyps during colonoscopy: When to resect and discard? Ana Wilson, BA, MD, MRCP Wolfson Unit for Endoscopy, St Mark's Hospital, Harrow HA1 3UJ, United Kingdom
a b s t r a c t Keywords: Colorectal polyps Optical diagnosis Electronic chromoendoscopy Narrow band imaging (NBI) Fujinon intelligent color enhancement (FICE) i-SCAN
E-mail address: [email protected]. http://dx.doi.org/10.1016/j.bpg.2015.06.007 1521-6918/© 2015 Elsevier Ltd. All rights reserved.
Colonoscopy with polypectomy has been shown to be effective in reducing incidence and mortality from colorectal cancer (CRC). The increase in use of colonoscopy in national bowel cancer screening programmes combined with improved technology has resulted in a large increase in detection of polyps. Most polyps detected at screening colonoscopy are small (<10 mm) or diminutive (<6 mm) and, in particular the latter, have a very small chance of containing advanced features or cancer. The main reason for resecting small adenomas and sending them to histopathology serves to inform on the future surveillance intervals. Being able to diagnose adenomas in vivo would allow for them to be resected and discarded, saving the costs associated with histopathology. Diagnosing distal hyperplastic polyps in vivo would allow for these to be left in situ reducing the risks associated with polypectomy. There are now a number of new technologies that could potentially make optical diagnosis a reality. Resect and discard policy is an attractive concept for patients, gastroenterologists and health service providers and would present an enticing change to current clinical practice. © 2015 Elsevier Ltd. All rights reserved.
640
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
Introduction Colorectal cancer (CRC) is one the leading causes of morbidity and mortality in the Western world [1]. It is the second most common cancer in Europe, with 447,000 new cancer cases diagnosed in 2012 (Cancer Research UK). Colorectal cancer screening reduces the incidence and mortality from CRC and is widely recommended and implemented in the Europe and USA [2,3]. The benefit of colonoscopy is two fold e first as adenomas are thought to be precursors of CRC, their removal at colonoscopy prevents development of CRC [4]. Cohort studies of colonoscopy and polypectomy have suggested that against SEER (Surveillance Epidemiology and End Results) data, the rate of CRC detected was 76e90% lower after polypectomy than expected for the population [5,6]. An Ontario population-based cohort study of 2,412,077 individuals 50e90 years of age followed over 14 years, found that for every 1% increase in complete colonoscopy rate, the hazard of death from CRC decreased by 3% [7]. Further evidence for a protective effect of colonoscopy with polypectomy can be extrapolated from flexible sigmoidoscopy trials, with the recent randomised controlled trial [8] that enrolled 170,432 participants demonstrating that the incidence of CRC in people attending for screening was reduced by 33% (0.67; 95% CI 0.60e0.76) and mortality by 43% (0.57; 95% CI 0.45e0.72). Nishihara et al. [9] followed 88,902 participants over 22 years and found that negative index colonoscopy was associated with reduced risk of all CRC (hazard ration 0.44, 95% CI, 0.38 to 0.52) and reduced incidence of proximal CRC (hazard ratio, 0.73; 95% CI, 0.57 to 0.92). Secondly, colonoscopy allows stratification of patients into risk categories, with those with higher risk having more frequent surveillance than those at lower risk. The efficacy of endoscopic surveillance has only been addressed in epidemiological series, however those studies have suggested that patients who are not entered into a surveillance programme have a 3-4 fold greater risk of CRC [10,11]. Most polyps detected at colonoscopy are either adenomas or hyperplastic polyps. The latter ones are not thought to be pre-malignant in general. However, as white light endoscopy cannot reliably differentiate between these two types of polyp, the current standard of care dictates that all polyps seen at colonoscopy are removed and sent for histopathology. This practice has several disadvantages. First, it incurs an unnecessary time and cost associated with polypectomy and subsequent histopathology of distal hyperplastic polyps. Secondly, although complications of colonoscopy are rare, bleeding and perforation are associated with polypectomy and given the large number of colonoscopies performed for screening and subsequent surveillance this could become clinically significant [12e15]. Therefore sending small polyps for histopathology purely serves to differentiate whether they are adenomas or hyperplastic and therefore decide on surveillance intervals. Being able to differentiate adenomas from hyperplastic polyps in vivo (optical diagnosis) would allow for adenomas to be resected and discarded and small distal hyperplastic polyps to be left in situ thereby reducing the time and cost associated with polypectomy and histopathology and giving a patient a surveillance interval immediately after the procedure. Prevalence and significance of small colorectal polyps More than 90% of polyps detected at colonoscopy are small (6e9 mm) or diminutive (5 mm), with the latter making up the majority [16e18]. In a study of 13,992 asymptomatic patients who had a screening colonoscopy, 6360 (45%) patients had polyps and 83% of those had a largest polyp that was 9 mm in size [16]. Furthermore, only 2549 out of 4942 (52%) were neoplastic with the rest composed of hyperplastic and inflammatory polyps and lymphoid aggregates. Similar findings were reported in a retrospective study of 10,034 patients who underwent colonoscopy over a five-year period [18]. Polyps 5 mm represented 81.6% of all polyps removed and of those 47.9% were tubular adenomas. In a cumulative analysis [19] of 18,549 patients who had a screening colonoscopy, half of diminutive polyps were adenomas (range 49e61%). Screening series have reported adenoma prevalence of up to 50% with the use of high-definition colonoscopy [20,21]. However, this proportion might be lower in the rectum and sigmoid colon where there is high prevalence of small hyperplastic polyps, reducing the reported prevalence of adenomas to below 20% [22]. Clinical significance of small polyps is not clear. Risk of advanced features (high-grade dysplasia or villous component >25%) in small and diminutive polyps is low, ranging from 0.1% to 26% with most
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
641
estimates reported towards the lower end of the range [23]. Similarly, the rate of cancer in diminutive polyps is low although not negligible (for polyps 5 mm, the risk ranges from 0% to 0.6%, again with most estimated at the lower end of the range) [23]. To reduce the risk of missing a cancer, it is generally advised that optical diagnosis should be avoided in all lesions with any suspicion of invasive neoplasia (eg, ulcerated Paris O-IIc lesions or depressed lesions). Furthermore, data are limited on the importance of advanced pathology in small and diminutive lesions in terms of future risk of advanced pathology or cancer. As diminutive polyps are of such limited significance, radiologists reporting CT colonography do not report them [18]. Two prospective Scandinavian studies followed up 194 diminutive polyps after two and three years [24,25]. In those studies no diminutive polyp reached a size of >5 mm and no case of high-grade dysplasia or cancer were detected. A frequently cited drawback of optical diagnosis is the lack of ability to distinguish between different grades of neoplasia (high vs. low), presence of villous features and inability to diagnose the serrated lesions. Correlation between increasing size and histology of adenomas mean that these two factors often act as confounders [16]. A systemic review and meta-analysis on baseline risk factors for advanced adenomas found that there was no difference between villous vs. tubular histology [26]. Although a few individual studies have suggested that high-grade dysplasia confers a higher risk of development of advanced adenomas [27,28], a large pooled study that used individual data from six other studies [29] has not confirmed those findings. Both high-grade dysplasia and villous histology have been inconsistent in terms of being able to predict the future risk of advanced neoplasia. In addition, there is a poor agreement between even expert pathologists for the diagnosis of villosity or high-grade dysplasia with up to 10-fold variation in reported rates [30e35]. The recent interest in serrated lesions has emerged as a new ‘serrated’ pathway to colorectal cancer has been described accounting for 20% of colorectal cancer [35] where dysplastic serrated lesion (sessile serated adenomas and traditional serrated adenomas) progresses to cancer [36]. Sessile serrated adenomas resemble hyperplastic polyps [37], both morphologically and histologically and are predominantly right sided, subtle and often difficult to detect at colonoscopy and as such are often missed or incompletely resected. In a meta-analysis that included nine studies and 34,084 patients, the prevalence of serrated polyps ranged from 5.6 to 28.7% [38]. Although surface appearances are often similar to simple hyperplastic polyps the prevalence of serrated polyps/adenomas in diminutive polyps appears to be very low, 0.3e0.5% [16,18,20]. Imaging modalities available for optical diagnosis White light colonoscopy Conventional white light colonoscopy has a poor accuracy (59e84%) [39e44] in differentiating hyperplastic polyps from adenomas and therefore is not good enough for optical diagnosis. Compared to standard definition white light colonoscopy, new high definition systems have much higher resolution (charge coupled devices with >1 million pixels compared to an average of 300,000 for older standard white light colonoscopes). This high resolution, along with high definition processor and monitor produces high definition image. However this improvement in technology has not had a significant impact on the accuracy of white light for optical diagnosis of small polyps. A single randomised controlled study comparing high definition versus standard definition white light for predicting histology in 293 polyps found no difference in terms of sensitivity (76% vs. 76%, p ¼ 0.96), specificity (59% vs. 67%, p ¼ 0.44) and accuracy (70% vs 73%, p ¼ 0.6) [45] although another prospective study found HD white light to have superior accuracy in predicting adenomas (73.2% vs. 68.5%, p < 0.0001) [46]. When compared to NBI, high definition white light colonoscopy had a significantly lower sensitivity (38% vs. 96%, p < 0.0001) and accuracy (61% vs. 93%, p < 0.0001) in differentiation of adenomas from hyperplastic polyps [47]. Chromoendoscopy Chromoendoscopy involves spraying of the dyes (methylene blue, indigo carmine, cresyl violet) through the catheter via the biopsy channel to delineate colonic crypts revealing a pit pattern. Kudo
642
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
et al. described the pit pattern using magnification colonoscopy [48]. Benign, non-neoplastic lesions have pit-pattern I and II whereas neoplastic lesions have pit-patterns III, IV and V with the later one often signifying invasive neoplasia. The accuracy of chromoendoscopy in polyp characterisation is comparable to histopathology when performed by experts, with accuracies of between 80% and 96% in differentiating adenomas from hyperplastic polyps [42,43,49,50]. However, chromoendoscopy has not proved popular in the West due to perceived increases in training, time and cost associated with the procedure. Furthermore, the original pit patterns were described using magnification and these colonoscopes are not widely available. In addition, the learning curve to achieve the optical diagnosis accuracy of 90% demonstrated by experts is in the region of 200e300 histologically confirmed polyps and that has been perceived as a significant barrier to implementation [49]. Electronic chromoendoscopy Several in vivo endoscopic technologies exist that allow for real time characterisation of diminutive colonic polyps. They are integrated in the current generation colonoscopes and are easy to use, requiring a push of a button to activate. Electronic chromoendoscopy technologies enable polyp characterisation by enhancing surface pit pattern in addition to highlighting microvasculature. As angiogenesis is one of the early features of neoplasia, adenomas appear more ‘brown’ compared to the background mucosa [51] and can be seen to have the typical-appearing meshed brown capillary network of adenomas [52e54]. These technologies include narrow band imaging (NBI) (Olympus, Tokyo, Japan), i-SCAN (Pentax, Tokyo, Japan) and Fujinon Intelligent Colour Enhancement (FICE) (Fujinon Inc, Saitama, Japan) [55]. A meta-analysis [56] (Table 1) that included 56 NBI studies showed that overall sensitivity for differentiation between neoplastic and non-neoplastic polyps was 91.0% (95% CI 88.6e93.0) with specificity of 85.6% (95% CI 81.3e89.0). For FICE this meta-analysis included 14 studies, with sensitivity of 91.8% (95% CI 87.1e94.9) and specificity of 83.5% (95% CI 77.2e88.3) and for ISCAN 10 studies with sensitivity of 89.3% (95% CI 83.3e93.3) and specificity of 88.2% (95% CI 80.3e93.2). Confocal laser endomicroscopy (CLE) CLE is a system that provides magnified images of the epithelium, similar to Histopathological images, by using a miniature confocal laser microscope. The latter can be either integrated into the endoscope (Pentax, Japan) or introduced down the biopsy channel via a probe (Mauna Kea Technologies, France). Two meta-analysis have been done to evaluate the efficacy of CLE in discriminating neoplastic and non-neoplastic polyps. Wanders et al. [56] reported on 11 studies and showed an overall sensitivity to be 93.3% (95% CI 88.4e96.2) and specificity 89.9% (95% CI 81.8e94.6). Autofluorescence imaging (AFI) AFI relies on differences in mucosal blood flow and endogenous fluorophores (eg, flavins, NADPH, collagen) which change fluorescent signal emitted after being excited with short wavelength illumination. A relatively small number of AFI studies had been done assessing its efficacy in differentiation of small colorectal polyps and some of them were found to have high risk of bias in patient selection and interpretation of index test. Nevertheless the overall sensitivity reported was 86.7% (95% CI 79.5e91.6) and specificity 65.9% (95% CI 50.9e78.2).
