Barrett's Esophagus: Sorting Out the Controversy

Barrett's Esophagus: Sorting Out the Controversy Learning Objectives 1. Identify the challenges in screening for Barrett s esophagus 2. Demonstrate comprehension of the risk of progression of Barrett s esophagus and dysplasia to cancer 3. Recognize the indications for referral for endoscopic eradication therapy 4. Outline the current limitations with surveillance endoscopy and endoscopic eradication therapies Faculty V. Raman Muthusamy, MD, FACG, FASGE Professor of Medicine Director of Endoscopy Division of Digestive Diseases David Geffen School of Medicine at UCLA Los Angeles, California Slides are current as of the time of printing and may differ from the live presentation due to copyright issues. Please reference www.pri-med.com/west for the most up-to-date version of slide sets. West Annual Conference Anaheim, California April 27-30, 2016

Barrett s esophagus: History Mr. Norman Barrett Pioneering thoracic surgeon (1903-1979) Originally believed a congenitally shortened esophagus was key in pathogenesis; later accepted the concept of sliding hiatal hernia Also initially supported a developmental etiology for columnar-lined esophagus Current Definition of Barrett s Esophagus Barrett s esophagus (BE) implies replacement of a portion of the squamous epithelium of distal esophagus by specialized intestinal epithelium (IM). Three Essential Steps for Endoscopic Diagnosis and Description Recognize the squamocolumnar junction Identify gastroesophageal junction Describe extent of columnar mucosa AGA Barrett s Workshop, 2003 Tytgat GN, et al. Gastroenterol Clin North Am 1997; 26:507 The major endoscopic landmark for the gastroesophageal junction: The top of the gastric mucosal folds Barrett s esophagus: Prevalence Prevalence of BE in adults: 0.4 1.6% 880,000 3.5 million U.S. adults have BE In GI-select populations: 6-20% of patients undergoing endoscopy for GERD Up to 25% of an asymptomatic VA population Cameron AJ. Gastroenterol Clin North Am 1997;26:487 94. Winters C, et al. Barrett s esophagus. Gastro 1987;92:118 24. Ronkainen J, et al. Gastroenterology 2005;129:1825 31. Sampliner RE, et al. Gastroenterology 2005;129:2101 2113. Rex DK, et al. Gastroenterology 2003;125:1670 7. Gerson LB, et al. Gastroenterology. 2002;123:461 7. Ward EM, et al. Am J Gastroenterol. 2006;101:12 17.

Evolution of Barrett s esophagus Injury Acid & bile reflux nitrous oxide Genetics Gender, race,? other factors (cox-2) Accumulate Genetic Changes Natural History of HGD Most studies suggest 6-10% progression to cancer per year Wani et al, Am Journal of Gastro, 2009 Meta-analysis suggested 6.6%/yr Shaheen et al, NEJM, 2009 RCT of RFA vs. Surveillance HGD progression rate of 19% in 1 year in surveillance arm Rigorous confirmation of HGD pre-enrollment Rigor f/u and biopsy protocol Natural History of LGD: Earlier Data Interval LGD to HGD/Cancer Sharma, et al. 1 year risk 2.7% 5 year risk 13.5% Weston, et al. 1 year risk 3.1% 5 year risk 15.5% Skacel, et al. 1 year risk 12.7% 5 year risk 63.5% Lim, et al 1 year risk 3.3% 5 year risk 16.5% Wani, et al. 1 year risk 1.7% (cancer) 5 year risk 8.5 % (cancer) More Recent Data Regarding LGD Progression Meta-analysis suggests 1.7%/yr risk of progression to cancer Recent study (N=210) suggests 0.44%/yr to cancer with HGD progression of 1.6%/yr Another study of 147 LGD pts (85% of whom were downstaged) suggests progression rate of 13.4%/yr to HGD/CA and 85% risk of progression at 109.1 months. Shaheen, NEJM, 2009 13.6% progression to HGD in 1 year in surveillance arm Wani et al, Am Journal of Gastro, 2009 Wani et al, Gastroenterology 2011;141(4):1179-86. Curvers WL, AM J Gastro 2010;105(7):1523-30 Shaheen et al, NEJM, 2009; 360 (22): 2277-88. Natural History: BE without dysplasia Sharma, et al. Shaheen, et al. NDBE Progression LGD HGD Cancer 1 year risk 4.0% 0.9% 0.5% 5 year risk 20.0% 4.5% 2.5% 1 year risk na na 0.5% (0.0-2.7%) 5 year risk na na 2.5% (0.0-13.5%) Labenz, et al. 1 year risk na na 1.3% 5 year risk na na 6.5% Hvid-Jensen 1 year risk 0.12% Wani et al 1 year risk 0.6% (Meta-analysis) Bhat et al 1 year risk 0.13% 5 year risk 0.65% Wani et al 1 year risk 3.6% 0.48% 0.27%

