- Email: [email protected]
Identification of volumetric laser endomicroscopy features of colon polyps with histologic correlation
Accepted Manuscript Identification of volumetric laser endomicroscopy features of colon polyps with histologic correlation Arvind J. Trindade, Arvind Rishi, Robert Hirten, Sumant Inamdar, Divyesh V. Sejpal, Jean-Frederic Colombel PII:
S0016-5107(18)30138-X
DOI:
10.1016/j.gie.2018.02.024
Reference:
YMGE 10956
To appear in:
Gastrointestinal Endoscopy
Received Date: 19 November 2017 Accepted Date: 15 February 2018
Please cite this article as: Trindade AJ, Rishi A, Hirten R, Inamdar S, Sejpal DV, Colombel J-F, Identification of volumetric laser endomicroscopy features of colon polyps with histologic correlation, Gastrointestinal Endoscopy (2018), doi: 10.1016/j.gie.2018.02.024. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Identification of volumetric laser endomicroscopy features of colon polyps with histologic correlation Arvind J Trindade* (1), Arvind Rishi* (2), Robert Hirten (3), Sumant Inamdar (1), Divyesh V Sejpal (1), Jean-Frederic Colombel (3)
RI PT
(1) Hofstra Northwell School of Medicine, Northwell Health System, Division of Gastroenterology, Department of Medicine, Long Island Jewish Medical Center, New Hyde Park, NY, USA
SC
(2) Hofstra Northwell School of Medicine, Northwell Health System, Department of Pathology, Long Island Jewish Medical Center, New Hyde Park, NY, USA
*These authors contributed equally GRANT SUPPORT: None
M AN U
(3) Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, NY
ADDRESS CORRESPONDENCE during the review and after publication:
EP
TE D
Arvind J. Trindade, MD Director of Endoscopy Hofstra Northwell School of Medicine Long Island Jewish Medical Center Northwell Health System Division of Gastroenterology 270-05 76th Avenue, New Hyde Park, NY 11040. Tel: (718) 470-7281; Fax: (718) 470-5509 e-mail: [email protected]
AC C
CONFLICT OF INTEREST DISCLOSURE: None AUTHORS CONTRIBUTIONS: Conception and design (AJT) Analysis and interpretation of the data (AJT,AR) Drafting of the article (AJT,AR) Critical revision of the article for important intellectual content (AJT, AR, RH, SI, DVS, JFC) Final approval of the article (AJT, AR, RH, SI, DVS, JFC)
ACCEPTED MANUSCRIPT
ABSTRACT Background and Aims: There is limited data on the use of volumetric laser endomicroscopy (VLE) in imaging for colon polyps. Our aim was to identify VLE features of colon polyps.
RI PT
Methods: A total of 45 patients were included. 43 underwent endoscopic mucosal resection of colorectal polyps 2 cm or greater. These polyps were then scanned with VLE immediately after resection. Two patients that underwent partial colonic resection served as controls.
M AN U
SC
Results: Forty-three polyps were included with review of matching histology: 3 intramucosal adenocarcinoma (IMCA), 5 tubular adenoma (TA)/tubulovillous adenoma (TVA) with high-grade dysplasia (HGD), 9 TVA with only low-grade dysplasia (LGD), 5 serrated adenoma (SA), and 21 TA with LGD. All TA and TVA were hyper-reflective compared with normal tissue. Effacement occurred in 82.4% (14/17) of the colonic polyps with advanced pathology (TVA with HGD/IMCA) compared with 11.6% (3/26) of non-advanced pathology (TA with LGD and SA) (P<0.0001). Forty-seven percent (8/17) of advanced pathology had greater than 5 glands on VLE as compared with none in the non-advanced pathology group (P=0.0001). An irregular surface mainly occurred in polyps with high-grade pathology (HGD/IMCA) versus tubular adenomas. Eight-eight percent of polyps with HGD/IMC had an irregular surface (7/8) versus 6% (2/35) of TA (P<0.0001).
EP
INTRODUCTION
TE D
Conclusions: In this ex-vivo clinicopathological study, we show that there are distinct VLE features of colon polyps that may help identify polyps or features of a higher-grade lesion. This may have implications for possible in-vivo application to aid in dysplasia or polyp detection.
Colorectal cancer (CRC) is the second leading cause of cancer-related deaths in the
AC C
United States 1. Screening for CRC decreases CRC related mortality and thus is recommended per the United States Preventative Services Task Force as well as gastroenterology society guidelines2,3. One of the most common forms of screening is colonoscopy. There is evidence showing that removal of polyps reduces the mortality from CRC 4. Despite the benefit of colonoscopy, there is the potential for colonoscopy to miss polyps for various reasons in up to 6% of colonoscopies5. Therefore, it may be beneficial to use advanced imaging to help locate polyps,
ACCEPTED MANUSCRIPT
especially in high-risk groups such as inflammatory bowel disease, genetic syndromes predisposing to CRC, patients with obesity, diabetes, or smokers 1.
Optical coherence tomography (OCT), an advanced imaging platform that uses
RI PT
infrared light to produce high-resolution cross-sectional microstructure imaging,
has been studied in the colon 6–8. These studies have shown that OCT can be used to image the colon and features of normal versus adenomatous tissue have been
defined based on light scattering and tissue organization. However, this technology
SC
was not readily available or commercially developed. Recently a second-generation OCT, volumetric laser endomicroscopy (nVisionVLE, NinePoint Medical Inc, Bedford,
M AN U
Mass, USA), has been commercially developed. Although VLE is FDA cleared for use in the gastrointestinal tract, the main focus of application has been in targeting dysplasia in Barrett’s esophagus 9,10. Only a single case report exists in the literature on the use of VLE in imaging a colon polyp11. The VLE imaging system consists of a console, monitor, and optical probe contained within a Mylar balloon on a 8F, 260 cm catheter (FIGURE 1). The distal end of the catheter connects to the console. The
TE D
optical probe is designed to fit in a standard adult gastroscope channel. The probe is commercially available in 14 mm, 17 mm, and 20 mm diameter balloons that are 6 cm in length. The inflated balloon allows for centering of the probe while helical scanning occurs. VLE can scan a 6-cm length in approximately 90 seconds, providing
EP
surface and subsurface wide-field cross-sectional imaging with an axial resolution of 7 um, and to a depth of 3 mm 9,10. The rapid imaging of a long segment of tissue makes
AC C
this technology ideal for imaging long segments of the colon.
Before this technology can be accessed in-vivo, it is important to understand the VLE characteristics of the normal colon versus the adenomatous, tubulovillous, and neoplastic colon polyps. Thus, the purpose of this study was to investigate the VLE ex-vivo characteristics of adenomatous, tubulovillous, and neoplastic colon polyps and compare them with the histologic characteristics. This was accomplished by performing a VLE scan on ex-vivo endoscopic mucosal resection (EMR) specimens and then comparing these ex-vivo VLE findings with the histology.
ACCEPTED MANUSCRIPT
METHODS
RI PT
Patients and EMR specimens
The Institutional review board approved this study. EMR specimens were obtained in patients who underwent an EMR at a tertiary care referral center by 2 advanced endoscopists (AJT and DVS) at Long Island Jewish Medical Center and North Shore
SC
University Hospital, Northwell Health System, Long Island, NY. EMR specimens had to be sessile (ls), flat elevated lesions (0-lla, 0-lla +c) or flat lesions (0-llb, 0-llc) per
M AN U
the Paris classification12 and at least 2 cm. The 2 cm size cut-off was chosen to ensure a high-quality scan could be achieved to interpret VLE features in this pilot study. Endoscopic mucosal resection was performed by a saline-methylene blue injection of the polyp into the submucosa and then the polyp was removed with a
VLE scanning
TE D
snare. Every attempt was made to remove the polyps in as few pieces as possible.
Immediately after resection the polyp pieces was positioned flat on a piece of gauze and was scanned with a probe within a 20 mm balloon using the nVision VLE system
EP
(Ninepoint Medical, Bedford, Mass, USA). The system consists of a probe within a balloon, a console, and a monitor. The entire EMR specimen was scanned with the
AC C
probe facing the mucosal side of the polyp. VLE scans to a depth of 3 mm and thus visualized the submucosa of the ex-vivo polyp specimens. Imaging was performed by automatic helical pullback of the probe from the distal to proximal end of the balloon over a 90-second period, creating real-time volumetric images. VLE scans are viewed by using a software interface that allows real-time viewing of crosssectional transverse and longitudinal views of the polyps.
