Submucosal Chromoendoscopy. a Technique That Highlights Epithelia

REVIEW

Submucosal chromoendoscopy. A technique that highlights epithelia and differentiates histological components, and renders colon polypectomy easier and safer Carlos Dolz-Abadía and Angels Vilella-Martorell Department of Digestive Diseases. Hospital Son Llatzer. Palma de Mallorca. Spain

ABSTRACT instance, in violet-colored candy), drinks, and ice cream. It also serves as pH indicator. Submucosal chromoendoscopy involves the injection of a The classical technique for topical chromoendoscopy solution containing a vital stain, usually indigo carmine, into the intestinal wall submucosal layer. This allows to: Better delimit involves the instillation of indigo carmine over the colon- and characterize the various epithelia present (colonic mucosa, ic epithelium and lesions. This requires an optimal or at adenoma, hyperplastic polyp, serrated polyp, small bowel mucosa); least adequate level of cleanliness. Some authors previ- expose and delimit lesion implantation areas; cooperate in the lifting ously administer a mucolytic agent (N-acetylcysteine) to of resectable lesions; ensure section across the submucosal plane; identify intestinal wall structures; render complex polypectomy cleanse as thoroughly as possible the target area. Atropine, feasible; and facilitate the identification of perforations. butylhyoscine bromide or glucagon may be administered The present paper offers information on the endoscopic to obtain a hypokinetic bowel (3). technique for submucosal injection, solution preparation and Then the dye is injected through a spray catheter at a con- concentration, and on the potential benefits it may provide for centration ranging from 0.1% to 0.8%. Once injected, the polypectomy or endocopic mucosal resection whether en block or piecemeal. This endoscopic technique simultaneously combines a dye may be washed out with water or air through the same diagnostic and a therapeutic aspect, since lesion lifting in association spray catheter should the stain be too dark (2). Mucosal with better delimited contours may improve not only accuracy but irregularities and pit patterns are then carefully examined, also endoscopic resection safety and feasibility. usually with Kudo’s pit pattern (4,5) (Fig. 1) and Sanos’s capillary pattern (Fig. 2) classifications (6,7). In non-con- Key words: Chromoendoscopy. Colonic polypectomy. trolled studies aimed at identifying colonic preneoplastic lesions, indigo carmine increased detection rates for flat and smaller lesions overlooked by conventional colonoscopy INTRODUCTION (8-11). Randomized controlled studies comparing conventional white-light colonoscopy and chromoendoscopy have Chromoendoscopy or chromoscopy is defined as the shown strong evidence that the latter enhances premalignant topical application of dyes during endoscopy in order polyp detection in the colon and rectum (12-16). Moreover, to highlight tissue characteristics (1). The dyes used for the sensitivity of indigo carmine to predict polyp histology digestive chromoendoscopy are categorized according to (adenomatous vs. hyperplastic) was 82% and 95%, with a their absorption. The first group includes absorbable dyes specificity of 64% and 95%, respectively. such as methylene blue, lugol, toluidine blue, and crystal The injection of indigo carmine into the submucosal violet. Dyes in the second group are not absorbed, and layer, or submucosal chromoendoscopy, is a variant of the include indigo carmine. Finally, dyes in the third group above classical modality used to verify the submucosal loca- modify their color according to pH, as is the case with tion of the injected solution. The injection of any solution congo red and phenol red (2). into the submucosal layer results in a bulging that will not Indigo carmine is the most widely used stain for colo- develop during injection into the muscular layer. The addi- rectal endoscopy. It occurs naturally in the sap of the shrub tion of indigo carmine to the injected solution colors the Indigofera tinctoria, but is currently manufactured syn- resulting bulge and confirms its location within the submu- thetically. It may be used as a food coloring in sweets (for cosa. This averts perforation risk during polypectomy (17).