Table 1 Pooled sensitivity, specificity and negative predictive values for real-time studies. Adapted from Ref. [57].
NBI FICE iSCAN
Number of real-time studies
Sensitivity (%)
Specificity (%)
Negative predictive value (%)
35 14 9
91.5 (88.2e93.9) 92.5 (87.6e95.6) 89.5 (82.7e93.8)
85.2 (80.0e89.3) 85.1 (78.7e89.8 89.3 (81.0e94.2)
82.5 (75.4e87.9) 83.7 (77.5e88.4) 86.5 (78.0e92.1)
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
643
Performance of imaging modalities Majority of the early studies evaluating the new electronic chromoendoscopy technologies were done using post-procedure assessments of still photographs of polyps. It is quite apparent that optical diagnosis based on a photograph does not equate to optical diagnosis done in vivo. Narrow band imaging is by far the most extensively studied of the three electronic chromoendoscopy systems. Initial studies used Kudo pit pattern to characterise polyps with NBI [39,54]. However, the assumption that Kudo pit pattern can be used with NBI was inaccurate e in one study agreement between chromoendoscopic and NBI pit pattern was only 0.23 [51]. Therefore newer classification systems have been described. East et al. [51] proposed a vascular pattern intensity classification, with strong vascular pattern intensity indicating an adenoma. A classification system (NICE) using non-magnified NBI images was developed and validated by a group of experts [57] and uses colour, vessel appearance and surface pattern to differentiate between hyperplastic polyps and adenomas (Table 2) Figs. 1 and 2. The first prospective study on the use of optical diagnosis using NBI during colonoscopy was performed a single expert endoscopist [58]. In this study of 451 consecutively identified patients, high confidence predictions of adenoma were correct for 91% of diminutive polyps and only 1/136 patients received different surveillance intervals when recommendations based on endoscopic and histopathological diagnosis were compared. Further studies from academic centres followed [59] and in 2011 American Society of Gastrointestinal Endoscopy (ASGE) published a Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) document defining performance thresholds that optical diagnosis technologies should meet before adoption in clinical practice [60]: 1. ‘In order for colorectal polyps 5 mm in size to be resected and discarded without pathologic assessment, endoscopic technology (when used with high confidence) used to determine histology of polyps 5 mm in size, when combined with the histopathologic assessment of polyps 5 mm in size, should provide a >90% agreement in assignment of post-polypectomy surveillance intervals when compared to decisions based on pathology assessment of all identified polyps. 2. In order for a technology to be used to guide the decision to leave suspected rectosigmoid hyperplastic polyps 5 mm in size in place (without resection), the technology should provide >90% negative predictive value (when used with high confidence) for adenomatous histology.’ Although PIVI was developed to help direct incorporation of endoscopic technologies into clinical practice rather than a legal standard of care, it has become a benchmark against which more recent ‘resect and discard’ studies have been assessed. ASGE Technology Committee has recently published a systematic review and meta-analysis assessing the above ASGE PIVI thresholds [61] including 20 NBI, 8 i-SCAN and 8 FICE studies. The majority of studies were conducted at academic medical centres (17/20 for NBI and 13/16 for i-SCAN and FICE). Negative predictive value was reported or could be calculated in 19 NBI studies, which gave pooled NPV of 91% (95% CI 88e94%). However this was associated with significant heterogeneity leading authors to conduct subgroup analyses. There was no significant difference noted in pooled NPV in studies conducted at academic medical centres compared to community practices although sensitivities and specificities reported from non-academic centres were significantly lower than those from academic ones with sensitivities ranging from 75 to 94% and specificity 65e76% [62e64]. However, when operator experience was taken into account when
Table 2 Adapted from Ref. [58]. Colour Vessel Surface pattern Most likely pathology
Same or lighter than background None or isolated lacy vessels coursing across the lesion Dark or white spots of uniform size or homogenous absence of pattern Hyperplastic
Brown relative to the background Thick brown vessels surrounding white structures Oval, tubular or brached white structures surrounded by brown vessels Adenoma
644
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
Fig. 1. Hyperplastic polyp viewed with white light (a) and NBI (b).
Fig. 2. Tubular adenoma viewed with white light (a) and NBI (b).
interpreting real time optical diagnosis, only experts met the PIVI threshold of 90% or higher NPV with pooled NPV 93% for experts (95% CI 91e96). Ten studies including 3082 patients reported on the degree of agreement between surveillance intervals based on histology and optical diagnosis. The overall pooled percentage of agreement was 89% (95% CI 85e93), with agreement higher for studies conducted in academic medical centres (91%) compared to community practice (82%). Again, experts reached the PIVI threshold of 90% or higher agreement (92%) compared to novice endoscopists who had lower pooled percentage of agreement (82%). Eight studies using iSCAN to characterize small colorectal polyps reported or provided information to enable NPV to be calculated. The pooled NPV was 84% (95% CI 76e91), again with experienced endoscopists meeting the PIVI NPV threshold (96% vs 72% for novices). Only one iSCAN study assessed post-polypectomy surveillance intervals and reported 69.5% agreement between histology and optical diagnosis surveillance intervals, not meeting the PIVI threshold. There were eight studies that evaluated the use of FICE in 1243 small colorectal polyps. The pooled NPV was 80% (95% CI 76e85) but in this case the operator experience did not have any effect on NPV. Only two studies assessed the degree of agreement between surveillance intervals as specified in PIVI statement e the agreement in these two studies was 100% (95% CI 91e100) and 97% (95% CI 89e100). Very few studies have assessed CLE in meeting PIVI thresholds e six studies, one of which was in abstract form reported real time classification using CLE. There was a significant variability amongst these studies and therefore a meaningful meta-analysis could not be performed. Among the CLE studies that classified polyps in real time NPV ranged from 79% to 100% with the smallest study yielding the highest NPV. Economic benefits of optical diagnosis There is moderate quality evidence from several large studies that in vivo optical diagnosis using NBI or FICE would be accost effective alternative to histopathology for diminutive polyps [59,65e67],
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
645
showing simple reduction in costs ranging from $62-210 per colonoscopy were resect and discard policy to be implemented. Kessler et al. [68] performed modelling of real time endoscopy histology followed by resection and discarding of <6 mm polyps versus resection and submission for histology on 4474 consecutive colonoscopies in which 10,400 polyps were removed (9042 of them diminutive). Assuming costs of $75 per specimen, at least $151 000 could be saved for each 1000 patients with at least one diminutive polyp and less than one in 1100 patients with diminutive polyps removed would have an undetected cancer in any removed polyp. Assuming 40% of 1.6 million annual colonoscopies performed in the USA detect at least one diminutive polyp, the potential saving from not sending these polyps for histopathological assessment is greater than $95 million. Using a ‘resect and discard’ policy for diminutive polyps would result in annual saving of $33 million when applied to colonoscopy screening of the US population (corresponding to overall saving of $330 million, assuming cumulative period of ten years to screen just under a quarter of the US population) [67]. Implementing optical diagnosis The use of vascular pattern to differentiate between neoplastic and non-neoplastic lesions appears to have a short learning curve [52,53,69], making it potentially an attractive option for optical diagnosis. In addition to NICE classification for NBI, separate classification systems have been described for FICE and iSCAN, based again on vessel and surface pattern of the polyps [70,71]. The fact that a number of studies based in community settings performed significantly below PIVI thresholds, demonstrate the ‘resect and discard’ policy is not quite ready for implementation into routine clinical practice. A prospective study from the USA involving community based gastroenterologists [63] found that only 25% of gastroenterologists assessed polyps with 90% accuracy using NBI. Although this study fulfilled the second PIVI threshold for diminutive recto-sigmoid polyps, with NPV for adenomas of 91% (95% CI 86%e97%) it failed to satisfy the first threshold with the agreement of 80% (95% CI 77%e82%) for surveillance intervals based on optical diagnosis when compared to histopathology. Similarly, two prospective studies from the Netherlands and Italy [62,64] failed to fulfil the first PIVI threshold, with agreement between surveillance intervals of only 81% and 83% respectively. One of the possible explanations for this discrepancy may be insufficient training of the less experienced endoscopists in studies based in community settings [72]. Early studies of impact of training suggested a short learning curve for using NBI for optical diagnosis. In a study by Rahavendra et al. [73] 37 physicians (including medical residents, gastroenterology fellows and faculty) participated in a 20 minute didactic teaching session on polyp characterisation using still images of polyps seen with NBI without magnification. The participants