VLE scans were then reviewed by 2 reviewers (A.J.T., S.I.) experienced in reading VLE (>50 scans). The following features were abstracted for each slide: reflectivity
ACCEPTED MANUSCRIPT
(normal vs hyper-reflective), effacement or loss of layered VLE architecture, irregularity of the surface, and degree of glandular structures (no glands, 1-5 glands, and >5 glands). These features were selected as features of importance based on prior literature6–8 and prior experience of the reviewers. Reflectivity refers to the
RI PT
comparison of grayscale of the epithelial layer with the subsurface. Hyper-reflective term is applied to the darker grayscale. In normal colon, the surface and subsurface have the same reflectivity (FIGURE 2). Effacement refers to the loss of normal layers. Normal rectum and colon show distinct layers. If all these layers are obscured this
SC
is termed effacement. For the purpose of this study, if the muscularis mucosae was intact, then effacement was considered absent; as this layer is very prominent when
M AN U
present versus the other layers. Irregularity of the surface is defined as lack of a smooth surface as seen in normal rectum and colon. In order for a surface to be irregular, over 50% of the polyp surface had to be irregular on VLE. The two reviewers then reviewed the scans together and any disagreements were discussed and a consensus was reached. VLE findings of the various polyp types were then
TE D
recorded based on the abstracted variables.
In order to have a reference of normal tissue, two partial colectomy specimens were scanned immediately after resection in the operating room. Both specimens were from the patients with unresectable polyps on endoscopy. Normal tissue and the
AC C
scanned.
EP
polyps were identified on each gross specimen. Only the normal areas were
Histology
Immediately after VLE scanning, the polyps were placed in 10% neutral buffered formalin. The polyps were then processed for histological examination. Histology slides were stained with hematoxylin and eosin (H&E) for histological analysis and evaluated by an expert gastroenterology pathologist (A.R.) who was familiar with the study protocol. After the histologic review, VLE images and histologic images were reviewed between an expert in VLE (A.J.T.) and the pathologist (A.R.) to correlate VLE findings
ACCEPTED MANUSCRIPT
to the histologic findings. Reasons for VLE findings based on histology were then recorded.
STATISTICS
RI PT
The categorical variables were compared using χ2 tests or Fisher exact tests. All
statistical tests were 2-sided, and P < 0.05 was considered significant. All analyses were conducted using SAS Version 9.4 (SAS Institute Inc, Cary, NC, USA).
SC
RESULTS
M AN U
A total of 45 specimens underwent VLE scanning. There were 43 ex-vivo colon polyps scanned immediately after endoscopic resection and 2 partial colectomy specimens scanned immediately after colectomy. Of the 43 polyps resected, the final histologic diagnosis was as follows: intramucosal adenocarcinoma (3), adenoma/tubulovillous adenoma with high-grade dysplasia (5), tubulovillous adenoma with only low-grade dysplasia (9), sessile serrated adenoma (5), tubular adenoma with only low-grade
TE D
dysplasia (21).
The VLE features of the controls/normal colon (FIGURE 2) were as follows. Layered colon architecture was seen. The surface was regular and no glands were seen. The
EP
surface had a normal reflectance pattern.
AC C
The VLE features for each colon polyp histology subtype are in TABLE1 and illustrated in FIGURES 3 (polyps with advanced pathology) and 4 (polyps with non-advanced pathology) . All adenomas and tubulovillous adenomas regardless of the degree of dysplasia showed hyper-reflectivity compared with the normal tissue. Serrated adenomas had the same reflectance pattern as the normal tissue. Effacement of the layers was mainly observed in advanced pathology, which consisted of tubulovillous adenomas, polyps with high-grade dysplasia (HGD) and intramucosal cancer (IMCA). Effacement was not frequently observed in non-advanced pathology, which consisted of tubular adenomas with low-grade dysplasia (LGD) and sessile serrated adenomas. Effacement
ACCEPTED MANUSCRIPT
occurred in 82.4% (14/17) of advanced pathology compared with 11.6% (3/26) of nonadvanced pathology (P<0.0001). Glands were present in 94% (16/17) of advanced pathology versus 42% (11/26) of non-advanced pathology (P=0.007). The number of glands present was a predictor of advanced pathology. Forty-seven percent (8/17) of
RI PT
advanced pathology has greater than 5 glands present versus zero in the non-advanced
pathology group (P=0.0001). An irregular surface mainly occurred in polyps with higher grade (HGD/IMCA) versus polyps with low-grade dysplasia (adenoma and tubulovillous adenomas with LGD and serrated adenomas). Eight-eight percent of polyps with
SC
advanced pathology had an irregular surface (7/8) versus 6% (2/35) of polyps with low-
M AN U
grade dysplasia (P<0.0001).
Histologic Correlation to VLE Findings
Histological pattern of the various polyp subtypes seen on lower magnification were comparable with the VLE findings. In general, tubular adenoma with only low-grade dysplasia (TA) had more uniform colonic crypt architecture and smoother surface
TE D
epithelial lining. The muscularis mucosae was uniform in most of the cases except for some cases, where there was associated mild mucosal prolapse due to the traction or mechanical effect of the polyp. Although, cytologically the lining epithelium of the colonic crypts in TA is dysplastic, the contour of the crypts has much uniform overall
EP
shape. This finding correlated with lesser number of “glands” identified on VLE in TA. On the contrary, tubular adenoma with high-grade dysplasia (HGD) have more frequent
AC C
fused crypts with cribriform architecture, which is more likely to give an appearance of greater number “glands” on VLE (fig 3 D). Histological review of advanced lesion, in particular, intramucosal adenocarcinoma, showed highly complex colonic crypts with high-grade cytological features and disruption of the basement membrane with infiltration of the neoplastic glands into the lamina propria (fig 3 H and 3L).
Infiltration of the lamina propria in IMCA cases could be the reason for irregular surface layer on VLE examination (3H). Similarly, a distorted colonic architecture in HGD could be attributed to the irregular surface findings on VLE. The muscularis mucosae had a
ACCEPTED MANUSCRIPT
higher degree of distortion in the tubulovillous adenoma and IMCA. This distortion is attributed to higher degree of mucosal traction in tubulovillous adenoma, likely due to their larger size and villous architecture of the lesion, leading to the microscopic areas of mucosal prolapse. The muscularis mucosae is not infiltrated by the neoplastic process but
RI PT
is distorted due to the physical architecture of the villous lesion. This could explain the
effacement seen on the VLE image for these lesions. In contrast, this structural distortion does not occur frequently in tubular adenomas due to the flatter surface of the lesion and hence no loss of layering is observed on VLE images. Some cases of TA may encounter
SC
effacement (10% in our series) because of the larger size of the lesion or a certain
location, where the size of lesion in conjunction with the bowel movements may lead to
M AN U
mucosal prolapse. Similarly, such obliteration in IMCA is likely due to the mechanics of the lesion where the lesional cells are infiltrating into the lamina propria. It is important to note that none of these cases had infiltration of dysplastic cells into the muscularis mucosae. Sessile serrated adenoma (SSA) showed uniform mild to moderately distended colonic crypts with mild crypt distortion and non-distended crypts (Fig 4H). The histological and architectural pattern of dysplasia in SSA showed a uniform involvement
TE D
of the mucosal layer. We did not numerically compare the number of colonic crypts on histology with the number of “glands” identified because of the differences in magnification of the tissue seen by VLE and with histology compound microscope.
EP
DISCUSSION
AC C
In this study we identify ex-vivo VLE features of various colon polyps that underwent endoscopic mucosal resection and correlated it to low magnification architectural features of polyps on histology. This is the first study to identify OCT or VLE features that could help identify the type and possibly the grade of dysplasia seen in real-time at the time of colonoscopy.