Are We Underestimating Risk? Hvid-Jensen study eliminated 131 cancers found in 1 st year of study ( incident disease ) Without exclusion, rate goes for 0.12% to 0.36% Also excluded all HGD in 1 st year Misclassification of BE 32.3% of those w/ BE were not confirmed (95% CI 24.4-41.1) in a community study VA study found 18% of LSBE and 33% of SSBE couldn t be confirmed Ganz GIE, 80:5 2014 Should We Screen? BE associated with increased risk for esophageal adenocarcinoma (EAC) Potentially this risk is up to 40x that of the general population Risk of cancer between 0.5-1.0%/yr (all levels of dysplasia combined) EAC has dismal 17% 5-yr survival BE, however, may have negligible effect on overall mortality Which GERD Patients to Screen? Men > 50 years Caucasian race GERD symptoms for > 5 years Nocturnal Reflux Hiatal Hernia Elevated BMI Tobacco use Intra-abdominal distribution of body fat Screen patients with multiple risk factors: Weak rec., low-moderate quality evidence Don t need to rescreen negative patients: <=2.4% yield on f/u endoscopy AGA Medical Positional Statement on Management of BE, Gastroenterology 2011; 140:1084-1091 ASGE Guideline on BE and Other Premalignant Conditions of the Esophagus: GIE 2012; 76:6, 1087-94 Upper Endoscopy for GERD: Best Practice Advice from CGC of the ACP: Ann Int Med 2012; 157: 808-816. The Barrett s Iceberg Prevalence of Barrett s in Subjects Undergoing Colonoscopy Most Barrett s Undetected Endoscopy: 22.6/100,000 Autopsy: 376.0/100,000 GERD symptoms not present in 60% of Barrett s patients in population based study Only 23/589 pts diagnosed with EAC in Kaiser study had known BE >= 6 months Cameron, et al. Gastroenterology 1990; 99: 918 Ronkainen et al, Gastroenterology, 2005; 129 (6): 1825-31. Corley DA et al, Gastroenterology 2002;122(3):633-40. % 10 8 6 4 2 0 8.3% 2.6% Heartburn n=384 5.7% 5.6% BE LSBE SSBE 0.4% 5.2% No Heartburn n=556 Rex, et al. Gastroenterology 2003; 125:1670