Received: 25-09-2014 Accepted: 07-01-2015 Dolz-Abadía C, Vilella-Martorell A. Submucosal chromoendoscopy. A technique that highlights epithelia and differentiates histological components, Correspondence: Carlos Dolz Abadía. Department of Digestive Diseases. Hos- and renders colon polypectomy easier and safer. Rev Esp Enferm Dig pital Son Llatzer. Ctra. Manacor, km. 4. 07198 Palma de Mallorca 2015;107:430-435. e-mail: [email protected] Vol. 107, N.º 7, 2015 Submucosal chromoendoscopy. A technique that highlights epithelia and differentiates 431 histological components, and renders colon polypectomy easier and safer

I Round pit (normal pit) Normal

II Asteroid pit Hyperplasic

Tubular or round pit that is smaller than the III Adenoma S normal pit (type I)

Tubular or round pit that is larger than the normal III Adenoma L pit (type I)

IV Dendritic or gyrus-like pit Adenoma

Irregular arrangement and sizes of III , III , IV type V S L Carcinoma I pit pattern

Loss or decrease of pits with an amorphous V Carcinoma N structure

Fig. 1. Kudo’s pit pattern classification for the colonic mucosa.

Capillary pattern I II IIIA IIIB

Schema

Endoscopic findings

Meshed capillary vessels characterized by: Blind ending, Meshed capillary vessels (-) branching and curtailed irregularly Capillary Meshed capillary vessels (-) Capillary vessel sureounds Lack of uniformity characteristics Nearly avascular or loose mucosal glands High density of capillary micro capillary vessels vessels A. Type I CP: MC (-), capillaries not visible with NBI. Hyperplastic polyp. Non-neoplastic lesions. B.1. and B.2. Type II CP: MC (+), capillaries visible, regularly distributed around crypts. Adenomas (tubular and villous, respectively). Neoplastic lesions. Type II I CP: MC (+), irregular capillaries visible. Cancer. Neoplastic lesions. C.1. Type IIIA CP: Increased capillary vessel density and loss of capillary vessel uniformity. Compromised m/ssm (m: mucosa; ssm: superficial submucosa < 1,000 mcm). C.2. Type IIIB CP: Avascular areas. Compromised dsm (deep submucosa ≥ 1,000 mcm).

Fig. 2. Sano’s microvascular pattern classification for colon adenomas.

Rev Esp Enferm Dig 2015; 107 (7): 430-435 432 C. Dolz-Abadía AND A. VILELLA-MARTORELL Rev Esp Enferm Dig (Madrid)

Both the muscularis propria and the submucosal fibrotic In this paper we shall not refer to the solution used for tissue are weakly stained by indigo carmine. The muscula- endoscopic submucosal dissection. Instead, we shall refer ris propria layer is seen as concentric white fibers whereas to the solution used for classical polypectomy, endoscop- fibrosis appears as a dense white-yellowish structure (3,18). ic mucosal resection or fragmented resection -anyway, a Most of the available information on the submucosal resection that will not in theory take longer than 15-20 injection of indigo carmine involves the study of adeno- minutes. The composition we recommend in table I is the mas, colon cancer, and ulcerative colitis (12-16,19,20). In result of a number of combinations tried in our practice, the last few years the submucosal injection of indigo car- the one we finally deemed more appropriate. mine has been widely used for endoscopic submucosal dis- The liquid base we shall use is a saline solution (0.9% section (21,22). However, overall, the injection of indigo sodium chloride) in a 100-cc container. Reloading from a carmine into the submucosa is a non-standard technique, 100-cc bag-container is more efficient and convenient as and information available is scarce on the how, when and compared to filling 10-cc syringes. We add adrenaline, 1 mg where of its use. per 100 cc of saline. This concentration is 10 times lower This manuscript is an attempt to offer information on than usual when attempting to achieve hemostasis for active the way submucosal chromoendoscopy may be carried out, bleeding. We must underscore two reasons for the addition of and on the benefits it may provide. adrenaline to a submucosal chromoendoscopy solution: First, to prevent potential bleeding; second, to obtain vasoconstriction in highly vascular epithelia, as is the case with some ade- ENDOSCOPIC TECHNIQUE nomas. We have seen that adenomas become paler following this sort of injection, and that such decoloration helps indigo First, the target lesion is identified and cleansed, and carmine in delimiting the various epithelia present. We deem secretions are aspirated. A hypokinetic agent is given for a higher adrenaline concentration unnecessary, but could be excessive intestinal peristaltism. We use glucagon rather used to manage potential bleeding following resection. than butylhyoscine bromide (Buscapine, Buscopan) as the Indigo carmine is available at several concentrations latter more commonly increases heart rate. ranging from 0.1% to 0.8%. Its dosing at such concentra- Once the lesion’s surface is attentively examined its tions, with no dilution, provides a deep blue color that attachment is delimited and established. Pedicled lesion precludes differentiating the various components of intes- resection usually entails no difficulties, but sessile or flat tinal wall layers. When injecting a dye into the wall our lesions may be challenging or even impossible to excise. goal is obtaining information on the various components In such cases delimiting the lesion’s attachment extent we may come across in order to make better decisions. and contour is crucial. It is here that submucosal chro- We must find a blue hue allowing differentiation between moendoscopy may be of invaluable help, as it will lift and collagen fibrous tissue and muscle tissue, veins and arter- more clearly delineate the lesion’s base by highlighting ies, fat from submucosa, etc. Having obtained a number its contours. This is feasible even for lesions spanning of dilutions we believe that adding 0.4 cc of 0.8% indigo two aspects of a haustrum or fold. In other words, it may carmine to 100 cc of saline results in an appropriate blue render feasible a lesion initially deemed impossible to hue for our purposes (Table I). Other compositions and resect. combinations have also been used, as the one suggested by Bourke -succinate-polymerized gelatin plus indigo carmine 80 mg in 500 ml of solution plus adrenaline 1:10.000 Injected solution composition (1 ml per 10 ml injected) (28).