Previous studies on first generation OCT in the colon helped show that polyps have abnormal light scattering (referred to reflectance in this study to be consistent with the VLE literature) and tissue organization. However, these studies did not differentiate
ACCEPTED MANUSCRIPT
types of polyps or grade of dysplasia. In our study we show that similar to previous OCT studies that tubular adenomas and tubulovillous adenomas have a high reflectance or dark surface compared with the subsurface6–8. New information added by our study shows that serrated adenomas without any associated high-grade dysplasia could have a normal
RI PT
reflectance pattern on VLE.
Also shown in this study is that advanced colon polyp pathology (tubulovillous adenomas and polyps with HGD/ IMCA) has a VLE signature distinct from non-advanced colon
SC
pathology (tubular adenomas with LGD and serrated adenomas). We show that the
advanced pathology polyps have a greater chance of having effacement/loss of layer VLE
M AN U
architecture and >5 glands. If these features are seen then there is a higher statistical chance of these polyps being of advanced pathology. Additionally, neoplastic polyps (HGD/IMCA) had a higher chance of having an irregular surface. Based on these patterns described, we have developed a possible algorithm to differentiate the different types of polyps using the VLE features (FIGURE 4). This will need to be validated prospectively. Not surprisingly the abnormal features seen in this study are similar to abnormal VLE
TE D
features seen in Barrett’s esophagus with dysplasia and intramucosal cancer13,14. Both of these neoplastic processes are epithelial based, which occur in columnar mucosal lining.
By defining the VLE criteria for various polyps and associated dysplasia grade in this
EP
study, we may be able to apply VLE to in-vivo use for a number of clinical scenarios. We especially think that VLE could play a role in IBD surveillance given that dysplasia can be difficult to find. The presence of dysplasia increases the risk of cancer significantly in
AC C
these patients18. VLE can potentially help differentiate pseudopolyps from dysplastic polyps and also identify normal versus dysplastic tissue in the setting of quiescent IBD. Given this potential promise, our group is commencing an ex-vivo study of VLE scanning on colectomy specimens in IBD performed for dysplasia and correlation to histology. In addition, this technology may be useful for surveillance of polypectomy sites after EMR of large, flat colon polyps. When residual type tissue is seen in the area of the polypectomy site, it is either inflammatory/hyperplastic tissue vs. residual dysplastic tissue. VLE can potentially help differentiate the residual tissue; dysplastic
ACCEPTED MANUSCRIPT
tissue would need further therapeutic treatment such as additional resection, argon plasma coagulation ablation or avulsion. VLE may also play a role in high-risk patients for colon cancer, ie, Lynch syndrome and personal or family history of colon cancer. Despite advanced optics of colonoscopes a significant amount of colon polyps can still be
RI PT
missed in these patients, up to 26% per one systematic review15 with interval cancers
occurring approximately 2% to 6% of the time16,17 With the advent of laser marking with VLE, superficial cautery marks can be placed on areas of suspected dysplasia for
SC
sampling or removal.
This ex-vivo study should provide a stimulus for further research into the potential use of
M AN U
VLE for imaging the colon. Next steps after ex-vivo identification of VLE features are to prospectively validate our findings and possibly scan colectomy specimens with smaller polyps or focal dysplasia to determine if VLE can identify focal disease. VLE in this study scanned larger polyps to identify VLE features of polyps, but clinically VLE would be useful for in-vivo detection of focal dysplasia or lesions that can be missed in routine colonoscopy. Finally, future studies would also be helpful in determining VLE features
invasive cancer.
TE D
that can distinguish adenomas from hyperplastic polyps and intramucosal cancer from
Modifications on the current probe would need to be made before in-vivo use of VLE in
EP
the colon. The current commercially available VLE probe is designed to fit through a standard upper endoscope. Although the cecum can be reached often with a
AC C
gastroscope20,21, creating a longer catheter so the probe can be compatible with the longer colonoscopes would be desirable. Another issue would be that the balloon containing the probe would need to be bigger as currently the largest balloon size is 20 mm, and this would not appose well to the colon epithelial lining. Fortunately, these hardware adjustments might not be difficult to perform. Rather, a more challenging issue would be of imaging around the flexures of the colon or a tortuous sigmoid. The long 6-cm balloon would make this challenging. A shorter balloon would be more ideal for these areas of the colon but would add more time to image other areas of the colon.
ACCEPTED MANUSCRIPT
In conclusion, we identify ex-vivo VLE features of colonic polyps in this study. We show that there are distinct VLE features of colon polyps, which correlate with the structural pattern of colonic polyps on histology at a lower magnification. These VLE features may help in identifying and classifying the colonic polypoid lesions into a low- or high-grade
RI PT
pattern. Our findings have implications for possible application of in-vivo use of this
technology to aid in dysplasia or polyp detection. Future ex-vivo and in-vivo studies as
SC
outlined in the paper, using the criteria outlined here, are warranted.
1.
M AN U
REFERENCES
Oluyemi AO, Welch AR, Yoo LJ, Lehman EB, McGarrity TJ, Chuang CH. Colorectal cancer screening in high-risk groups is increasing, although current smokers fall behind. Cancer [Internet]. 2014 Jul 15;120:2106–13. Available from: http://doi.wiley.com/10.1002/cncr.28707
Rex DK, Boland CR, Dominitz JA, Giardiello FM, Johnson DA, Kaltenbach T, et
TE D
2.
al. Colorectal Cancer Screening: Recommendations for Physicians and Patients From the U.S. Multi-Society Task Force on Colorectal Cancer. Gastroenterology [Internet]. 2017 Jul;153:307–23. Available from:
3.
EP
http://www.ncbi.nlm.nih.gov/pubmed/28600072 US Preventive Services Task Force, Bibbins-Domingo K, Grossman DC, Curry
AC C
SJ, Davidson KW, Epling JW, et al. Screening for Colorectal Cancer: US Preventive Services Task Force Recommendation Statement. JAMA [Internet]. 2016;315:2564–75. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27304597
4.
Lieberman D, Ladabaum U, Cruz-Correa M, Ginsburg C, Inadomi JM, Kim LS, et al. Screening for Colorectal Cancer and Evolving Issues for Physicians and Patients. JAMA [Internet]. 2016;316:2135. Available from: http://jama.jamanetwork.com/article.aspx?doi=10.1001/jama.2016.17418
5.
Lee J, Park SW, Kim YS, Lee KJ, Sung H, Song PH, et al. Risk factors of missed
ACCEPTED MANUSCRIPT
colorectal lesions after colonoscopy. Medicine (Baltimore) [Internet]. 2017;96:e7468. Available from: http://www.ncbi.nlm.nih.gov/pubmed/28682916 6.
Sivak M V, Kobayashi K, Izatt JA, Rollins AM, Ung-Runyawee R, Chak A, et al.
RI PT
High-resolution endoscopic imaging of the GI tract using optical coherence tomography. Gastrointest Endosc [Internet]. 2000 Apr;51(4 Pt 1):474–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10744825 7.
Pfau PR, Sivak M V, Chak A, Kinnard M, Wong RCK, Isenberg GA, et al. Criteria
SC
for the diagnosis of dysplasia by endoscopic optical coherence tomography. Gastrointest Endosc [Internet]. 2003;58:196–202. Available from:
M AN U
http://www.ncbi.nlm.nih.gov/pubmed/12872085 8.
Westphal V, Rollins AM, Willis J, Sivak M V, Izatt JA. Correlation of endoscopic optical coherence tomography with histology in the lower-GI tract. Gastrointest Endosc [Internet]. 2005;61:537–46. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15812406
9.
Trindade AJ, Smith MS, Pleskow DK. The new kid on the block for advanced
TE D
imaging in Barrett’s esophagus: a review of volumetric laser endomicroscopy. Therap Adv Gastroenterol [Internet]. 2016;9:408–16. Available from: http://tag.sagepub.com/cgi/doi/10.1177/1756283X16639003 10.
Trindade AJ, Leggett CL, Chang KJ. Volumetric laser endomicroscopy in the
EP
management of Barrett’s esophagus. Curr Opin Gastroenterol [Internet]. 2017 Apr 11; Available from: http://www.ncbi.nlm.nih.gov/pubmed/28402993 Trindade AJ, Sultan K, Vamadevan AS, Fan C, Sejpal D V. Successful use of
AC C
11.
volumetric laser endomicroscopy in imaging a rectal polyp. Therap Adv Gastroenterol [Internet]. 2016;9:128–31. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26770274
12.