How Should We Screen? Standard Endoscopy Unsedated Endoscopy Capsule Endoscopy? Sponge Cell Acquisition Device? Something else? (Blood/Saliva) Cytosponge for BE Screening Office-based, < 10 min Can be done by PCPs Tests for trefoil factor 3 Study of 504 pts 99% swallowed sponge 3% diagnosed with BE 73% Se and 94% Sp for >=1 cm of BE 90% Se and 93.5% Sp for >=2 cm of BE Kadri et al, BMJ 2010; 341 How Often or Long Should We Screen? CORI National endoscopic database 1/1/2000-12/31/2004 24,406 pts had an initial EGD negative for BE and at least 1 subsequent EGD during the study period 2.3% (N=561) had? BE on f/u EGD (3.1% M:1.2% F) 9.9% BE on f/u EGD when esophagitis on initial exam vs. only 1.8% when no initial esophagitis Rodriguez S et al, Am Jl Gastro, 2008, 103:1892 1897 Technique of Biopsy in Endoscopic Surveillance Technique: NDBE: 4 quadrant q 2 cm Q 1 yr x 2; then q 3-5 yr LGD: 4 quadrant q 1 cm Repeat: Q 6-12 months HGD: 4 quadrant q 1 cm Repeat q 3 months Abrams JA, Clin Gastroenterol Hepatol 2009;7(7):736-42. Curvers WL Eur J Gastroenterol Hepatol 2008; 20(7):601-7. Compliance: US Study of 2245 cases Adherence rate was 51.2% Lower compliance with longer BE (N=150; Netherlands) 0-5 cm: 79% 5-10 cm: 50% 0-15 cm: 30% Problems with Surveillance Expensive ($596,000/QALY saved) Detects, not prevents, cancer Leads to patient anxiety Samples only 3-5% of mucosa Poor compliance associated with reduced dysplasia detection No quality data showing mortality reduction from EAC Population study showed only 23/589 pts diagnosed with EAC had known BE >= 6 mo Inadomi J, et al. Ann Intern Med 2003; 138:178 Corley DA, Gastroenterology 2002;122(3):633-40.

More on Surveillance Corley et al Gastro 145(2), Aug 2013 8272 BE patients; 351 with adenocarcinoma 70 pts with prior dx of BE (51 dead; 38 from CA) Surveillance hx contrasted with 101 living BE pts Surveillance within 3 yrs: Not associated with decreased risk from CA (OR 0.99 [.36-2.75] Fatal cases received surveillance 55.3% of time; controls: 60.4% El-Serag HB et al, Gut, 2015 29,536 veterans with BE; 424 developed cancer over 5 yr f/u Surveillance associated with reduced risk of cancer related death (HR 0.47; 0.35-0.64) Surveillance associated with earlier diagnosis, longer survival, and lower cancer related mortality Increased Yield with Specialized Brush 39.8% increase in Barrett s esophagus detection in GERD patients 42.1% increase in dysplasia detection c/t biopsy in patients w/ dysplasia undergoing surveillance Anandasapathy, Dig Dis Sci, 2011 Johanson, Dig Dis Sci, 2011 Surveillance Post Endoscopic Treatment: Is a Brush Better than Biopsies? AIM: Evaluate adjunctive value of specialized brush biopsy to forceps biopsy for detection of residual/recurrent BE after Barrett s ablation. 2 centers Specialized brush and standard biopsies taken at same session Specialized brushings taken first, then 4 quadrant q 1 cm random biopsies Abstract #345, Iorio et al, DDW 2015 Surveillance Post -Treatment: Is a Brush Better than Biopsies? Mean BE Length = 3.88 cm 18.8% (39/208) w/ IM, dysplasia, or neoplasia on forceps biopsy alone Specialized brush detected 24 additional cases (incremental yield of 11.5%) increasing detection of endpoint from 18.8% to 30.3% This is a 61% augmentation of the yield for IM, dyplasia, or neoplasia NNT = 8.7 to find one additional recurrence Abstract #345, Iorio et al, DDW 2015 Probe Based Confocal Laser Endomicroscopy A unique endomicroscopy system Probe-based system, used during endoscopic procedures Compatible with existing endoscopes Real-time visualization of tissue at the cellular level Access the entire GI tract Different probe types for different indications and needs Providing real-time visualization of tissues