completed a pre and post training test and the accuracy of optical diagnosis and improved significantly, from 48% to 91% (p ¼ 0.001). This approach to training, however, is labour-intensive and expensive, and an alternative ways of delivering training were designed. Using a computer based training module and with pre and post training testing [74], 21 participants of different levels of experience (novices, trainees and experienced but not expert colonoscopists) evaluated NBI images of 30 small polyps. Accuracy of characterisation improved significantly in the post-training test for all participants, with trainees achieving accuracy of 90%, comparable to experts. Similarly there was an improvement in interobserver agreement following the training. However, the improvement in accuracy of optical diagnosis in both of these studies may have been driven largely by the artificial test setting in which the best polyp images were selected for inclusion in the test set, perhaps not representative of real life performance. An attempt to circumnavigate this bias by using still images was made by Rastogi et al. [75] who developed a video-based teaching module. The accuracy of optical diagnosis achieved by non-experts in both academic and community settings, although improved from pre-training testing, did not reach the level of accuracy of experts (81%). The presumption that ex vivo training does not necessarily translate into clinical practice is further supported by the study by Ladabaum et al. [63] where 12/13 community based gastroenterologists identified adenomas with >90% accuracy at the end of the training module and only 3/12 did so at the end of in vivo study. Similarly in DISCARD two study (abstract form only, personal communication), all colonoscopists achieved >90% accuracy on post-training test module but this did not translate to real-life, where sensitivity for
646
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
adenoma characterisation was 76.1% (95% CI, 72.8%e79.1%). It is likely that training will need to be improved, with some of it performed in real-time with appropriate feedback in order to achieve and maintain the required competency. The second obstacle towards implementation of optical diagnosis comes from personal views of colonoscopists. It is acknowledged that there is always resistance amongst medical professionals when dramatic changes to clinical practice are proposed [76]. A recent survey of the American College of Gastroenterology confirmed wide variability in the management of diminutive polyps [77]. Although the majority of responders (78.3%) reported that they never or rarely leave diminutive polyps in place in average-risk individuals, they were likely to leave them always or nearly always in certain scenarios, including patients with advanced age or co-morbidities, those who are anticoagulated, in patients with history of non-adenomatous polyps in the same area, and when the appearances of polyp suggests non-adenomatous histology. However the same survey showed that, despite the above, when it comes to resect and discard policy, around 50% of participants were not at all agreeable to this approach but less than 30% were not at all agreeable to leaving diminutive polyps in place if the guidelines endorsed this practice. Both European Society of Gastrointestinal Endoscopy (ESGE) and ASGE have recently moved in support of optical diagnosis. European Society of Gastrointestinal Endoscopy guidelines [23] on advanced imaging for detection and differentiation of colorectal neoplasia suggest that virtual chromoendoscopy (NBI, iSCAN and FICE) can be used under strictly controlled conditions for real time optical diagnosis of diminutive colorectal polyps. The optical diagnosis has to be reported using validated scales, must be clearly photo-documented, and can be performed only by experienced endoscopists who are adequately trained and audited. ASGE similarly ‘endorses the use of NBI for both diagnose and leave strategy for diminutive recto-sigmoid hyperplastic polyps and the resect and discard strategy for diminutive adenomatous polyps’ provided that endoscopists are trained and make assessments with high confidence [61]. This concept of ‘degree of confidence of optical diagnosis’ may provide a solution for less experienced and less confident endoscopists. First introduced by Rex [58] it essentially divides diagnoses into high and low confidence. Polyps that fulfil diagnostic criteria based on classification are diagnosed with ‘high confidence’. Polyps that do not have enough distinguishing features for an endoscopist to make the optical diagnosis with high confidence would still be resected and sent for histopathology. It has been shown that novice operators approached PIVI NPV threshold when their optical diagnosis was performed with high confidence (90%, 95% CI 86e94) whereas experienced operators exceeded both PIVI thresholds when their high confidence optical diagnoses were evaluated [61]. It is anticipated that the percentage of high confidence optical diagnoses would therefore increase with experience. Conclusion The concept of optical diagnosis is potentially attractive to endoscopists, patients and health service providers. An ability to diagnose diminutive distal hyperplastic polyps would allow for them to be left in situ reducing the risk of polypectomy and the cost of histopathology. Similarly being able to diagnose a diminutive adenoma in vivo would allow for them to be resected and discarded, without the need for associated histopathology. In addition, patients who only had polyps smaller than 6 mm that could be classified by optical diagnosis with high confidence could be given their surveillance interval immediately after the procedure. At present we are still some way off from implementing this policy in routine clinical setting in non-academic centres. An approved training programme would have to be developed along with a reliable form of accreditation and then followed up by a process of audit and on-going quality assurance. A major obstacle to implementing optical diagnosis remains the fear of litigation as the responsibility for patient outcomes is no longer shared with histopathologists. Quality assurance measures still need to be developed including adequate documentation of optical diagnosis. It still remains unclear whether high definition still images or videos would be sufficient as records but both present a storage problem for endoscopy units. However, a carefully controlled resect and discard strategy for diminutive polyps has now been endorsed both in Europe and USA, signifying a potential paradigm change to the clinical routine practice.
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
647
Practice points Standards of optical diagnosis have now been defined Optical diagnosis carries major advantages for patients and healthcare providers Electronic chromoendoscopy is easy to use and learn and accurate when performed by experts There is need for established training, accreaditation and quality assurance of optical diagnosis
Research agenda
Developing training and accreditation programmes for optical diagnosis Developing a quality assurance programme What is the role of computer aided diagnosis? Can optical diagnosis using advanced imaging differentiate between adenomas, hyperplastic polyps and serrated lesions?
Conflict of interest None. References [1] Edwards BK, Ward E, Kohler BA, et al. Annual report to the nation on the status of cancer, 1975e2006, featuring colorectal cancer trends and impact of interventions (risk factors, screening, and treatment) to reduce future rates. Cancer 2010;116: 544e73. [2] von Karsa L, Patnick J, Segnan N, et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis: overview and introduction to the full supplement publication. Endoscopy 2013;45:51e9. [3] Levin B, Lieberman DA, McFarland B, et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008;134:1570e95. [4] Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic alterations during colorectal-tumor development. N Engl J Med 1988; 319:525e32. [5] Citarda F, Tomaselli G, Capocaccia R, et al. Efficacy in standard clinical practice of colonoscopic polypectomy in reducing colorectal cancer incidence. Gut 2001;48:812e5. [6] Winawer SJ, Zauber AG, Ho MN, et al. The national polyp study. Eur J Cancer Prev 1993;2(Suppl. 2):83e7. [7] Rabeneck L, Paszat LF, Saskin R, et al. Association between colonoscopy rates and colorectal cancer mortality. Am J Gastroenterol 2010;105:1627e32. [8] Atkin WS, Edwards R, Kralj-Hans I, et al. Once-only flexible sigmoidoscopy screening in prevention of colorectal cancer: a multicentre randomised controlled trial. Lancet 2010;375:1624e33. [9] Nishihara R, Ogino S, Chan AT. Colorectal-cancer incidence and mortality after screening. N Engl J Med 2013;369:2355. [10] Cottet V, Jooste V, Fournel I, et al. Long-term risk of colorectal cancer after adenoma removal: a population-based cohort study. Gut 2012;61:1180e6. [11] Brenner H, Chang-Claude J, Jansen L, et al. Role of colonoscopy and polyp characteristics in colorectal cancer after colonoscopic polyp detection: a population-based case-control study. Ann Intern Med 2012;157:225e32. [12] Wexner SD, Garbus JE, Singh JJ. A prospective analysis of 13,580 colonoscopies. Reevaluation of credentialing guidelines. Surg Endosc 2001;15:251e61. [13] Paspatis GA, Vardas E, Theodoropoulou A, et al. Complications of colonoscopy in a large public county hospital in Greece. A 10-year study. Dig Liver Dis 2008;40:951e7. [14] Eckardt V, Kanzler G, Scmitt T, et al. Complications and adverse effects of colonoscopy with selective sedation. Gastrointest Endosc 1999;49:6. [15] Nelson D, McQuaid K, Bond J, et al. Procedural success and complications of large-scale screening colonoscopy. Gastrointest Endosc 2002;55:7. [16] Lieberman D, Moravec M, Holub J, et al. Polyp size and advanced histology in patients undergoing colonoscopy screening: implications for CT colonography. Gastroenterology 2008;135:1100e5.