The Paris endoscopic classification of superficial neoplastic lesions:
esophagus, stomach, and colon: November 30 to December 1, 2002. Gastrointest Endosc [Internet]. 2003;58(6 Suppl):S3-43. Available from: http://www.ncbi.nlm.nih.gov/pubmed/14652541 13.
Swager A-F, Tearney GJ, Leggett CL, van Oijen MGH, Meijer SL, Weusten BL, et
ACCEPTED MANUSCRIPT
al. Identification of volumetric laser endomicroscopy features predictive for early neoplasia in Barrett’s esophagus using high-quality histological correlation. Gastrointest Endosc [Internet]. 2016 Sep 19; Available from: http://dx.doi.org/10.1016/j.gie.2016.09.012 Leggett CL, Gorospe EC, Chan DK, Muppa P, Owens V, Smyrk TC, et al.
RI PT
14.
Comparative diagnostic performance of volumetric laser endomicroscopy and confocal laser endomicroscopy in the detection of dysplasia associated with Barrett’s esophagus. Gastrointest Endosc [Internet]. 2015; Available from:
15.
SC
http://linkinghub.elsevier.com/retrieve/pii/S0016510715028229
van Rijn JC, Reitsma JB, Stoker J, Bossuyt PM, van Deventer SJ, Dekker E. Polyp
M AN U
miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol [Internet]. 2006;101:343–50. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16454841 16.
Brenner H, Chang-Claude J, Seiler CM, Hoffmeister M. Interval cancers after negative colonoscopy: population-based case-control study. Gut [Internet]. 2012;61:1576–82. Available from:
17.
TE D
http://www.ncbi.nlm.nih.gov/pubmed/22200840 Bressler B, Paszat LF, Chen Z, Rothwell DM, Vinden C, Rabeneck L. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology [Internet]. 2007;132:96–102.
18.
EP
Available from: http://www.ncbi.nlm.nih.gov/pubmed/17241863 Ullman T, Croog V, Harpaz N, Sachar D, Itzkowitz S. Progression of flat low-
AC C
grade dysplasia to advanced neoplasia in patients with ulcerative colitis. Gastroenterology [Internet]. 2003;125:1311–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/14598247
19.
Beintaris I, Rutter M. Advanced imaging in colonoscopy: contemporary
approach to dysplasia surveillance in inflammatory bowel disease. Frontline Gastroenterol [Internet]. 2016;7:308–15. Available from: http://www.ncbi.nlm.nih.gov/pubmed/28839872
20.
Bick BL, Vemulapalli KC, Rex DK. Regional center for complex colonoscopy: yield of neoplasia in patients with prior incomplete colonoscopy. Gastrointest
ACCEPTED MANUSCRIPT
Endosc [Internet]. 2016;83:1239–44. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26584785 21.
Rex DK, Chen SC, Overhiser AJ. Colonoscopy Technique in Consecutive
Hepatol [Internet]. 2007;5:879–83. Available from:
RI PT
Patients Referred for Prior Incomplete Colonoscopy. Clin Gastroenterol
SC
http://linkinghub.elsevier.com/retrieve/pii/S1542356507002881
FIGURE 1: Components of the volumetric laser endomicroscopy system. A, Console
into a gastroscope.
M AN U
in which the distal end of the catheter attaches. B, Balloon catheter inserted
FIGURE 2: Volumetric laser endomicroscopy features in the colon. A, Normal colon. There is normal reflectance (orange arrow). The red arrow is the muscularis
TE D
mucosa. Above this layer is the epithelium. Below it is the submucosal layer. B, VLE of a tubular adenoma. The epithelium is darker and thus hyperreflective compared with normal colon (orange arrow). Layering is present (defined as presence of muscularis mucosa; red arrow). C, VLE of a
EP
tubulovillous adenoma. There is a darker hyperreflective epithelium (orange arrow). There is lack of layering or effacement given no muscularis mucosae is
AC C
visualized. D, A tubular adenoma with high-grade dysplasia. There is a hyperreflective surface (orange arrow), atypical glands (short yellow arrows),
and effacement. E) A polyp with intramucosal cancer. There is a hyperreflective surface (orange arrow), an irregular surface (blue arrow), and
effacement.
FIGURE 3: Representative colon polyps with high risk volumetric laser endomicroscopy and matched histologic features. A, A tubulovillous adenoma with high-grade dysplasia on colonoscopy. B, post-endoscopic mucosal
ACCEPTED MANUSCRIPT
resection of the polyp seen in A C, VLE features of the polyp in A showing an irregular surface, effacement, and greater than 5 glands. D, Matched histology of the polyp in A showing distended and irregular crypts in scattered areas and fused crypts with cribriform architecture (HE; x 40 magnification). E, A
RI PT
polyp with intramucosal cancer on colonoscopy. F, post EMR of the polyp seen in E. G, VLE features of the polyp in E showing an irregular surface and
effacement. H, Matched histology with areas of highly irregular dysplastic
colonic crypts and associated lamina propria stromal reaction. Note that some
SC
fused crypts are identified. However, distended glands are not identified in
region with low magnification pattern of intramucosal adenocarcinoma (H &
M AN U
E; orig. mag. x40). I, A tubular adenonma with high-grade dysplasia and a focus of intramucosal cancer seen on colonoscopy. J, post-EMR of the polyp seen in I. K, VLE features of the polyp in I showing an irregular surface, effacement, and greater than 5 glands. L, Matched histology with fused complex colonic crypts lined by high-grade epithelium (H & E; orig. mag.
TE D
x100).
FIGURE 4: Representative colon polyps with low risk volumetric laser endomicroscopy and matched histologic features. A, A adenoma with lowgrade dysplasia on colonoscopy. B, post-endoscopic mucosal resection of the
EP
polyp seen in A. C, VLE features of the polyp in A showing regular surface contour, fewer than 5 glands, and no effacement. D, Matched histology of the
AC C
polyp in A showing uniform colonic crypts and smooth surface contour (H & E; orig. mag. x40). E, A serrated adenoma on colonoscopym F, post-endoscopic mucosal resection of the polyp seen in E. G, VLE features of the polyp in E showing normal reflectance pattern and no effacement. H, Matched histology of the polyp in E showing uniform serrated crypts and smooth surface contour. Also note that serrated lesions may have prominent benign adipose tissue in the submucosa (arrow) (H & E; orig. mag. x40).
FIGURE 5: Proposed algorithm for identification of the type of polyp based on
ACCEPTED MANUSCRIPT
volumetric laser endomicroscopy features.
TABLE 1: Volumetric laser endomicroscopy features of the colon polyps.