No recommendations are available regarding the components of solutions for submucosal chromoendoscopy and Perilesional injection their concentrations. There is only some consensus in the endoscopic submucosal dissection setting, but modifica- The method and sequence of the injection procedure tions according to authors are many (23-28). are very important. The final result of a consensus on the

Table I. Composition and concentrations of the solution used for submucosal chromoendoscopy Substance and material Dose/Measure Comments Saline 100 ml More practical than filling 10-20 cc syringes. Enough for a complete day of work Adrenaline 1 mg Concentration 10 times lower than for hemostasis. Highlights more vascular epithelia Indigo carmine 0.4 cc of a 0.8% solution Pastel blue hue adequate for the purposes pursued Injection needle 25 G = 0.5 mm Greater diameter = solution leaks from puncture holes

Rev Esp Enferm Dig 2015; 107 (7): 430-435 Vol. 107, N.º 7, 2015 Submucosal chromoendoscopy. A technique that highlights epithelia and differentiates 433 histological components, and renders colon polypectomy easier and safer technique, with the different authors using differing meth- Lesion resection ods. However, the recommendations issued by Bourke are highly useful (28) and may be considered a reference. A After submucosal injection completion we shall have first option would consist of starting the injection at the a clearer view of the lesion’s attachment base, which will lesion’s posterior aspect so that the bulging obtained brings the lesion toward us, from back to front. Once the posterior aspect is lifted, the bulging should be induced on A B both sides, left and right, in a symmetrical, matching way. Finally, we must examine the lesion on its anterior aspect. Injection at this level may also be required, but always in smaller volumes as compared to the posterior and lateral procedures; failure to comply with this would mar the previously obtained visual plane. This is common in narrower colonic segments such as the sigmoid colon. It should be noted the perilesional mucosal bulging detracts from endoscopic view, hence it is recommended that the needle be slowly withdrawn into the endoscope’s working channel as the bulging develops (Fig. 3-6). The needle recommended for injection should have a Fig. 4. LST-G polyp in the ascending colon. With and without (white light) rather small bore to prevent the injected solution from (A) and submucosal chromoendoscopy (B). leaking towards the intestinal lumen. A 25 G (0.5 mm) size seems appropriate to us. The volume to be injected oscillates according to lesion dimensions. A sessile lesion about 25 mm in diameter may admit 5-20 cc of solution. More is better than less since the bulging slowly but steadily flattens over time; however, the injected volume is not to block the visual field thus precluding adequate lesion delimitation. Some drops of injected solution usually leak from puncture orifices. This should entail no difficulties -rather, superficial epithelial staining helps define their characteristics and better delineate lesion borders. Fig. 5. Flat polyp. Tubular adenoma in the ascending colon. A small trickle of blood from the injection site may be seen.

Fig. 6. Pedunculated polyp. Tubular adenoma. A submucosal injection Fig. 3. Ileocecal valve. Transition between ileum and tubular adenoma of indigo carmine allows to precisely delimit the transition between the epithelium (arrow). pedicle’s normal mucosa and the adenoma.