648
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
[17] Chen SC, Rex DK. Endoscopist can be more powerful than age and male gender in predicting adenoma detection at colonoscopy. Am J Gastroenterol 2007;102:856e61. [18] Rex DK, Overhiser AJ, Chen SC, et al. Estimation of impact of American college of radiology recommendations on CT colonography reporting for resection of high-risk adenoma findings. Am J Gastroenterol 2009;104:149e53. [19] Hassan C, Repici A, Zullo A, et al. Colonic polyps: are we ready to resect and discard? Gastrointest Endosc Clin N Am 2013; 23:663e78. [20] Gupta N, Bansal A, Rao D, et al. Prevalence of advanced histological features in diminutive and small colon polyps. Gastrointest Endosc 2012;75:1022e30. [21] Kahi CJ, Hewett DG, Norton DL, et al. Prevalence and variable detection of proximal colon serrated polyps during screening colonoscopy. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc 2011;9:42e6. [22] Hewett DG, Huffman ME, Rex DK. Leaving distal colorectal hyperplastic polyps in place can be achieved with high accuracy by using narrow-band imaging: an observational study. Gastrointest Endosc 2012;76:374e80. [23] Kaminski MF, Hassan C, Bisschops R, et al. Advanced imaging for detection and differentiation of colorectal neoplasia: European Society of Gastrointestinal Endoscopy (ESGE) guideline. Endoscopy 2014;46:435e49. [24] Hoff G, Foerster A, Vatn MH, et al. Epidemiology of polyps in the rectum and colon. Recovery and evaluation of unresected polyps 2 years after detection. Scand J Gastroenterol 1986;21:853e62. [25] Hofstad B, Vatn MH, Andersen SN, et al. Growth of colorectal polyps: redetection and evaluation of unresected polyps for a period of three years. Gut 1996;39:449e56. [26] Saini SD, Kim HM, Schoenfeld P. Incidence of advanced adenomas at surveillance colonoscopy in patients with a personal history of colon adenomas: a meta-analysis and systematic review. Gastrointest Endosc 2006;64:614e26. [27] Bonithon-Kopp C, Piard F, Fenger C, et al. Colorectal adenoma characteristics as predictors of recurrence. Dis Colon Rectum 2004;47:323e33. [28] Lieberman DA, Weiss DG, Harford WV, et al. Five-year colon surveillance after screening colonoscopy. Gastroenterology 2007;133:1077e85. [29] Martinez ME, Baron JA, Lieberman DA, et al. A pooled analysis of advanced colorectal neoplasia diagnoses after colonoscopic polypectomy. Gastroenterology 2009;136:832e41. [30] Terry MB, Neugut AI, Bostick RM, et al. Reliability in the classification of advanced colorectal adenomas. Cancer Epidemiol Biomarkers Prev 2002;11:660e3. [31] Costantini M, Sciallero S, Giannini A, et al. Interobserver agreement in the histologic diagnosis of colorectal polyps. The experience of the multicenter adenoma colorectal study (SMAC). J Clin Epidemiol 2003;56:209e14. [32] Lasisi F, Mouchli A, Riddell R, et al. Agreement in interpreting villous elements and dysplasia in adenomas less than one centimetre in size. Dig Liver Dis 2013;45:1049e55. [33] Rex DK, Alikhan M, Cummings OW, et al. Accuracy of pathologic interpretation of colorectal polyps by general pathologists in community practice. Gastrointest Endosc 1999;50:468e74. [34] Turner JK, Williams GT, Morgan M, et al. Interobserver agreement in the reporting of colorectal polyp pathology among bowel cancer screening pathologists in Wales. Histopathology 2013;62:916e24. [35] Leggett B, Whitehall V. Role of the serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010;138: 2088e100. [36] O'Brien MJ, Yang S, Mack C, et al. Comparison of microsatellite instability, CpG island methylation phenotype, BRAF and KRAS status in serrated polyps and traditional adenomas indicates separate pathways to distinct colorectal carcinoma end points. Am J Surg Pathol 2006;30:1491e501. [37] Torlakovic E, Skovlund E, Snover DC, et al. Morphologic reappraisal of serrated colorectal polyps. Am J Surg Pathol 2003; 27:65e81. [38] Gao Q, Tsoi KK, Hirai HW, et al. Serrated polyps and the risk of synchronous colorectal advanced neoplasia: a systematic review and meta-analysis. Am J Gastroenterol 2015;110:501e9. [39] Machida H, Sano Y, Hamamoto Y, et al. Narrow-band imaging in the diagnosis of colorectal mucosal lesions: a pilot study. Endoscopy 2004;36:1094e8. [40] Apel D, Jakobs R, Schilling D, et al. Accuracy of high resolution of chromoendoscopy in prediction of histologic findings in diminutive lesions of the rectosigmoid. Gastrointest Endosc 2006;63:824e8. [41] Tischendorf JJ, Wasmuth HE, Koch A, et al. Value of magnifying chromoendoscopy and narrow band imaging (NBI) in classifying colorectal polyps: a prospective controlled study. Endoscopy 2007;39:1092e6. [42] De Palma GD, Rega M, Masone S, et al. Conventional colonoscopy and magnified chromoendoscopy for the endoscopic histological prediction of diminutive colorectal polyps: a single operator study. World J Gastroenterol 2006;12:2402e5. [43] Fu KI, Sano Y, Kato S, et al. Chromoendoscopy using indigo carmine dye spraying with magnifying observation is the most reliable method for differential diagnosis between non-neoplastic and neoplastic colorectal lesions: a prospective study. Endoscopy 2004;36:1089e93. [44] Su MY, Ho YP, Chen PC, et al. Magnifying endoscopy with indigo carmine contrast for differential diagnisis of neoplastic and non-neoplastic colonic polyps. Dig Dis Sci 2004;49:1123e7. [45] Longcroft-Wheaton G, Brown J, Cowlishaw D, et al. High-definition vs. standard-definition colonoscopy in the characterization of small colonic polyps: results from a randomized trial. Endoscopy 2012;44:905e10. [46] Rastogi A, Early DS, Gupta N, et al. Randomized, controlled trial of standard-definition white-light, high-definition whitelight, and narrow-band imaging colonoscopy for the detection of colon polyps and prediction of polyp histology. Gastrointest Endosc 2011;74:593e602. [47] Rastogi A, Keighley J, Singh V, et al. High accuracy of narrow band imaging without magnification for the real-time characterization of polyp histology and its comparison with high-definition white light colonoscopy: a prospective study. Am J Gastroenterol 2009;104:2422e30. [48] Kudo S, Tamura S, Nakajima T, et al. Diagnosis of colorectal tumorous lesions by magnifying endoscopy. Gastrointest Endosc 1996;44:8e14. [49] Togashi K, Konishi F, Ishizuka T, et al. Efficacy of magnifying endoscopy in the differential diagnosis of neoplastic and nonneoplastic polyps of the large bowel. Dis Colon Rectum 1999;42:1602e8.
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
649
[50] Konishi K, Kaneko K, Kurahashi T, et al. A comparison of magnifying and nonmagnifying colonoscopy for diagnosis of colorectal polyps: a prospective study. Gastrointest Endosc 2003;57:48e53. [51] East JE, Suzuki N, Saunders BP. Comparison of magnified pit pattern interpretation with narrow band imaging versus chromoendoscopy for diminutive colonic polyps: a pilot study. Gastrointest Endosc 2007;66:310e6. [52] Konerding MA, Fait E, Gaumann A. 3D microvascular architecture of pre-cancerous lesions and invasive carcinomas of the colon. Br J Cancer 2001;84:1354e62. [53] Sano Y, Ikematsu H, Fu KI, et al. Meshed capillary vessels by use of narrow-band imaging for differential diagnosis of small colorectal polyps. Gastrointest Endosc 2009;69:278e83. [54] Hirata M, Tanaka S, Oka S, et al. Magnifying endoscopy with narrow band imaging for diagnosis of colorectal tumors. Gastrointest Endosc 2007;65:988e95. [55] Hussain ZH, Pohl H. Ancillary imaging techniques and adenoma detection. Gastroenterol Clin North Am 2013;42:547e65. [56] Wanders LK, East JE, Uitentuis SE, et al. Diagnostic performance of narrowed spectrum endoscopy, autofluorescence imaging, and confocal laser endomicroscopy for optical diagnosis of colonic polyps: a meta-analysis. Lancet Oncol 2013; 14:1337e47. [57] Hewett DG, Kaltenbach T, Sano Y, et al. Validation of a simple classification system for endoscopic diagnosis of small colorectal polyps using narrow-band imaging. Gastroenterology 2012;143:599e607. e1. [58] Rex DK. Narrow-band imaging without optical magnification for histologic analysis of colorectal polyps. Gastroenterology 2009;136:1174e81. [59] Ignjatovic A, East JE, Suzuki N, et al. Optical diagnosis of small colorectal polyps at routine colonoscopy (Detect InSpect ChAracterise Resect and Discard; DISCARD trial): a prospective cohort study. Lancet Oncol 2009;10:1171e8. [60] Rex DK, Kahi C, O'Brien M, et al. The American Society for Gastrointestinal Endoscopy PIVI (Preservation and Incorporation of Valuable Endoscopic Innovations) on real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc 2011;73:419e22. [61] Abu Dayyeh BK, Thosani N, Konda V, et al. ASGE Technology Committee systematic review and meta-analysis assessing the ASGE PIVI thresholds for adopting real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc 2015;81. 502.e1-502.e16. [62] Kuiper T, Marsman WA, Jansen JM, et al. Accuracy for optical diagnosis of small colorectal polyps in nonacademic settings. Clin Gastroenterol Hepatol 2012;10:1016e20. quiz e79. [63] Ladabaum U, Fioritto A, Mitani A, et al. 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;144:81e91. [64] Paggi S, Rondonotti E, Amato A, et al. Resect and discard strategy in clinical practice: a prospective cohort study. Endoscopy 2012;44:899e904. [65] Gupta N, Bansal A, Rao D, et al. Accuracy of in vivo optical diagnosis of colon polyp histology by narrow-band imaging in predicting colonoscopy surveillance intervals. Gastrointest Endosc 2012;75:494e502. [66] Longcroft-Wheaton GR, Higgins B, Bhandari P. Flexible spectral imaging color enhancement and indigo carmine in neoplasia diagnosis during colonoscopy: a large prospective UK series. Eur J Gastroenterol Hepatol 2011;23:903e11. [67] Hassan C, Pickhardt PJ, Rex DK. A resect and discard strategy would improve cost-effectiveness of colorectal cancer screening. Clin Gastroenterol Hepatol 2010;8:865e9. 869 e1e3. [68] Kessler WR, Imperiale TF, Klein RW, et al. A quantitative assessment of the risks and cost savings of forgoing histologic examination of diminutive polyps. Endoscopy 2011;43:683e91. [69] Hirata M, Tanaka S, Oka S, et al. Evaluation of microvessels in colorectal tumors by narrow band imaging magnification. Gastrointest Endosc 2007;66:945e52. [70] Teixeira CR, Torresini RS, Canali C, et al. Endoscopic classification of the capillary-vessel pattern of colorectal lesions by spectral estimation technology and magnifying zoom imaging. Gastrointest Endosc 2009;69:750e6. [71] Basford PJ, Longcroft-Wheaton G, Higgins B, et al. High-definition endoscopy with i-Scan for evaluation of small colon polyps: the HiSCOPE study. Gastrointest Endosc 2014;79:111e8. [72] Rastogi A. Optical diagnosis of small colorectal polyp histology with high-definition colonoscopy using narrow band imaging. Clin Endosc 2013;46:120e9. [73] Raghavendra M, Hewett DG, Rex DK. Differentiating adenomas from hyperplastic colorectal polyps: narrow-band imaging can be learned in 20 minutes. Gastrointest Endosc 2010;72:572e6. [74] Ignjatovic A, Thomas-Gibson S, East JE, et al. 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;73:128e33. [75] Rastogi A, Rao DS, Gupta N, et al. Impact of a computer-based teaching module on characterization of diminutive colon polyps by using narrow-band imaging by non-experts in academic and community practice: a video-based study. Gastrointest Endosc 2013. [76] Hogan 3rd RB, Brill JV, Littenberg G, et al. Predict, resect, and discard… really? Gastrointest Endosc 2012;75:503e5. [77] Gellad ZF, Voils CI, Lin L, et al. Clinical practice variation in the management of diminutive colorectal polyps: results of a national survey of gastroenterologists. Am J Gastroenterol 2013;108:873e8.