Glands 1-5
Glands >5
-
-
-
-
0% (0/21)
48% (10/21 )
M AN U
CONTROL n=2 Normal n=2 NON-ADVANCED PATHOLOGY n=26 Tubular 100 % 10% Adenoma (TA) (21/21) (2/21) with LGD n=21 Sessile Serrated 0% 20% Adenoma (0/5) (1/5) n=5 ADVANCED PATHOLOGY n=17 Tubulovillous 100% 67% adenoma (TVA) (9/9) (6/9) with LGD n=9 TA or TVA with 100% 100% HGD n=5 (5/5) (5/5)
Irregula No r Surface Glands
RI PT
Effacemen t of all layers
SC
HyperReflectivit y
AC C
100% (8/8)
0% 0/21
0% (0/5)
100% (5/5)
0% (0/5)
0% (0/5)
22% (2/9)
11% (1/9)
44% (4/9)
44% (4/9)
80% (4/5)
0% (0/5)
40% (2/5)
60% (3/5)
100% (3/3)
100% (3/3)
0% (0/3)
67% (2/3) (67%)
33% (1/3)
100% (8/8)
88% (7/8)
0% (0/8)
50% (4/8)
50% (4/8)
TE D
100% (3/3)
EP
Intramucosal Adenocarcinoma (IMCA) n=3 Composite IMCA/HGD N=8
52% (11/21 )
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
SC
M AN U
VLE= Volumetric Laser Endomicroscopy EMR= Endoscopic Mucosal Resection OCT= Optical Coherence Tomography HGD= high-grade dysplasia IMCA= intramucosal cancer TA= tubular adenoma TVA= tubulovillous adenoma LGD= low grade dysplasia SSA= sessile serrated adenoma CRC= colorectal cancer H&E= with hematoxylin and eosin
RI PT
Acronyms
S0016-5107(18)30138-X
DOI:
10.1016/j.gie.2018.02.024
Reference:
YMGE 10956
To appear in:
Gastrointestinal Endoscopy
Received Date: 19 November 2017 Accepted Date: 15 February 2018
Please cite this article as: Trindade AJ, Rishi A, Hirten R, Inamdar S, Sejpal DV, Colombel J-F, Identification of volumetric laser endomicroscopy features of colon polyps with histologic correlation, Gastrointestinal Endoscopy (2018), doi: 10.1016/j.gie.2018.02.024. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Identification of volumetric laser endomicroscopy features of colon polyps with histologic correlation Arvind J Trindade* (1), Arvind Rishi* (2), Robert Hirten (3), Sumant Inamdar (1), Divyesh V Sejpal (1), Jean-Frederic Colombel (3)
RI PT
(1) Hofstra Northwell School of Medicine, Northwell Health System, Division of Gastroenterology, Department of Medicine, Long Island Jewish Medical Center, New Hyde Park, NY, USA
SC
(2) Hofstra Northwell School of Medicine, Northwell Health System, Department of Pathology, Long Island Jewish Medical Center, New Hyde Park, NY, USA
*These authors contributed equally GRANT SUPPORT: None
M AN U
(3) Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, NY
ADDRESS CORRESPONDENCE during the review and after publication:
EP
TE D
Arvind J. Trindade, MD Director of Endoscopy Hofstra Northwell School of Medicine Long Island Jewish Medical Center Northwell Health System Division of Gastroenterology 270-05 76th Avenue, New Hyde Park, NY 11040. Tel: (718) 470-7281; Fax: (718) 470-5509 e-mail: [email protected]
AC C
CONFLICT OF INTEREST DISCLOSURE: None AUTHORS CONTRIBUTIONS: Conception and design (AJT) Analysis and interpretation of the data (AJT,AR) Drafting of the article (AJT,AR) Critical revision of the article for important intellectual content (AJT, AR, RH, SI, DVS, JFC) Final approval of the article (AJT, AR, RH, SI, DVS, JFC)
ACCEPTED MANUSCRIPT
ABSTRACT Background and Aims: There is limited data on the use of volumetric laser endomicroscopy (VLE) in imaging for colon polyps. Our aim was to identify VLE features of colon polyps.
RI PT
Methods: A total of 45 patients were included. 43 underwent endoscopic mucosal resection of colorectal polyps 2 cm or greater. These polyps were then scanned with VLE immediately after resection. Two patients that underwent partial colonic resection served as controls.
M AN U
SC
Results: Forty-three polyps were included with review of matching histology: 3 intramucosal adenocarcinoma (IMCA), 5 tubular adenoma (TA)/tubulovillous adenoma (TVA) with high-grade dysplasia (HGD), 9 TVA with only low-grade dysplasia (LGD), 5 serrated adenoma (SA), and 21 TA with LGD. All TA and TVA were hyper-reflective compared with normal tissue. Effacement occurred in 82.4% (14/17) of the colonic polyps with advanced pathology (TVA with HGD/IMCA) compared with 11.6% (3/26) of non-advanced pathology (TA with LGD and SA) (P<0.0001). Forty-seven percent (8/17) of advanced pathology had greater than 5 glands on VLE as compared with none in the non-advanced pathology group (P=0.0001). An irregular surface mainly occurred in polyps with high-grade pathology (HGD/IMCA) versus tubular adenomas. Eight-eight percent of polyps with HGD/IMC had an irregular surface (7/8) versus 6% (2/35) of TA (P<0.0001).
EP
INTRODUCTION
TE D
Conclusions: In this ex-vivo clinicopathological study, we show that there are distinct VLE features of colon polyps that may help identify polyps or features of a higher-grade lesion. This may have implications for possible in-vivo application to aid in dysplasia or polyp detection.
Colorectal cancer (CRC) is the second leading cause of cancer-related deaths in the
AC C
United States 1. Screening for CRC decreases CRC related mortality and thus is recommended per the United States Preventative Services Task Force as well as gastroenterology society guidelines2,3. One of the most common forms of screening is colonoscopy. There is evidence showing that removal of polyps reduces the mortality from CRC 4. Despite the benefit of colonoscopy, there is the potential for colonoscopy to miss polyps for various reasons in up to 6% of colonoscopies5. Therefore, it may be beneficial to use advanced imaging to help locate polyps,
ACCEPTED MANUSCRIPT
especially in high-risk groups such as inflammatory bowel disease, genetic syndromes predisposing to CRC, patients with obesity, diabetes, or smokers 1.
Optical coherence tomography (OCT), an advanced imaging platform that uses
RI PT
infrared light to produce high-resolution cross-sectional microstructure imaging,
has been studied in the colon 6–8. These studies have shown that OCT can be used to image the colon and features of normal versus adenomatous tissue have been
defined based on light scattering and tissue organization. However, this technology
SC
was not readily available or commercially developed. Recently a second-generation OCT, volumetric laser endomicroscopy (nVisionVLE, NinePoint Medical Inc, Bedford,
M AN U
Mass, USA), has been commercially developed. Although VLE is FDA cleared for use in the gastrointestinal tract, the main focus of application has been in targeting dysplasia in Barrett’s esophagus 9,10. Only a single case report exists in the literature on the use of VLE in imaging a colon polyp11. The VLE imaging system consists of a console, monitor, and optical probe contained within a Mylar balloon on a 8F, 260 cm catheter (FIGURE 1). The distal end of the catheter connects to the console. The
TE D
optical probe is designed to fit in a standard adult gastroscope channel. The probe is commercially available in 14 mm, 17 mm, and 20 mm diameter balloons that are 6 cm in length. The inflated balloon allows for centering of the probe while helical scanning occurs. VLE can scan a 6-cm length in approximately 90 seconds, providing
EP
surface and subsurface wide-field cross-sectional imaging with an axial resolution of 7 um, and to a depth of 3 mm 9,10. The rapid imaging of a long segment of tissue makes
AC C
this technology ideal for imaging long segments of the colon.
Before this technology can be accessed in-vivo, it is important to understand the VLE characteristics of the normal colon versus the adenomatous, tubulovillous, and neoplastic colon polyps. Thus, the purpose of this study was to investigate the VLE ex-vivo characteristics of adenomatous, tubulovillous, and neoplastic colon polyps and compare them with the histologic characteristics. This was accomplished by performing a VLE scan on ex-vivo endoscopic mucosal resection (EMR) specimens and then comparing these ex-vivo VLE findings with the histology.
ACCEPTED MANUSCRIPT
METHODS
RI PT
Patients and EMR specimens
The Institutional review board approved this study. EMR specimens were obtained in patients who underwent an EMR at a tertiary care referral center by 2 advanced endoscopists (AJT and DVS) at Long Island Jewish Medical Center and North Shore
SC
University Hospital, Northwell Health System, Long Island, NY. EMR specimens had to be sessile (ls), flat elevated lesions (0-lla, 0-lla +c) or flat lesions (0-llb, 0-llc) per
M AN U
the Paris classification12 and at least 2 cm. The 2 cm size cut-off was chosen to ensure a high-quality scan could be achieved to interpret VLE features in this pilot study. Endoscopic mucosal resection was performed by a saline-methylene blue injection of the polyp into the submucosa and then the polyp was removed with a
VLE scanning
TE D
snare. Every attempt was made to remove the polyps in as few pieces as possible.
Immediately after resection the polyp pieces was positioned flat on a piece of gauze and was scanned with a probe within a 20 mm balloon using the nVision VLE system
EP
(Ninepoint Medical, Bedford, Mass, USA). The system consists of a probe within a balloon, a console, and a monitor. The entire EMR specimen was scanned with the
AC C
probe facing the mucosal side of the polyp. VLE scans to a depth of 3 mm and thus visualized the submucosa of the ex-vivo polyp specimens. Imaging was performed by automatic helical pullback of the probe from the distal to proximal end of the balloon over a 90-second period, creating real-time volumetric images. VLE scans are viewed by using a software interface that allows real-time viewing of crosssectional transverse and longitudinal views of the polyps.