Rev Esp Enferm Dig 2015; 107 (7): 430-435 434 C. Dolz-Abadía AND A. VILELLA-MARTORELL Rev Esp Enferm Dig (Madrid)

allow an estimation of the lesion’s histological nature risk for complications following polypectomy for lesions and invasion depth by using the Kudo and Sano classi- greater than 2 cm in size, although controlled, randomized fications (4,5,7). With this we are in a better position to studies are needed to demonstrate this in a prospective man- decide whether the lesion should be excised and the best ner. Reinjecting the indigo carmine solution is also useful therapeutic modality to be used (Figs. 1 and 2). Be as it when in doubt regarding potential microperforations. Rein- may, the final objective should be visual reassurance that jection may confirm an appropriate resection plane, thus a complete resection, that is, an R0 resection was obtained reducing the risk for delayed perforation (17). of the neoformed tissue. Factors in favor of submucosal In summary, submucosal chromoendoscopy may attain injection (submucosal chromoendoscopy) include bet- several goals: ter lesion exposure from the lifting action exerted with a – b etter delimitation and characterization of the differ- potential disappearance of the haustra effect, and complete ent epithelia present, thus telling the normal colonic exposure of difficult-to-identify neoplastic tissue as occurs mucosa from that of lesions. with some serrated polyps. Some of these serrated polyps, – e xposure and definition of a lesion’s attachment the so-called SSA/P-D (sessile serrated adenomas/polyps) base and its morphology (this may render feasible growths, have a non-dysplastic component that may be the resection of a lesion previously considered non- easily overlooked, thus resulting in incomplete endoscopic resectable). resection (29-33). – Confirmation that resection took place along the sub- Better lesion exposure permits a safer, more accurate mucosal layer. endoscopic resection. Similarly, lesion lifting averts perfo- – Improved perforation detection. ration risk by increasing intestinal wall thickness, and may minimize bleeding risk as a result of the vasoconstriction provided by the diluted adrenaline associated with indigo REFERENCES carmine. Often, a lesion initially considered non-resectable 1. Devuni D, Vaziri H, Anderson JC. Chromocolonoscopy. Gastroenterol Clin North Am 2013;42:521-45. DOI: 10.1016/j.gtc.2013.05.002 becomes endoscopically resectable in just one fragment, 2. Wong KeeSong LM, Adler DG, Chand B, et al. Chromoendoscopy. which allows the pathology study of a single block and ASGE Technology Committee. Gastrointest Endosc 2007;66:639-49. ensures a precise histological characterization. A second 3. Mönkemüller K, Wilcox CM. Interventional chromoendoscopy. Gas- option is fragmented (piecemeal) resection following sub- trointest Endosc 2013;78:346-50. DOI: 10.1016/j.gie.2013.04.181 4. Kudo S, Hirota S, Nakajima T, et al. Colorectal tumours and pit pattern. mucosal injection. Piecemeal resection entails uncertain J Clin Pathol 1994;47:880-5. DOI: 10.1136/jcp.47.10.880 pathological diagnosis and uncertain cure, hence must be 5. Kudo S, Tamura S, Nakajima T, et al. Diagnosis of colorrectal tumour- considered a valid though not preferable option. With good ous lesions by magnifying endoscopy. Gastrointest Enodsc 1996;44:8- visual access, a