Contents lists available at ScienceDirect
Best Practice & Research Clinical Gastroenterology
10
Optical diagnosis of small colorectal polyps during colonoscopy: When to resect and discard? Ana Wilson, BA, MD, MRCP Wolfson Unit for Endoscopy, St Mark's Hospital, Harrow HA1 3UJ, United Kingdom
a b s t r a c t Keywords: Colorectal polyps Optical diagnosis Electronic chromoendoscopy Narrow band imaging (NBI) Fujinon intelligent color enhancement (FICE) i-SCAN
E-mail address: [email protected]. http://dx.doi.org/10.1016/j.bpg.2015.06.007 1521-6918/© 2015 Elsevier Ltd. All rights reserved.
Colonoscopy with polypectomy has been shown to be effective in reducing incidence and mortality from colorectal cancer (CRC). The increase in use of colonoscopy in national bowel cancer screening programmes combined with improved technology has resulted in a large increase in detection of polyps. Most polyps detected at screening colonoscopy are small (<10 mm) or diminutive (<6 mm) and, in particular the latter, have a very small chance of containing advanced features or cancer. The main reason for resecting small adenomas and sending them to histopathology serves to inform on the future surveillance intervals. Being able to diagnose adenomas in vivo would allow for them to be resected and discarded, saving the costs associated with histopathology. Diagnosing distal hyperplastic polyps in vivo would allow for these to be left in situ reducing the risks associated with polypectomy. There are now a number of new technologies that could potentially make optical diagnosis a reality. Resect and discard policy is an attractive concept for patients, gastroenterologists and health service providers and would present an enticing change to current clinical practice. © 2015 Elsevier Ltd. All rights reserved.
640
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
Introduction Colorectal cancer (CRC) is one the leading causes of morbidity and mortality in the Western world [1]. It is the second most common cancer in Europe, with 447,000 new cancer cases diagnosed in 2012 (Cancer Research UK). Colorectal cancer screening reduces the incidence and mortality from CRC and is widely recommended and implemented in the Europe and USA [2,3]. The benefit of colonoscopy is two fold e first as adenomas are thought to be precursors of CRC, their removal at colonoscopy prevents development of CRC [4]. Cohort studies of colonoscopy and polypectomy have suggested that against SEER (Surveillance Epidemiology and End Results) data, the rate of CRC detected was 76e90% lower after polypectomy than expected for the population [5,6]. An Ontario population-based cohort study of 2,412,077 individuals 50e90 years of age followed over 14 years, found that for every 1% increase in complete colonoscopy rate, the hazard of death from CRC decreased by 3% [7]. Further evidence for a protective effect of colonoscopy with polypectomy can be extrapolated from flexible sigmoidoscopy trials, with the recent randomised controlled trial [8] that enrolled 170,432 participants demonstrating that the incidence of CRC in people attending for screening was reduced by 33% (0.67; 95% CI 0.60e0.76) and mortality by 43% (0.57; 95% CI 0.45e0.72). Nishihara et al. [9] followed 88,902 participants over 22 years and found that negative index colonoscopy was associated with reduced risk of all CRC (hazard ration 0.44, 95% CI, 0.38 to 0.52) and reduced incidence of proximal CRC (hazard ratio, 0.73; 95% CI, 0.57 to 0.92). Secondly, colonoscopy allows stratification of patients into risk categories, with those with higher risk having more frequent surveillance than those at lower risk. The efficacy of endoscopic surveillance has only been addressed in epidemiological series, however those studies have suggested that patients who are not entered into a surveillance programme have a 3-4 fold greater risk of CRC [10,11]. Most polyps detected at colonoscopy are either adenomas or hyperplastic polyps. The latter ones are not thought to be pre-malignant in general. However, as white light endoscopy cannot reliably differentiate between these two types of polyp, the current standard of care dictates that all polyps seen at colonoscopy are removed and sent for histopathology. This practice has several disadvantages. First, it incurs an unnecessary time and cost associated with polypectomy and subsequent histopathology of distal hyperplastic polyps. Secondly, although complications of colonoscopy are rare, bleeding and perforation are associated with polypectomy and given the large number of colonoscopies performed for screening and subsequent surveillance this could become clinically significant [12e15]. Therefore sending small polyps for histopathology purely serves to differentiate whether they are adenomas or hyperplastic and therefore decide on surveillance intervals. Being able to differentiate adenomas from hyperplastic polyps in vivo (optical diagnosis) would allow for adenomas to be resected and discarded and small distal hyperplastic polyps to be left in situ thereby reducing the time and cost associated with polypectomy and histopathology and giving a patient a surveillance interval immediately after the procedure. Prevalence and significance of small colorectal polyps More than 90% of polyps detected at colonoscopy are small (6e9 mm) or diminutive (5 mm), with the latter making up the majority [16e18]. In a study of 13,992 asymptomatic patients who had a screening colonoscopy, 6360 (45%) patients had polyps and 83% of those had a largest polyp that was 9 mm in size [16]. Furthermore, only 2549 out of 4942 (52%) were neoplastic with the rest composed of hyperplastic and inflammatory polyps and lymphoid aggregates. Similar findings were reported in a retrospective study of 10,034 patients who underwent colonoscopy over a five-year period [18]. Polyps 5 mm represented 81.6% of all polyps removed and of those 47.9% were tubular adenomas. In a cumulative analysis [19] of 18,549 patients who had a screening colonoscopy, half of diminutive polyps were adenomas (range 49e61%). Screening series have reported adenoma prevalence of up to 50% with the use of high-definition colonoscopy [20,21]. However, this proportion might be lower in the rectum and sigmoid colon where there is high prevalence of small hyperplastic polyps, reducing the reported prevalence of adenomas to below 20% [22]. Clinical significance of small polyps is not clear. Risk of advanced features (high-grade dysplasia or villous component >25%) in small and diminutive polyps is low, ranging from 0.1% to 26% with most
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
641
estimates reported towards the lower end of the range [23]. Similarly, the rate of cancer in diminutive polyps is low although not negligible (for polyps 5 mm, the risk ranges from 0% to 0.6%, again with most estimated at the lower end of the range) [23]. To reduce the risk of missing a cancer, it is generally advised that optical diagnosis should be avoided in all lesions with any suspicion of invasive neoplasia (eg, ulcerated Paris O-IIc lesions or depressed lesions). Furthermore, data are limited on the importance of advanced pathology in small and diminutive lesions in terms of future risk of advanced pathology or cancer. As diminutive polyps are of such limited significance, radiologists reporting CT colonography do not report them [18]. Two prospective Scandinavian studies followed up 194 diminutive polyps after two and three years [24,25]. In those studies no diminutive polyp reached a size of >5 mm and no case of high-grade dysplasia or cancer were detected. A frequently cited drawback of optical diagnosis is the lack of ability to distinguish between different grades of neoplasia (high vs. low), presence of villous features and inability to diagnose the serrated lesions. Correlation between increasing size and histology of adenomas mean that these two factors often act as confounders [16]. A systemic review and meta-analysis on baseline risk factors for advanced adenomas found that there was no difference between villous vs. tubular histology [26]. Although a few individual studies have suggested that high-grade dysplasia confers a higher risk of development of advanced adenomas [27,28], a large pooled study that used individual data from six other studies [29] has not confirmed those findings. Both high-grade dysplasia and villous histology have been inconsistent in terms of being able to predict the future risk of advanced neoplasia. In addition, there is a poor agreement between even expert pathologists for the diagnosis of villosity or high-grade dysplasia with up to 10-fold variation in reported rates [30e35]. The recent interest in serrated lesions has emerged as a new ‘serrated’ pathway to colorectal cancer has been described accounting for 20% of colorectal cancer [35] where dysplastic serrated lesion (sessile serated adenomas and traditional serrated adenomas) progresses to cancer [36]. Sessile serrated adenomas resemble hyperplastic polyps [37], both morphologically and histologically and are predominantly right sided, subtle and often difficult to detect at colonoscopy and as such are often missed or incompletely resected. In a meta-analysis that included nine studies and 34,084 patients, the prevalence of serrated polyps ranged from 5.6 to 28.7% [38]. Although surface appearances are often similar to simple hyperplastic polyps the prevalence of serrated polyps/adenomas in diminutive polyps appears to be very low, 0.3e0.5% [16,18,20]. Imaging modalities available for optical diagnosis White light colonoscopy Conventional white light colonoscopy has a poor accuracy (59e84%) [39e44] in differentiating hyperplastic polyps from adenomas and therefore is not good enough for optical diagnosis. Compared to standard definition white light colonoscopy, new high definition systems have much higher resolution (charge coupled devices with >1 million pixels compared to an average of 300,000 for older standard white light colonoscopes). This high resolution, along with high definition processor and monitor produces high definition image. However this improvement in technology has not had a significant impact on the accuracy of white light for optical diagnosis of small polyps. A single randomised controlled study comparing high definition versus standard definition white light for predicting histology in 293 polyps found no difference in terms of sensitivity (76% vs. 76%, p ¼ 0.96), specificity (59% vs. 67%, p ¼ 0.44) and accuracy (70% vs 73%, p ¼ 0.6) [45] although another prospective study found HD white light to have superior accuracy in predicting adenomas (73.2% vs. 68.5%, p < 0.0001) [46]. When compared to NBI, high definition white light colonoscopy had a significantly lower sensitivity (38% vs. 96%, p < 0.0001) and accuracy (61% vs. 93%, p < 0.0001) in differentiation of adenomas from hyperplastic polyps [47]. Chromoendoscopy Chromoendoscopy involves spraying of the dyes (methylene blue, indigo carmine, cresyl violet) through the catheter via the biopsy channel to delineate colonic crypts revealing a pit pattern. Kudo
642
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
et al. described the pit pattern using magnification colonoscopy [48]. Benign, non-neoplastic lesions have pit-pattern I and II whereas neoplastic lesions have pit-patterns III, IV and V with the later one often signifying invasive neoplasia. The accuracy of chromoendoscopy in polyp characterisation is comparable to histopathology when performed by experts, with accuracies of between 80% and 96% in differentiating adenomas from hyperplastic polyps [42,43,49,50]. However, chromoendoscopy has not proved popular in the West due to perceived increases in training, time and cost associated with the procedure. Furthermore, the original pit patterns were described using magnification and these colonoscopes are not widely available. In addition, the learning curve to achieve the optical diagnosis accuracy of 90% demonstrated by experts is in the region of 200e300 histologically confirmed polyps and that has been perceived as a significant barrier to implementation [49]. Electronic chromoendoscopy Several in vivo endoscopic technologies exist that allow for real time characterisation of diminutive colonic polyps. They are integrated in the current generation colonoscopes and are easy to use, requiring a push of a button to activate. Electronic chromoendoscopy technologies enable polyp characterisation by enhancing surface pit pattern in addition to highlighting microvasculature. As angiogenesis is one of the early features of neoplasia, adenomas appear more ‘brown’ compared to the background mucosa [51] and can be seen to have the typical-appearing meshed brown capillary network of adenomas [52e54]. These technologies include narrow band imaging (NBI) (Olympus, Tokyo, Japan), i-SCAN (Pentax, Tokyo, Japan) and Fujinon Intelligent Colour Enhancement (FICE) (Fujinon Inc, Saitama, Japan) [55]. A meta-analysis [56] (Table 1) that included 56 NBI studies showed that overall sensitivity for differentiation between neoplastic and non-neoplastic polyps was 91.0% (95% CI 88.6e93.0) with specificity of 85.6% (95% CI 81.3e89.0). For FICE this meta-analysis included 14 studies, with sensitivity of 91.8% (95% CI 87.1e94.9) and specificity of 83.5% (95% CI 77.2e88.3) and for ISCAN 10 studies with sensitivity of 89.3% (95% CI 83.3e93.3) and specificity of 88.2% (95% CI 80.3e93.2). Confocal laser endomicroscopy (CLE) CLE is a system that provides magnified images of the epithelium, similar to Histopathological images, by using a miniature confocal laser microscope. The latter can be either integrated into the endoscope (Pentax, Japan) or introduced down the biopsy channel via a probe (Mauna Kea Technologies, France). Two meta-analysis have been done to evaluate the efficacy of CLE in discriminating neoplastic and non-neoplastic polyps. Wanders et al. [56] reported on 11 studies and showed an overall sensitivity to be 93.3% (95% CI 88.4e96.2) and specificity 89.9% (95% CI 81.8e94.6). Autofluorescence imaging (AFI) AFI relies on differences in mucosal blood flow and endogenous fluorophores (eg, flavins, NADPH, collagen) which change fluorescent signal emitted after being excited with short wavelength illumination. A relatively small number of AFI studies had been done assessing its efficacy in differentiation of small colorectal polyps and some of them were found to have high risk of bias in patient selection and interpretation of index test. Nevertheless the overall sensitivity reported was 86.7% (95% CI 79.5e91.6) and specificity 65.9% (95% CI 50.9e78.2).