VLE scans were then reviewed by 2 reviewers (A.J.T., S.I.) experienced in reading VLE (>50 scans). The following features were abstracted for each slide: reflectivity
ACCEPTED MANUSCRIPT
(normal vs hyper-reflective), effacement or loss of layered VLE architecture, irregularity of the surface, and degree of glandular structures (no glands, 1-5 glands, and >5 glands). These features were selected as features of importance based on prior literature6–8 and prior experience of the reviewers. Reflectivity refers to the
RI PT
comparison of grayscale of the epithelial layer with the subsurface. Hyper-reflective term is applied to the darker grayscale. In normal colon, the surface and subsurface have the same reflectivity (FIGURE 2). Effacement refers to the loss of normal layers. Normal rectum and colon show distinct layers. If all these layers are obscured this
SC
is termed effacement. For the purpose of this study, if the muscularis mucosae was intact, then effacement was considered absent; as this layer is very prominent when
M AN U
present versus the other layers. Irregularity of the surface is defined as lack of a smooth surface as seen in normal rectum and colon. In order for a surface to be irregular, over 50% of the polyp surface had to be irregular on VLE. The two reviewers then reviewed the scans together and any disagreements were discussed and a consensus was reached. VLE findings of the various polyp types were then
TE D
recorded based on the abstracted variables.
In order to have a reference of normal tissue, two partial colectomy specimens were scanned immediately after resection in the operating room. Both specimens were from the patients with unresectable polyps on endoscopy. Normal tissue and the
AC C
scanned.
EP
polyps were identified on each gross specimen. Only the normal areas were
Histology
Immediately after VLE scanning, the polyps were placed in 10% neutral buffered formalin. The polyps were then processed for histological examination. Histology slides were stained with hematoxylin and eosin (H&E) for histological analysis and evaluated by an expert gastroenterology pathologist (A.R.) who was familiar with the study protocol. After the histologic review, VLE images and histologic images were reviewed between an expert in VLE (A.J.T.) and the pathologist (A.R.) to correlate VLE findings
ACCEPTED MANUSCRIPT
to the histologic findings. Reasons for VLE findings based on histology were then recorded.
STATISTICS
RI PT
The categorical variables were compared using χ2 tests or Fisher exact tests. All
statistical tests were 2-sided, and P < 0.05 was considered significant. All analyses were conducted using SAS Version 9.4 (SAS Institute Inc, Cary, NC, USA).
SC
RESULTS
M AN U
A total of 45 specimens underwent VLE scanning. There were 43 ex-vivo colon polyps scanned immediately after endoscopic resection and 2 partial colectomy specimens scanned immediately after colectomy. Of the 43 polyps resected, the final histologic diagnosis was as follows: intramucosal adenocarcinoma (3), adenoma/tubulovillous adenoma with high-grade dysplasia (5), tubulovillous adenoma with only low-grade dysplasia (9), sessile serrated adenoma (5), tubular adenoma with only low-grade
TE D
dysplasia (21).
The VLE features of the controls/normal colon (FIGURE 2) were as follows. Layered colon architecture was seen. The surface was regular and no glands were seen. The
EP
surface had a normal reflectance pattern.
AC C
The VLE features for each colon polyp histology subtype are in TABLE1 and illustrated in FIGURES 3 (polyps with advanced pathology) and 4 (polyps with non-advanced pathology) . All adenomas and tubulovillous adenomas regardless of the degree of dysplasia showed hyper-reflectivity compared with the normal tissue. Serrated adenomas had the same reflectance pattern as the normal tissue. Effacement of the layers was mainly observed in advanced pathology, which consisted of tubulovillous adenomas, polyps with high-grade dysplasia (HGD) and intramucosal cancer (IMCA). Effacement was not frequently observed in non-advanced pathology, which consisted of tubular adenomas with low-grade dysplasia (LGD) and sessile serrated adenomas. Effacement
ACCEPTED MANUSCRIPT
occurred in 82.4% (14/17) of advanced pathology compared with 11.6% (3/26) of nonadvanced pathology (P<0.0001). Glands were present in 94% (16/17) of advanced pathology versus 42% (11/26) of non-advanced pathology (P=0.007). The number of glands present was a predictor of advanced pathology. Forty-seven percent (8/17) of
RI PT
advanced pathology has greater than 5 glands present versus zero in the non-advanced
pathology group (P=0.0001). An irregular surface mainly occurred in polyps with higher grade (HGD/IMCA) versus polyps with low-grade dysplasia (adenoma and tubulovillous adenomas with LGD and serrated adenomas). Eight-eight percent of polyps with
SC
advanced pathology had an irregular surface (7/8) versus 6% (2/35) of polyps with low-
M AN U
grade dysplasia (P<0.0001).
Histologic Correlation to VLE Findings
Histological pattern of the various polyp subtypes seen on lower magnification were comparable with the VLE findings. In general, tubular adenoma with only low-grade dysplasia (TA) had more uniform colonic crypt architecture and smoother surface
TE D
epithelial lining. The muscularis mucosae was uniform in most of the cases except for some cases, where there was associated mild mucosal prolapse due to the traction or mechanical effect of the polyp. Although, cytologically the lining epithelium of the colonic crypts in TA is dysplastic, the contour of the crypts has much uniform overall
EP
shape. This finding correlated with lesser number of “glands” identified on VLE in TA. On the contrary, tubular adenoma with high-grade dysplasia (HGD) have more frequent
AC C
fused crypts with cribriform architecture, which is more likely to give an appearance of greater number “glands” on VLE (fig 3 D). Histological review of advanced lesion, in particular, intramucosal adenocarcinoma, showed highly complex colonic crypts with high-grade cytological features and disruption of the basement membrane with infiltration of the neoplastic glands into the lamina propria (fig 3 H and 3L).
Infiltration of the lamina propria in IMCA cases could be the reason for irregular surface layer on VLE examination (3H). Similarly, a distorted colonic architecture in HGD could be attributed to the irregular surface findings on VLE. The muscularis mucosae had a
ACCEPTED MANUSCRIPT
higher degree of distortion in the tubulovillous adenoma and IMCA. This distortion is attributed to higher degree of mucosal traction in tubulovillous adenoma, likely due to their larger size and villous architecture of the lesion, leading to the microscopic areas of mucosal prolapse. The muscularis mucosae is not infiltrated by the neoplastic process but
RI PT
is distorted due to the physical architecture of the villous lesion. This could explain the
effacement seen on the VLE image for these lesions. In contrast, this structural distortion does not occur frequently in tubular adenomas due to the flatter surface of the lesion and hence no loss of layering is observed on VLE images. Some cases of TA may encounter
SC
effacement (10% in our series) because of the larger size of the lesion or a certain
location, where the size of lesion in conjunction with the bowel movements may lead to
M AN U
mucosal prolapse. Similarly, such obliteration in IMCA is likely due to the mechanics of the lesion where the lesional cells are infiltrating into the lamina propria. It is important to note that none of these cases had infiltration of dysplastic cells into the muscularis mucosae. Sessile serrated adenoma (SSA) showed uniform mild to moderately distended colonic crypts with mild crypt distortion and non-distended crypts (Fig 4H). The histological and architectural pattern of dysplasia in SSA showed a uniform involvement
TE D
of the mucosal layer. We did not numerically compare the number of colonic crypts on histology with the number of “glands” identified because of the differences in magnification of the tissue seen by VLE and with histology compound microscope.
EP
DISCUSSION
AC C
In this study we identify ex-vivo VLE features of various colon polyps that underwent endoscopic mucosal resection and correlated it to low magnification architectural features of polyps on histology. This is the first study to identify OCT or VLE features that could help identify the type and possibly the grade of dysplasia seen in real-time at the time of colonoscopy.
Previous studies on first generation OCT in the colon helped show that polyps have abnormal light scattering (referred to reflectance in this study to be consistent with the VLE literature) and tissue organization. However, these studies did not differentiate
ACCEPTED MANUSCRIPT
types of polyps or grade of dysplasia. In our study we show that similar to previous OCT studies that tubular adenomas and tubulovillous adenomas have a high reflectance or dark surface compared with the subsurface6–8. New information added by our study shows that serrated adenomas without any associated high-grade dysplasia could have a normal
RI PT
reflectance pattern on VLE.