resection en block should have margins 14. DOI: 10.1016/S0016-5107(96)70222-5 6. Fu KL, Sano Y, Kato S, et al. Chromoendoscopy using indigo carmine < 25-30 mm in the distal colon and rectum, and < 20-25 dye sparring with magnifying observation is the most reliable method mm in the right colon (28). for differential diagnosis between non-neoplastic and neoplastic color- Once resection is completed, the bluish tinge of the sub- rectal lesions: A prospective study. Endoscopy 2004;36:1089-93. DOI: 10.1055/s-2004-826039 mucosa grants that it took place along that plane, and lower 7. Uraoka T, Saito Y, Ikematsu H, et al. Sano’s capillary pattern classifica- perforation risks. The sight of linear, parallel, whitish struc- tion for narrow-band imaging of early colorectal lesions. Dig Endosc tures will suggest the muscle layer was reached, and depth 2011;23(Supl. 1):112-5. DOI: 10.1111/j.1443-1661.2011.01118.x was excessive. Small ectopic fat collections may be identi- 8. Hurlstone DP, Sanders DS, Cross SS, et al. Colonoscopic resection of lateral spreading tumours: A prospective analysis of endoscopic fied in the submucosa, particularly in the proximal colon, mucosal resect. Gut 2004;53:1334-9. DOI: 10.1136/gut.2003.036913 which should not be mistaken for peritoneal fat; the latter 9. Rembacken BJ, Fujii T, Cairns A, et al. Flat and depressed colonic neo- would suggest perforation. Red areas may also be identi- plasms: A prospective study of 1000 colonoscopies in the UK. Lancet 2000;355:1211-4. DOI: 10.1016/S0140-6736(00)02086-9 fied that represent submucosal bleeding and should not be 10. Lee JH, Kim JW, Cho YK, et al. Detection of colorectal adenomas by mistaken for muscle fibers, which are a whitish pale color routine chromoendoscopy with indigo carmine. Am J Gastroenterol (18). Prior staining with India ink, as in a tattoo, may confer 2003;98:1284-8. DOI: 10.1111/j.1572-0241.2003.07473.x a typical blackish hue to the submucosal layer, and result 11. Brown SR. Baraza W. Chromoscopy versus convetional endoscopy for the detection of polyps in the colon and rectum. Cochrane Database in wall fibrosis and loss of elasticity, which translates into Syst Rev 2010;(10):CD006439. higher perforation risk after endoscopic resection; from all 12. Brooker JC, Saunders BP, Shah SG, et al. Total colonic dye-spray the above, endoscopic resection should be avoided on previ- increases the detection of diminutive adenomas during routine colonos- ously tattooed segments if possible (34,35). The use of pro- copy: A randomized controlled trial. Gastrointest Endosc 2002;56:333- 8. DOI: 10.1016/S0016-5107(02)70034-5 phylactic endoscopic procedures to avoid digestive bleed- 13. Hurlstone DP, Cross SS, Slater R, et al. Detecting diminutive colorectal ing after the endoscopic resection of large colonic lesions lesions at colonoscopy: A randomised controlled trial of pan-colonic (including the use of clips to bring together the borders of versus targeted chromoscopy. Gut 2004;53:376-80. DOI: 10.1136/ gut.2003.029868 the mucosal defect left after the excision) is now open to 14. Lapalus MG, Helbert T, Napoleon B, et al. Does chromoendoscopy debate (36). Closing the resection base may reduce bleeding with structure enhance the colonoscopic adenoma detection rate? risk (37). Thus, the prophylactic use of clips may reduce the Endoscopy 2006;38:444-8.