Table 1 Pooled sensitivity, specificity and negative predictive values for real-time studies. Adapted from Ref. [57].
NBI FICE iSCAN
Number of real-time studies
Sensitivity (%)
Specificity (%)
Negative predictive value (%)
35 14 9
91.5 (88.2e93.9) 92.5 (87.6e95.6) 89.5 (82.7e93.8)
85.2 (80.0e89.3) 85.1 (78.7e89.8 89.3 (81.0e94.2)
82.5 (75.4e87.9) 83.7 (77.5e88.4) 86.5 (78.0e92.1)
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
643
Performance of imaging modalities Majority of the early studies evaluating the new electronic chromoendoscopy technologies were done using post-procedure assessments of still photographs of polyps. It is quite apparent that optical diagnosis based on a photograph does not equate to optical diagnosis done in vivo. Narrow band imaging is by far the most extensively studied of the three electronic chromoendoscopy systems. Initial studies used Kudo pit pattern to characterise polyps with NBI [39,54]. However, the assumption that Kudo pit pattern can be used with NBI was inaccurate e in one study agreement between chromoendoscopic and NBI pit pattern was only 0.23 [51]. Therefore newer classification systems have been described. East et al. [51] proposed a vascular pattern intensity classification, with strong vascular pattern intensity indicating an adenoma. A classification system (NICE) using non-magnified NBI images was developed and validated by a group of experts [57] and uses colour, vessel appearance and surface pattern to differentiate between hyperplastic polyps and adenomas (Table 2) Figs. 1 and 2. The first prospective study on the use of optical diagnosis using NBI during colonoscopy was performed a single expert endoscopist [58]. In this study of 451 consecutively identified patients, high confidence predictions of adenoma were correct for 91% of diminutive polyps and only 1/136 patients received different surveillance intervals when recommendations based on endoscopic and histopathological diagnosis were compared. Further studies from academic centres followed [59] and in 2011 American Society of Gastrointestinal Endoscopy (ASGE) published a Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) document defining performance thresholds that optical diagnosis technologies should meet before adoption in clinical practice [60]: 1. ‘In order for colorectal polyps 5 mm in size to be resected and discarded without pathologic assessment, endoscopic technology (when used with high confidence) used to determine histology of polyps 5 mm in size, when combined with the histopathologic assessment of polyps 5 mm in size, should provide a >90% agreement in assignment of post-polypectomy surveillance intervals when compared to decisions based on pathology assessment of all identified polyps. 2. In order for a technology to be used to guide the decision to leave suspected rectosigmoid hyperplastic polyps 5 mm in size in place (without resection), the technology should provide >90% negative predictive value (when used with high confidence) for adenomatous histology.’ Although PIVI was developed to help direct incorporation of endoscopic technologies into clinical practice rather than a legal standard of care, it has become a benchmark against which more recent ‘resect and discard’ studies have been assessed. ASGE Technology Committee has recently published a systematic review and meta-analysis assessing the above ASGE PIVI thresholds [61] including 20 NBI, 8 i-SCAN and 8 FICE studies. The majority of studies were conducted at academic medical centres (17/20 for NBI and 13/16 for i-SCAN and FICE). Negative predictive value was reported or could be calculated in 19 NBI studies, which gave pooled NPV of 91% (95% CI 88e94%). However this was associated with significant heterogeneity leading authors to conduct subgroup analyses. There was no significant difference noted in pooled NPV in studies conducted at academic medical centres compared to community practices although sensitivities and specificities reported from non-academic centres were significantly lower than those from academic ones with sensitivities ranging from 75 to 94% and specificity 65e76% [62e64]. However, when operator experience was taken into account when
Table 2 Adapted from Ref. [58]. Colour Vessel Surface pattern Most likely pathology
Same or lighter than background None or isolated lacy vessels coursing across the lesion Dark or white spots of uniform size or homogenous absence of pattern Hyperplastic
Brown relative to the background Thick brown vessels surrounding white structures Oval, tubular or brached white structures surrounded by brown vessels Adenoma
644
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
Fig. 1. Hyperplastic polyp viewed with white light (a) and NBI (b).
Fig. 2. Tubular adenoma viewed with white light (a) and NBI (b).
interpreting real time optical diagnosis, only experts met the PIVI threshold of 90% or higher NPV with pooled NPV 93% for experts (95% CI 91e96). Ten studies including 3082 patients reported on the degree of agreement between surveillance intervals based on histology and optical diagnosis. The overall pooled percentage of agreement was 89% (95% CI 85e93), with agreement higher for studies conducted in academic medical centres (91%) compared to community practice (82%). Again, experts reached the PIVI threshold of 90% or higher agreement (92%) compared to novice endoscopists who had lower pooled percentage of agreement (82%). Eight studies using iSCAN to characterize small colorectal polyps reported or provided information to enable NPV to be calculated. The pooled NPV was 84% (95% CI 76e91), again with experienced endoscopists meeting the PIVI NPV threshold (96% vs 72% for novices). Only one iSCAN study assessed post-polypectomy surveillance intervals and reported 69.5% agreement between histology and optical diagnosis surveillance intervals, not meeting the PIVI threshold. There were eight studies that evaluated the use of FICE in 1243 small colorectal polyps. The pooled NPV was 80% (95% CI 76e85) but in this case the operator experience did not have any effect on NPV. Only two studies assessed the degree of agreement between surveillance intervals as specified in PIVI statement e the agreement in these two studies was 100% (95% CI 91e100) and 97% (95% CI 89e100). Very few studies have assessed CLE in meeting PIVI thresholds e six studies, one of which was in abstract form reported real time classification using CLE. There was a significant variability amongst these studies and therefore a meaningful meta-analysis could not be performed. Among the CLE studies that classified polyps in real time NPV ranged from 79% to 100% with the smallest study yielding the highest NPV. Economic benefits of optical diagnosis There is moderate quality evidence from several large studies that in vivo optical diagnosis using NBI or FICE would be accost effective alternative to histopathology for diminutive polyps [59,65e67],
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
645
showing simple reduction in costs ranging from $62-210 per colonoscopy were resect and discard policy to be implemented. Kessler et al. [68] performed modelling of real time endoscopy histology followed by resection and discarding of <6 mm polyps versus resection and submission for histology on 4474 consecutive colonoscopies in which 10,400 polyps were removed (9042 of them diminutive). Assuming costs of $75 per specimen, at least $151 000 could be saved for each 1000 patients with at least one diminutive polyp and less than one in 1100 patients with diminutive polyps removed would have an undetected cancer in any removed polyp. Assuming 40% of 1.6 million annual colonoscopies performed in the USA detect at least one diminutive polyp, the potential saving from not sending these polyps for histopathological assessment is greater than $95 million. Using a ‘resect and discard’ policy for diminutive polyps would result in annual saving of $33 million when applied to colonoscopy screening of the US population (corresponding to overall saving of $330 million, assuming cumulative period of ten years to screen just under a quarter of the US population) [67]. Implementing optical diagnosis The use of vascular pattern to differentiate between neoplastic and non-neoplastic lesions appears to have a short learning curve [52,53,69], making it potentially an attractive option for optical diagnosis. In addition to NICE classification for NBI, separate classification systems have been described for FICE and iSCAN, based again on vessel and surface pattern of the polyps [70,71]. The fact that a number of studies based in community settings performed significantly below PIVI thresholds, demonstrate the ‘resect and discard’ policy is not quite ready for implementation into routine clinical practice. A prospective study from the USA involving community based gastroenterologists [63] found that only 25% of gastroenterologists assessed polyps with 90% accuracy using NBI. Although this study fulfilled the second PIVI threshold for diminutive recto-sigmoid polyps, with NPV for adenomas of 91% (95% CI 86%e97%) it failed to satisfy the first threshold with the agreement of 80% (95% CI 77%e82%) for surveillance intervals based on optical diagnosis when compared to histopathology. Similarly, two prospective studies from the Netherlands and Italy [62,64] failed to fulfil the first PIVI threshold, with agreement between surveillance intervals of only 81% and 83% respectively. One of the possible explanations for this discrepancy may be insufficient training of the less experienced endoscopists in studies based in community settings [72]. Early studies of impact of training suggested a short learning curve for using NBI for optical diagnosis. In a study by Rahavendra et al. [73] 37 physicians (including medical residents, gastroenterology fellows and faculty) participated in a 20 minute didactic teaching session on polyp characterisation using still images of polyps seen with NBI without magnification. The participants