Also shown in this study is that advanced colon polyp pathology (tubulovillous adenomas and polyps with HGD/ IMCA) has a VLE signature distinct from non-advanced colon
SC
pathology (tubular adenomas with LGD and serrated adenomas). We show that the
advanced pathology polyps have a greater chance of having effacement/loss of layer VLE
M AN U
architecture and >5 glands. If these features are seen then there is a higher statistical chance of these polyps being of advanced pathology. Additionally, neoplastic polyps (HGD/IMCA) had a higher chance of having an irregular surface. Based on these patterns described, we have developed a possible algorithm to differentiate the different types of polyps using the VLE features (FIGURE 4). This will need to be validated prospectively. Not surprisingly the abnormal features seen in this study are similar to abnormal VLE
TE D
features seen in Barrett’s esophagus with dysplasia and intramucosal cancer13,14. Both of these neoplastic processes are epithelial based, which occur in columnar mucosal lining.
By defining the VLE criteria for various polyps and associated dysplasia grade in this
EP
study, we may be able to apply VLE to in-vivo use for a number of clinical scenarios. We especially think that VLE could play a role in IBD surveillance given that dysplasia can be difficult to find. The presence of dysplasia increases the risk of cancer significantly in
AC C
these patients18. VLE can potentially help differentiate pseudopolyps from dysplastic polyps and also identify normal versus dysplastic tissue in the setting of quiescent IBD. Given this potential promise, our group is commencing an ex-vivo study of VLE scanning on colectomy specimens in IBD performed for dysplasia and correlation to histology. In addition, this technology may be useful for surveillance of polypectomy sites after EMR of large, flat colon polyps. When residual type tissue is seen in the area of the polypectomy site, it is either inflammatory/hyperplastic tissue vs. residual dysplastic tissue. VLE can potentially help differentiate the residual tissue; dysplastic
ACCEPTED MANUSCRIPT
tissue would need further therapeutic treatment such as additional resection, argon plasma coagulation ablation or avulsion. VLE may also play a role in high-risk patients for colon cancer, ie, Lynch syndrome and personal or family history of colon cancer. Despite advanced optics of colonoscopes a significant amount of colon polyps can still be
RI PT
missed in these patients, up to 26% per one systematic review15 with interval cancers
occurring approximately 2% to 6% of the time16,17 With the advent of laser marking with VLE, superficial cautery marks can be placed on areas of suspected dysplasia for
SC
sampling or removal.
This ex-vivo study should provide a stimulus for further research into the potential use of
M AN U
VLE for imaging the colon. Next steps after ex-vivo identification of VLE features are to prospectively validate our findings and possibly scan colectomy specimens with smaller polyps or focal dysplasia to determine if VLE can identify focal disease. VLE in this study scanned larger polyps to identify VLE features of polyps, but clinically VLE would be useful for in-vivo detection of focal dysplasia or lesions that can be missed in routine colonoscopy. Finally, future studies would also be helpful in determining VLE features
invasive cancer.
TE D
that can distinguish adenomas from hyperplastic polyps and intramucosal cancer from
Modifications on the current probe would need to be made before in-vivo use of VLE in
EP
the colon. The current commercially available VLE probe is designed to fit through a standard upper endoscope. Although the cecum can be reached often with a
AC C
gastroscope20,21, creating a longer catheter so the probe can be compatible with the longer colonoscopes would be desirable. Another issue would be that the balloon containing the probe would need to be bigger as currently the largest balloon size is 20 mm, and this would not appose well to the colon epithelial lining. Fortunately, these hardware adjustments might not be difficult to perform. Rather, a more challenging issue would be of imaging around the flexures of the colon or a tortuous sigmoid. The long 6-cm balloon would make this challenging. A shorter balloon would be more ideal for these areas of the colon but would add more time to image other areas of the colon.
ACCEPTED MANUSCRIPT
In conclusion, we identify ex-vivo VLE features of colonic polyps in this study. We show that there are distinct VLE features of colon polyps, which correlate with the structural pattern of colonic polyps on histology at a lower magnification. These VLE features may help in identifying and classifying the colonic polypoid lesions into a low- or high-grade
RI PT
pattern. Our findings have implications for possible application of in-vivo use of this
technology to aid in dysplasia or polyp detection. Future ex-vivo and in-vivo studies as
SC
outlined in the paper, using the criteria outlined here, are warranted.
1.
M AN U
REFERENCES
Oluyemi AO, Welch AR, Yoo LJ, Lehman EB, McGarrity TJ, Chuang CH. Colorectal cancer screening in high-risk groups is increasing, although current smokers fall behind. Cancer [Internet]. 2014 Jul 15;120:2106–13. Available from: http://doi.wiley.com/10.1002/cncr.28707
Rex DK, Boland CR, Dominitz JA, Giardiello FM, Johnson DA, Kaltenbach T, et
TE D
2.
al. Colorectal Cancer Screening: Recommendations for Physicians and Patients From the U.S. Multi-Society Task Force on Colorectal Cancer. Gastroenterology [Internet]. 2017 Jul;153:307–23. Available from:
3.
EP
http://www.ncbi.nlm.nih.gov/pubmed/28600072 US Preventive Services Task Force, Bibbins-Domingo K, Grossman DC, Curry
AC C
SJ, Davidson KW, Epling JW, et al. Screening for Colorectal Cancer: US Preventive Services Task Force Recommendation Statement. JAMA [Internet]. 2016;315:2564–75. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27304597
4.
Lieberman D, Ladabaum U, Cruz-Correa M, Ginsburg C, Inadomi JM, Kim LS, et al. Screening for Colorectal Cancer and Evolving Issues for Physicians and Patients. JAMA [Internet]. 2016;316:2135. Available from: http://jama.jamanetwork.com/article.aspx?doi=10.1001/jama.2016.17418
5.
Lee J, Park SW, Kim YS, Lee KJ, Sung H, Song PH, et al. Risk factors of missed
ACCEPTED MANUSCRIPT
colorectal lesions after colonoscopy. Medicine (Baltimore) [Internet]. 2017;96:e7468. Available from: http://www.ncbi.nlm.nih.gov/pubmed/28682916 6.
Sivak M V, Kobayashi K, Izatt JA, Rollins AM, Ung-Runyawee R, Chak A, et al.
RI PT
High-resolution endoscopic imaging of the GI tract using optical coherence tomography. Gastrointest Endosc [Internet]. 2000 Apr;51(4 Pt 1):474–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/10744825 7.
Pfau PR, Sivak M V, Chak A, Kinnard M, Wong RCK, Isenberg GA, et al. Criteria
SC
for the diagnosis of dysplasia by endoscopic optical coherence tomography. Gastrointest Endosc [Internet]. 2003;58:196–202. Available from:
M AN U
http://www.ncbi.nlm.nih.gov/pubmed/12872085 8.
Westphal V, Rollins AM, Willis J, Sivak M V, Izatt JA. Correlation of endoscopic optical coherence tomography with histology in the lower-GI tract. Gastrointest Endosc [Internet]. 2005;61:537–46. Available from: http://www.ncbi.nlm.nih.gov/pubmed/15812406
9.
Trindade AJ, Smith MS, Pleskow DK. The new kid on the block for advanced
TE D
imaging in Barrett’s esophagus: a review of volumetric laser endomicroscopy. Therap Adv Gastroenterol [Internet]. 2016;9:408–16. Available from: http://tag.sagepub.com/cgi/doi/10.1177/1756283X16639003 10.
Trindade AJ, Leggett CL, Chang KJ. Volumetric laser endomicroscopy in the
EP
management of Barrett’s esophagus. Curr Opin Gastroenterol [Internet]. 2017 Apr 11; Available from: http://www.ncbi.nlm.nih.gov/pubmed/28402993 Trindade AJ, Sultan K, Vamadevan AS, Fan C, Sejpal D V. Successful use of
AC C
11.
volumetric laser endomicroscopy in imaging a rectal polyp. Therap Adv Gastroenterol [Internet]. 2016;9:128–31. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26770274
12.