Rev Esp Enferm Dig 2015; 107 (7): 430-435 Vol. 107, N.º 7, 2015 Submucosal chromoendoscopy. A technique that highlights epithelia and differentiates 435 histological components, and renders colon polypectomy easier and safer

15. Le Rhun M, Coron E, Parlier D, et al. High resolution colonoscopy 26. Matsui Y, Inomata M, Izumi K, et al. Hyaluronic acid stimulates tumor- with chromoscopy versus standard colonoscopy for the detection of cell proliferation at wound sites. Gastrointest Endosc 2004;60:539-43. colonic neoplasia: A randomized study. Clin Gastroenterol Hepatol DOI: 10.1016/S0016-5107(04)01890-5 2006;4:349-54. DOI: 10.1016/j.cgh.2005.12.009 27. Bures J, Kopácová M, Kvetina J, et al. Different solutions used for sub- 16. Stoffel EM, Turgeon DK, Stockwell DH, et al. Great Lakes New Eng- mucosal injection influence lyhealing of gastric endoscopic mucosal land Clinical Epidemiology and Validation Center of the Early Detection resection in a preclinical study in experimental pigs. SurgEndosc Research Network. Chromoendoscopy detects more adenomas than colo- 2009;23:2094-101. noscopy using intensive inspection without dye spraying. Cancer Prev 28. Bourke MJ. Endoscopic mucosal resection in the colon: A practical Res (Phila) 2008;7:507-13. DOI: 10.1158/1940-6207.CAPR-08-0096 guide. Tech Gastroientest Endosc 2011;13:35-49. DOI: 10.1016/j. 17. Holt BA, Jayasekeran V, Sonson R, et al. Topical submucosal chro- tgie.2011.01.002 moendoscopy defines the level of resection in colonic EMR and 29. Pohl H, Srivastava A, Bensen SP, et al. Incomplete polyp resec- may improve procedural safety (with video). Gastrointest Endosc tion during colonoscopy-results of the complete adenoma resection 2013;77:949-53. DOI: 10.1016/j.gie.2013.01.021 (CARE) study. Gastroenterology 2013;144:74-80 DOI: 10.1053/j. 18. Tutticci N, Bourke MJ. Interventional chromoendoscopy: Specific gastro.2012.09.043 aspects for the colon. Gastrointest Endosc 2014;79:536-8. DOI: 30. Gancayco J, Siddiqui UD, Jain D, et al. Narrow band imaging features 10.1016/j.gie.2013.09.002 and pathological correlations of sessile serrated polyps. Am J Gastro- 19. Matsumoto T, Iwao Y, Igarashi M, et al. Endoscopic and chromoendo- enterol 2011;106:1559-60 DOI: 10.1038/ajg.2011.122 scopic atlas featuring dysplastic lesions in surveillance colonoscopy 31. Tadepalli US, Feihel D, Miller KM, et al. A morphologic analysis of for patients with long-standing ulcerative colitis. Inflamm Bowel Dis sessile serrated polyps observed during routine colonoscopy. Gastroin- 2008;14:259-64. DOI: 10.1002/ibd.20267 test Endosc 2011;74:1360-8. DOI: 10.1016/j.gie.2011.08.008 20. Kiesslich R, Neurath MF. Chromoendoscopy in inflammatory bow- 32. Hazewinkel Y, López-Cerón M, East JE, et al. Endoscopic features el disease. Gastroenterol Clin North Am 2012;41:291-302. DOI: of sessile serrated adenomas: Validation by international experts 10.1016/j.gtc.2012.01.016 using high-resolution white-light endoscopy and narrow-band 21. Herreros de Tejada A. ESD training: A challenging path to excel- imaging. Gastrointest Endosc 2013;77:916-24. DOI: 10.1016/j. lence. World J Gastrointest Endosc 2014;6:112-20. DOI: 10.4253/ gie.2012.12.018 wjge.v6.i4.112 33. Burgess NG, Tutticci NJ, Pellise M, et al. Sessile serrated adenomas/ 22. Saito Y, Uraoka T, Yamaguchi Y, et al. A prospective, multicenter study polyps with cytologic dysplasia: A triple threat for interval cancer. of 1111 colorectal endoscopic submucosal dissections (with video). Gastrointest Endosc 2014;80:307-10. DOI: 10.1016/j.gie.2014.03.050 Gastrointest Endosc 2010;72:1217-25. DOI: 10.1016/j.gie.2010.08.004 34. Trakarnsanga A, Akaraviputh T. Endoscopic tattooing of colorec- 23. Conio M, Rajan E, Sorbi D, et al. Comparative performance in the tal lesions: Is it a risk-free procedure? World J Gastrointest Endosc porcine esophagus of different solutions used for submucosal injec- 2011;3:256-60. tion. Gastrointest Endosc 2002;56:513-6. DOI: 10.1016/S0016- 35. Moss A, Bourke MJ, Pathmanathan N. Safety of colonic tattoo with 5107(02)70435-5 sterile carbon particle suspension: A proposed guideline with illus- 24. Hyun JJ, Chun HR, Chun HJ, et al. Comparison of the charac- trative cases. Gastrointest Endosc 2011;74:214-8. DOI: 10.1016/j. teristics of submucosal injection solutions used in endoscopic gie.2011.01.056 mucosal resection. Scand J Gastroenterol 2006;41:488-92. DOI: 36. Bourke MJ. Bleeding following wide-field endoscopic resection in the 10.1080/00365520500325994 colon. Gastroenterol Hepatol 2011;7:814-7. 25. Kantsevoy SV, Adler DG, Conway JD, et al. Endoscopic mucosal 37. Liaquat H, Rohn E, Rex DK. Prophylactic clip closure reduced the risk resection and endoscopic submucosal dissection. ASGE Technol- of delayed postpolypectomy hemorrhage: Experience in 277 clip large ogy Committee. Gastrointest Endosc 2008;68:11-8 DOI: 10.1016/j. sessile or flat colorectal lesions and 247 control lesions. Gastrointest gie.2008.01.037 Endosc 2013;77:401-7. DOI: 10.1016/j.gie.2012.10.024

Rev Esp Enferm Dig 2015; 107 (7): 430-435