completed a pre and post training test and the accuracy of optical diagnosis and improved significantly, from 48% to 91% (p ¼ 0.001). This approach to training, however, is labour-intensive and expensive, and an alternative ways of delivering training were designed. Using a computer based training module and with pre and post training testing [74], 21 participants of different levels of experience (novices, trainees and experienced but not expert colonoscopists) evaluated NBI images of 30 small polyps. Accuracy of characterisation improved significantly in the post-training test for all participants, with trainees achieving accuracy of 90%, comparable to experts. Similarly there was an improvement in interobserver agreement following the training. However, the improvement in accuracy of optical diagnosis in both of these studies may have been driven largely by the artificial test setting in which the best polyp images were selected for inclusion in the test set, perhaps not representative of real life performance. An attempt to circumnavigate this bias by using still images was made by Rastogi et al. [75] who developed a video-based teaching module. The accuracy of optical diagnosis achieved by non-experts in both academic and community settings, although improved from pre-training testing, did not reach the level of accuracy of experts (81%). The presumption that ex vivo training does not necessarily translate into clinical practice is further supported by the study by Ladabaum et al. [63] where 12/13 community based gastroenterologists identified adenomas with >90% accuracy at the end of the training module and only 3/12 did so at the end of in vivo study. Similarly in DISCARD two study (abstract form only, personal communication), all colonoscopists achieved >90% accuracy on post-training test module but this did not translate to real-life, where sensitivity for
646
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
adenoma characterisation was 76.1% (95% CI, 72.8%e79.1%). It is likely that training will need to be improved, with some of it performed in real-time with appropriate feedback in order to achieve and maintain the required competency. The second obstacle towards implementation of optical diagnosis comes from personal views of colonoscopists. It is acknowledged that there is always resistance amongst medical professionals when dramatic changes to clinical practice are proposed [76]. A recent survey of the American College of Gastroenterology confirmed wide variability in the management of diminutive polyps [77]. Although the majority of responders (78.3%) reported that they never or rarely leave diminutive polyps in place in average-risk individuals, they were likely to leave them always or nearly always in certain scenarios, including patients with advanced age or co-morbidities, those who are anticoagulated, in patients with history of non-adenomatous polyps in the same area, and when the appearances of polyp suggests non-adenomatous histology. However the same survey showed that, despite the above, when it comes to resect and discard policy, around 50% of participants were not at all agreeable to this approach but less than 30% were not at all agreeable to leaving diminutive polyps in place if the guidelines endorsed this practice. Both European Society of Gastrointestinal Endoscopy (ESGE) and ASGE have recently moved in support of optical diagnosis. European Society of Gastrointestinal Endoscopy guidelines [23] on advanced imaging for detection and differentiation of colorectal neoplasia suggest that virtual chromoendoscopy (NBI, iSCAN and FICE) can be used under strictly controlled conditions for real time optical diagnosis of diminutive colorectal polyps. The optical diagnosis has to be reported using validated scales, must be clearly photo-documented, and can be performed only by experienced endoscopists who are adequately trained and audited. ASGE similarly ‘endorses the use of NBI for both diagnose and leave strategy for diminutive recto-sigmoid hyperplastic polyps and the resect and discard strategy for diminutive adenomatous polyps’ provided that endoscopists are trained and make assessments with high confidence [61]. This concept of ‘degree of confidence of optical diagnosis’ may provide a solution for less experienced and less confident endoscopists. First introduced by Rex [58] it essentially divides diagnoses into high and low confidence. Polyps that fulfil diagnostic criteria based on classification are diagnosed with ‘high confidence’. Polyps that do not have enough distinguishing features for an endoscopist to make the optical diagnosis with high confidence would still be resected and sent for histopathology. It has been shown that novice operators approached PIVI NPV threshold when their optical diagnosis was performed with high confidence (90%, 95% CI 86e94) whereas experienced operators exceeded both PIVI thresholds when their high confidence optical diagnoses were evaluated [61]. It is anticipated that the percentage of high confidence optical diagnoses would therefore increase with experience. Conclusion The concept of optical diagnosis is potentially attractive to endoscopists, patients and health service providers. An ability to diagnose diminutive distal hyperplastic polyps would allow for them to be left in situ reducing the risk of polypectomy and the cost of histopathology. Similarly being able to diagnose a diminutive adenoma in vivo would allow for them to be resected and discarded, without the need for associated histopathology. In addition, patients who only had polyps smaller than 6 mm that could be classified by optical diagnosis with high confidence could be given their surveillance interval immediately after the procedure. At present we are still some way off from implementing this policy in routine clinical setting in non-academic centres. An approved training programme would have to be developed along with a reliable form of accreditation and then followed up by a process of audit and on-going quality assurance. A major obstacle to implementing optical diagnosis remains the fear of litigation as the responsibility for patient outcomes is no longer shared with histopathologists. Quality assurance measures still need to be developed including adequate documentation of optical diagnosis. It still remains unclear whether high definition still images or videos would be sufficient as records but both present a storage problem for endoscopy units. However, a carefully controlled resect and discard strategy for diminutive polyps has now been endorsed both in Europe and USA, signifying a potential paradigm change to the clinical routine practice.
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
647
Practice points Standards of optical diagnosis have now been defined Optical diagnosis carries major advantages for patients and healthcare providers Electronic chromoendoscopy is easy to use and learn and accurate when performed by experts There is need for established training, accreaditation and quality assurance of optical diagnosis
Research agenda
Developing training and accreditation programmes for optical diagnosis Developing a quality assurance programme What is the role of computer aided diagnosis? Can optical diagnosis using advanced imaging differentiate between adenomas, hyperplastic polyps and serrated lesions?
Conflict of interest None. References [1] Edwards BK, Ward E, Kohler BA, et al. Annual report to the nation on the status of cancer, 1975e2006, featuring colorectal cancer trends and impact of interventions (risk factors, screening, and treatment) to reduce future rates. Cancer 2010;116: 544e73. [2] von Karsa L, Patnick J, Segnan N, et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis: overview and introduction to the full supplement publication. Endoscopy 2013;45:51e9. [3] Levin B, Lieberman DA, McFarland B, et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008;134:1570e95. [4] Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic alterations during colorectal-tumor development. N Engl J Med 1988; 319:525e32. [5] Citarda F, Tomaselli G, Capocaccia R, et al. Efficacy in standard clinical practice of colonoscopic polypectomy in reducing colorectal cancer incidence. Gut 2001;48:812e5. [6] Winawer SJ, Zauber AG, Ho MN, et al. The national polyp study. Eur J Cancer Prev 1993;2(Suppl. 2):83e7. [7] Rabeneck L, Paszat LF, Saskin R, et al. Association between colonoscopy rates and colorectal cancer mortality. Am J Gastroenterol 2010;105:1627e32. [8] Atkin WS, Edwards R, Kralj-Hans I, et al. Once-only flexible sigmoidoscopy screening in prevention of colorectal cancer: a multicentre randomised controlled trial. Lancet 2010;375:1624e33. [9] Nishihara R, Ogino S, Chan AT. Colorectal-cancer incidence and mortality after screening. N Engl J Med 2013;369:2355. [10] Cottet V, Jooste V, Fournel I, et al. Long-term risk of colorectal cancer after adenoma removal: a population-based cohort study. Gut 2012;61:1180e6. [11] Brenner H, Chang-Claude J, Jansen L, et al. Role of colonoscopy and polyp characteristics in colorectal cancer after colonoscopic polyp detection: a population-based case-control study. Ann Intern Med 2012;157:225e32. [12] Wexner SD, Garbus JE, Singh JJ. A prospective analysis of 13,580 colonoscopies. Reevaluation of credentialing guidelines. Surg Endosc 2001;15:251e61. [13] Paspatis GA, Vardas E, Theodoropoulou A, et al. Complications of colonoscopy in a large public county hospital in Greece. A 10-year study. Dig Liver Dis 2008;40:951e7. [14] Eckardt V, Kanzler G, Scmitt T, et al. Complications and adverse effects of colonoscopy with selective sedation. Gastrointest Endosc 1999;49:6. [15] Nelson D, McQuaid K, Bond J, et al. Procedural success and complications of large-scale screening colonoscopy. Gastrointest Endosc 2002;55:7. [16] Lieberman D, Moravec M, Holub J, et al. Polyp size and advanced histology in patients undergoing colonoscopy screening: implications for CT colonography. Gastroenterology 2008;135:1100e5.