The Paris endoscopic classification of superficial neoplastic lesions:
esophagus, stomach, and colon: November 30 to December 1, 2002. Gastrointest Endosc [Internet]. 2003;58(6 Suppl):S3-43. Available from: http://www.ncbi.nlm.nih.gov/pubmed/14652541 13.
Swager A-F, Tearney GJ, Leggett CL, van Oijen MGH, Meijer SL, Weusten BL, et
ACCEPTED MANUSCRIPT
al. Identification of volumetric laser endomicroscopy features predictive for early neoplasia in Barrett’s esophagus using high-quality histological correlation. Gastrointest Endosc [Internet]. 2016 Sep 19; Available from: http://dx.doi.org/10.1016/j.gie.2016.09.012 Leggett CL, Gorospe EC, Chan DK, Muppa P, Owens V, Smyrk TC, et al.
RI PT
14.
Comparative diagnostic performance of volumetric laser endomicroscopy and confocal laser endomicroscopy in the detection of dysplasia associated with Barrett’s esophagus. Gastrointest Endosc [Internet]. 2015; Available from:
15.
SC
http://linkinghub.elsevier.com/retrieve/pii/S0016510715028229
van Rijn JC, Reitsma JB, Stoker J, Bossuyt PM, van Deventer SJ, Dekker E. Polyp
M AN U
miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol [Internet]. 2006;101:343–50. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16454841 16.
Brenner H, Chang-Claude J, Seiler CM, Hoffmeister M. Interval cancers after negative colonoscopy: population-based case-control study. Gut [Internet]. 2012;61:1576–82. Available from:
17.
TE D
http://www.ncbi.nlm.nih.gov/pubmed/22200840 Bressler B, Paszat LF, Chen Z, Rothwell DM, Vinden C, Rabeneck L. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology [Internet]. 2007;132:96–102.
18.
EP
Available from: http://www.ncbi.nlm.nih.gov/pubmed/17241863 Ullman T, Croog V, Harpaz N, Sachar D, Itzkowitz S. Progression of flat low-
AC C
grade dysplasia to advanced neoplasia in patients with ulcerative colitis. Gastroenterology [Internet]. 2003;125:1311–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/14598247
19.
Beintaris I, Rutter M. Advanced imaging in colonoscopy: contemporary
approach to dysplasia surveillance in inflammatory bowel disease. Frontline Gastroenterol [Internet]. 2016;7:308–15. Available from: http://www.ncbi.nlm.nih.gov/pubmed/28839872
20.
Bick BL, Vemulapalli KC, Rex DK. Regional center for complex colonoscopy: yield of neoplasia in patients with prior incomplete colonoscopy. Gastrointest
ACCEPTED MANUSCRIPT
Endosc [Internet]. 2016;83:1239–44. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26584785 21.
Rex DK, Chen SC, Overhiser AJ. Colonoscopy Technique in Consecutive
Hepatol [Internet]. 2007;5:879–83. Available from:
RI PT
Patients Referred for Prior Incomplete Colonoscopy. Clin Gastroenterol
SC
http://linkinghub.elsevier.com/retrieve/pii/S1542356507002881
FIGURE 1: Components of the volumetric laser endomicroscopy system. A, Console
into a gastroscope.
M AN U
in which the distal end of the catheter attaches. B, Balloon catheter inserted
FIGURE 2: Volumetric laser endomicroscopy features in the colon. A, Normal colon. There is normal reflectance (orange arrow). The red arrow is the muscularis
TE D
mucosa. Above this layer is the epithelium. Below it is the submucosal layer. B, VLE of a tubular adenoma. The epithelium is darker and thus hyperreflective compared with normal colon (orange arrow). Layering is present (defined as presence of muscularis mucosa; red arrow). C, VLE of a
EP
tubulovillous adenoma. There is a darker hyperreflective epithelium (orange arrow). There is lack of layering or effacement given no muscularis mucosae is
AC C
visualized. D, A tubular adenoma with high-grade dysplasia. There is a hyperreflective surface (orange arrow), atypical glands (short yellow arrows),
and effacement. E) A polyp with intramucosal cancer. There is a hyperreflective surface (orange arrow), an irregular surface (blue arrow), and
effacement.
FIGURE 3: Representative colon polyps with high risk volumetric laser endomicroscopy and matched histologic features. A, A tubulovillous adenoma with high-grade dysplasia on colonoscopy. B, post-endoscopic mucosal
ACCEPTED MANUSCRIPT
resection of the polyp seen in A C, VLE features of the polyp in A showing an irregular surface, effacement, and greater than 5 glands. D, Matched histology of the polyp in A showing distended and irregular crypts in scattered areas and fused crypts with cribriform architecture (HE; x 40 magnification). E, A
RI PT
polyp with intramucosal cancer on colonoscopy. F, post EMR of the polyp seen in E. G, VLE features of the polyp in E showing an irregular surface and
effacement. H, Matched histology with areas of highly irregular dysplastic
colonic crypts and associated lamina propria stromal reaction. Note that some
SC
fused crypts are identified. However, distended glands are not identified in
region with low magnification pattern of intramucosal adenocarcinoma (H &
M AN U
E; orig. mag. x40). I, A tubular adenonma with high-grade dysplasia and a focus of intramucosal cancer seen on colonoscopy. J, post-EMR of the polyp seen in I. K, VLE features of the polyp in I showing an irregular surface, effacement, and greater than 5 glands. L, Matched histology with fused complex colonic crypts lined by high-grade epithelium (H & E; orig. mag.
TE D
x100).
FIGURE 4: Representative colon polyps with low risk volumetric laser endomicroscopy and matched histologic features. A, A adenoma with lowgrade dysplasia on colonoscopy. B, post-endoscopic mucosal resection of the
EP
polyp seen in A. C, VLE features of the polyp in A showing regular surface contour, fewer than 5 glands, and no effacement. D, Matched histology of the
AC C
polyp in A showing uniform colonic crypts and smooth surface contour (H & E; orig. mag. x40). E, A serrated adenoma on colonoscopym F, post-endoscopic mucosal resection of the polyp seen in E. G, VLE features of the polyp in E showing normal reflectance pattern and no effacement. H, Matched histology of the polyp in E showing uniform serrated crypts and smooth surface contour. Also note that serrated lesions may have prominent benign adipose tissue in the submucosa (arrow) (H & E; orig. mag. x40).
FIGURE 5: Proposed algorithm for identification of the type of polyp based on
ACCEPTED MANUSCRIPT
volumetric laser endomicroscopy features.
TABLE 1: Volumetric laser endomicroscopy features of the colon polyps.
Glands 1-5
Glands >5
-
-
-
-
0% (0/21)
48% (10/21 )
M AN U
CONTROL n=2 Normal n=2 NON-ADVANCED PATHOLOGY n=26 Tubular 100 % 10% Adenoma (TA) (21/21) (2/21) with LGD n=21 Sessile Serrated 0% 20% Adenoma (0/5) (1/5) n=5 ADVANCED PATHOLOGY n=17 Tubulovillous 100% 67% adenoma (TVA) (9/9) (6/9) with LGD n=9 TA or TVA with 100% 100% HGD n=5 (5/5) (5/5)
Irregula No r Surface Glands
RI PT
Effacemen t of all layers
SC
HyperReflectivit y
AC C
100% (8/8)
0% 0/21
0% (0/5)
100% (5/5)
0% (0/5)
0% (0/5)
22% (2/9)
11% (1/9)
44% (4/9)
44% (4/9)
80% (4/5)
0% (0/5)
40% (2/5)
60% (3/5)
100% (3/3)
100% (3/3)
0% (0/3)
67% (2/3) (67%)
33% (1/3)
100% (8/8)
88% (7/8)
0% (0/8)
50% (4/8)
50% (4/8)
TE D
100% (3/3)
EP
Intramucosal Adenocarcinoma (IMCA) n=3 Composite IMCA/HGD N=8
52% (11/21 )
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
SC
M AN U
VLE= Volumetric Laser Endomicroscopy EMR= Endoscopic Mucosal Resection OCT= Optical Coherence Tomography HGD= high-grade dysplasia IMCA= intramucosal cancer TA= tubular adenoma TVA= tubulovillous adenoma LGD= low grade dysplasia SSA= sessile serrated adenoma CRC= colorectal cancer H&E= with hematoxylin and eosin
RI PT
Acronyms