648
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
[17] Chen SC, Rex DK. Endoscopist can be more powerful than age and male gender in predicting adenoma detection at colonoscopy. Am J Gastroenterol 2007;102:856e61. [18] Rex DK, Overhiser AJ, Chen SC, et al. Estimation of impact of American college of radiology recommendations on CT colonography reporting for resection of high-risk adenoma findings. Am J Gastroenterol 2009;104:149e53. [19] Hassan C, Repici A, Zullo A, et al. Colonic polyps: are we ready to resect and discard? Gastrointest Endosc Clin N Am 2013; 23:663e78. [20] Gupta N, Bansal A, Rao D, et al. Prevalence of advanced histological features in diminutive and small colon polyps. Gastrointest Endosc 2012;75:1022e30. [21] Kahi CJ, Hewett DG, Norton DL, et al. Prevalence and variable detection of proximal colon serrated polyps during screening colonoscopy. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc 2011;9:42e6. [22] Hewett DG, Huffman ME, Rex DK. Leaving distal colorectal hyperplastic polyps in place can be achieved with high accuracy by using narrow-band imaging: an observational study. Gastrointest Endosc 2012;76:374e80. [23] Kaminski MF, Hassan C, Bisschops R, et al. Advanced imaging for detection and differentiation of colorectal neoplasia: European Society of Gastrointestinal Endoscopy (ESGE) guideline. Endoscopy 2014;46:435e49. [24] Hoff G, Foerster A, Vatn MH, et al. Epidemiology of polyps in the rectum and colon. Recovery and evaluation of unresected polyps 2 years after detection. Scand J Gastroenterol 1986;21:853e62. [25] Hofstad B, Vatn MH, Andersen SN, et al. Growth of colorectal polyps: redetection and evaluation of unresected polyps for a period of three years. Gut 1996;39:449e56. [26] Saini SD, Kim HM, Schoenfeld P. Incidence of advanced adenomas at surveillance colonoscopy in patients with a personal history of colon adenomas: a meta-analysis and systematic review. Gastrointest Endosc 2006;64:614e26. [27] Bonithon-Kopp C, Piard F, Fenger C, et al. Colorectal adenoma characteristics as predictors of recurrence. Dis Colon Rectum 2004;47:323e33. [28] Lieberman DA, Weiss DG, Harford WV, et al. Five-year colon surveillance after screening colonoscopy. Gastroenterology 2007;133:1077e85. [29] Martinez ME, Baron JA, Lieberman DA, et al. A pooled analysis of advanced colorectal neoplasia diagnoses after colonoscopic polypectomy. Gastroenterology 2009;136:832e41. [30] Terry MB, Neugut AI, Bostick RM, et al. Reliability in the classification of advanced colorectal adenomas. Cancer Epidemiol Biomarkers Prev 2002;11:660e3. [31] Costantini M, Sciallero S, Giannini A, et al. Interobserver agreement in the histologic diagnosis of colorectal polyps. The experience of the multicenter adenoma colorectal study (SMAC). J Clin Epidemiol 2003;56:209e14. [32] Lasisi F, Mouchli A, Riddell R, et al. Agreement in interpreting villous elements and dysplasia in adenomas less than one centimetre in size. Dig Liver Dis 2013;45:1049e55. [33] Rex DK, Alikhan M, Cummings OW, et al. Accuracy of pathologic interpretation of colorectal polyps by general pathologists in community practice. Gastrointest Endosc 1999;50:468e74. [34] Turner JK, Williams GT, Morgan M, et al. Interobserver agreement in the reporting of colorectal polyp pathology among bowel cancer screening pathologists in Wales. Histopathology 2013;62:916e24. [35] Leggett B, Whitehall V. Role of the serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010;138: 2088e100. [36] O'Brien MJ, Yang S, Mack C, et al. Comparison of microsatellite instability, CpG island methylation phenotype, BRAF and KRAS status in serrated polyps and traditional adenomas indicates separate pathways to distinct colorectal carcinoma end points. Am J Surg Pathol 2006;30:1491e501. [37] Torlakovic E, Skovlund E, Snover DC, et al. Morphologic reappraisal of serrated colorectal polyps. Am J Surg Pathol 2003; 27:65e81. [38] Gao Q, Tsoi KK, Hirai HW, et al. Serrated polyps and the risk of synchronous colorectal advanced neoplasia: a systematic review and meta-analysis. Am J Gastroenterol 2015;110:501e9. [39] Machida H, Sano Y, Hamamoto Y, et al. Narrow-band imaging in the diagnosis of colorectal mucosal lesions: a pilot study. Endoscopy 2004;36:1094e8. [40] Apel D, Jakobs R, Schilling D, et al. Accuracy of high resolution of chromoendoscopy in prediction of histologic findings in diminutive lesions of the rectosigmoid. Gastrointest Endosc 2006;63:824e8. [41] Tischendorf JJ, Wasmuth HE, Koch A, et al. Value of magnifying chromoendoscopy and narrow band imaging (NBI) in classifying colorectal polyps: a prospective controlled study. Endoscopy 2007;39:1092e6. [42] De Palma GD, Rega M, Masone S, et al. Conventional colonoscopy and magnified chromoendoscopy for the endoscopic histological prediction of diminutive colorectal polyps: a single operator study. World J Gastroenterol 2006;12:2402e5. [43] Fu KI, Sano Y, Kato S, et al. Chromoendoscopy using indigo carmine dye spraying with magnifying observation is the most reliable method for differential diagnosis between non-neoplastic and neoplastic colorectal lesions: a prospective study. Endoscopy 2004;36:1089e93. [44] Su MY, Ho YP, Chen PC, et al. Magnifying endoscopy with indigo carmine contrast for differential diagnisis of neoplastic and non-neoplastic colonic polyps. Dig Dis Sci 2004;49:1123e7. [45] Longcroft-Wheaton G, Brown J, Cowlishaw D, et al. High-definition vs. standard-definition colonoscopy in the characterization of small colonic polyps: results from a randomized trial. Endoscopy 2012;44:905e10. [46] Rastogi A, Early DS, Gupta N, et al. Randomized, controlled trial of standard-definition white-light, high-definition whitelight, and narrow-band imaging colonoscopy for the detection of colon polyps and prediction of polyp histology. Gastrointest Endosc 2011;74:593e602. [47] Rastogi A, Keighley J, Singh V, et al. High accuracy of narrow band imaging without magnification for the real-time characterization of polyp histology and its comparison with high-definition white light colonoscopy: a prospective study. Am J Gastroenterol 2009;104:2422e30. [48] Kudo S, Tamura S, Nakajima T, et al. Diagnosis of colorectal tumorous lesions by magnifying endoscopy. Gastrointest Endosc 1996;44:8e14. [49] Togashi K, Konishi F, Ishizuka T, et al. Efficacy of magnifying endoscopy in the differential diagnosis of neoplastic and nonneoplastic polyps of the large bowel. Dis Colon Rectum 1999;42:1602e8.
A. Wilson / Best Practice & Research Clinical Gastroenterology 29 (2015) 639e649
649
[50] Konishi K, Kaneko K, Kurahashi T, et al. A comparison of magnifying and nonmagnifying colonoscopy for diagnosis of colorectal polyps: a prospective study. Gastrointest Endosc 2003;57:48e53. [51] East JE, Suzuki N, Saunders BP. Comparison of magnified pit pattern interpretation with narrow band imaging versus chromoendoscopy for diminutive colonic polyps: a pilot study. Gastrointest Endosc 2007;66:310e6. [52] Konerding MA, Fait E, Gaumann A. 3D microvascular architecture of pre-cancerous lesions and invasive carcinomas of the colon. Br J Cancer 2001;84:1354e62. [53] Sano Y, Ikematsu H, Fu KI, et al. Meshed capillary vessels by use of narrow-band imaging for differential diagnosis of small colorectal polyps. Gastrointest Endosc 2009;69:278e83. [54] Hirata M, Tanaka S, Oka S, et al. Magnifying endoscopy with narrow band imaging for diagnosis of colorectal tumors. Gastrointest Endosc 2007;65:988e95. [55] Hussain ZH, Pohl H. Ancillary imaging techniques and adenoma detection. Gastroenterol Clin North Am 2013;42:547e65. [56] Wanders LK, East JE, Uitentuis SE, et al. Diagnostic performance of narrowed spectrum endoscopy, autofluorescence imaging, and confocal laser endomicroscopy for optical diagnosis of colonic polyps: a meta-analysis. Lancet Oncol 2013; 14:1337e47. [57] Hewett DG, Kaltenbach T, Sano Y, et al. Validation of a simple classification system for endoscopic diagnosis of small colorectal polyps using narrow-band imaging. Gastroenterology 2012;143:599e607. e1. [58] Rex DK. Narrow-band imaging without optical magnification for histologic analysis of colorectal polyps. Gastroenterology 2009;136:1174e81. [59] Ignjatovic A, East JE, Suzuki N, et al. Optical diagnosis of small colorectal polyps at routine colonoscopy (Detect InSpect ChAracterise Resect and Discard; DISCARD trial): a prospective cohort study. Lancet Oncol 2009;10:1171e8. [60] Rex DK, Kahi C, O'Brien M, et al. The American Society for Gastrointestinal Endoscopy PIVI (Preservation and Incorporation of Valuable Endoscopic Innovations) on real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc 2011;73:419e22. [61] Abu Dayyeh BK, Thosani N, Konda V, et al. ASGE Technology Committee systematic review and meta-analysis assessing the ASGE PIVI thresholds for adopting real-time endoscopic assessment of the histology of diminutive colorectal polyps. Gastrointest Endosc 2015;81. 502.e1-502.e16. [62] Kuiper T, Marsman WA, Jansen JM, et al. Accuracy for optical diagnosis of small colorectal polyps in nonacademic settings. Clin Gastroenterol Hepatol 2012;10:1016e20. quiz e79. [63] Ladabaum U, Fioritto A, Mitani A, et al. 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;144:81e91. [64] Paggi S, Rondonotti E, Amato A, et al. Resect and discard strategy in clinical practice: a prospective cohort study. Endoscopy 2012;44:899e904. [65] Gupta N, Bansal A, Rao D, et al. Accuracy of in vivo optical diagnosis of colon polyp histology by narrow-band imaging in predicting colonoscopy surveillance intervals. Gastrointest Endosc 2012;75:494e502. [66] Longcroft-Wheaton GR, Higgins B, Bhandari P. Flexible spectral imaging color enhancement and indigo carmine in neoplasia diagnosis during colonoscopy: a large prospective UK series. Eur J Gastroenterol Hepatol 2011;23:903e11. [67] Hassan C, Pickhardt PJ, Rex DK. A resect and discard strategy would improve cost-effectiveness of colorectal cancer screening. Clin Gastroenterol Hepatol 2010;8:865e9. 869 e1e3. [68] Kessler WR, Imperiale TF, Klein RW, et al. A quantitative assessment of the risks and cost savings of forgoing histologic examination of diminutive polyps. Endoscopy 2011;43:683e91. [69] Hirata M, Tanaka S, Oka S, et al. Evaluation of microvessels in colorectal tumors by narrow band imaging magnification. Gastrointest Endosc 2007;66:945e52. [70] Teixeira CR, Torresini RS, Canali C, et al. Endoscopic classification of the capillary-vessel pattern of colorectal lesions by spectral estimation technology and magnifying zoom imaging. Gastrointest Endosc 2009;69:750e6. [71] Basford PJ, Longcroft-Wheaton G, Higgins B, et al. High-definition endoscopy with i-Scan for evaluation of small colon polyps: the HiSCOPE study. Gastrointest Endosc 2014;79:111e8. [72] Rastogi A. Optical diagnosis of small colorectal polyp histology with high-definition colonoscopy using narrow band imaging. Clin Endosc 2013;46:120e9. [73] Raghavendra M, Hewett DG, Rex DK. Differentiating adenomas from hyperplastic colorectal polyps: narrow-band imaging can be learned in 20 minutes. Gastrointest Endosc 2010;72:572e6. [74] Ignjatovic A, Thomas-Gibson S, East JE, et al. 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;73:128e33. [75] Rastogi A, Rao DS, Gupta N, et al. Impact of a computer-based teaching module on characterization of diminutive colon polyps by using narrow-band imaging by non-experts in academic and community practice: a video-based study. Gastrointest Endosc 2013. [76] Hogan 3rd RB, Brill JV, Littenberg G, et al. Predict, resect, and discard… really? Gastrointest Endosc 2012;75:503e5. [77] Gellad ZF, Voils CI, Lin L, et al. Clinical practice variation in the management of diminutive colorectal polyps: results of a national survey of gastroenterologists. Am J Gastroenterol 2013;108:873e8.