Endoscopic Eradication Techiques (EET) for BE Endoscopic Mucosal Resection Thermal Ablation Argon Plasma Coagulation Multipolar Electrocoagulation Laser Ablation Radiofrequency Ablation Photochemical Ablation Photodynamic Therapy (PDT) Cryotherapy Ablation CO2 based Liquid Nitrogen based Endoscopic Mucosal Resection Diagnosis Staging Treatment When to perform EMR Evidence of any focal abnormality (typically raised, can be depressed) Lesions < 20 mm that require en bloc resection STAGING EMR 135 patients (HGD 92, EAC 43) undergoing EET at 3 centers Biopsy (HGD/EAC) EMR (within 3 months) Change in diagnosis in up to 45% Larger lesions can be resected in a piecemeal fashion Wani S et al, Dig Dis Sci 2013 Efficacy and Durability of RFA: Systematic Review and Meta-Analysis (Orman, Clin Gastroenterol Hepatol, 2013) Authors assessed: 18 studies of 3802 pts reporting efficacy 6 studies of 540 pts reporting durability CE-IM in 78% of pts; CE-D in 91% After eradication, IM recurred in 13% Median follow-up 16.5 months Progression to cancer: 0.2% of pts during treatment 0.7% of pts following CE-EM Stricture was most common adverse event, occurring in 5%Conclusion: Treatment of BE with RFA results in CE-D and CE-IM in a high proportion of pts, with few recurrences of IM after treatment and a low rate of adverse events. 52

AGA Medical Position Statement Gastroenterology 2011;140:1084-1091 HGD: Endotherapy with RFA, PDT, or EMR is recommended rather than surveillance LGD: RFA should be a therapeutic option for treatment of patients with confirmed LGD NDBE: RFA with or without EMR should be a therapeutic option for select individuals with NDBE who are judged to be at increased risk for progression to HGD or cancer SAGES Guideline Guideline for surgical treatment of GERD Includes section on management of Barrett s esophagus with evidence grading HGIN and IMC can be managed with RFA +/- EMR for eradication of lesion and reduction in cancer (Level I evidence) Surgery remains option for HGIN and IMC NDBE (IM), IND, LGD may be effectively eradicated with RFA (Level I evidence) Anti-reflux surgery for GERD may be performed in patients with NDBE (IM), IND, LGD in conjunction with endoscopic eradication therapy (i.e., before, during, after RFA) Source: http://www.sages.org/publication/id/22/ ACG Guidelines 2015 For T1a cancer, endoscopic therapy is the preferred approach Endoscopic therapy is the preferred approach for confirmed LGD and HGD Agreement by two pathologists For LGD, some debate if LGD needs to be present on two occasions Non-dysplastic Barrett s should not be routinely treated with endoscopic ablative therapies due to its low risk of progression to esophageal cancer. Ablation Effect on Natural History Natural History (53 studies) After Ablation (65 studies) NDBE LGD HGD 0.6% 1.7% 6.6% 0.16% 0.16% 1.7% NNT=45 NNT=13 NNT= 4 Polyp 0.58% 0.06% NNT= 38 Progression risk expressed as Per-patient-per-year (%) risk of developing EAC NNT calculated on 5-year basis (number needed to treat to avoid one cancer over 5 years) Esophageal adenocarcinoma in BE: a meta-analysis. Wani S, et al. Am J Gastro 2009 Prevention of colorectal cancer by colonoscopic polypectomy. Winawer SJ, et al. NEJM 1993

NDBE: Who is High Risk? Family history? Long-segment? (>=3 cm) Young age? (<=60) History of prior dysplasia? Patient with poorly controlled GERD? Elevated BMI? Men? Hernia size? Rationale 50 y.o. man diagnosed with Long Segment NDBE Currently would get endoscopy at 51, 55, 59, 63, 67, 71, and 75 Assuming 3 ablation sessions, 1 yr f/u, 5 yr f/u, and 10 year f/u, would save 1 endoscopy and benefit from reduction in risk of progression. The Future: 7 Key Questions How do we best screen/identify patients with Barrett s Esophagus? What are the clinical and biomarker risk factors that predict progression of non-dysplastic BE? How can surveillance technique/yield be maximized? What are the risk factors for prolonged/failed ablation and recurrence? What are the appropriate post-ablation surveillance intervals? When, if ever, can surveillance post-ablation be stopped? In Whom? Is there a role for chemoprophylaxis? Proposed Future Algorithm for BE Inexpensive, Non-Invasive (Saliva/Blood), Highly Sensitive Screening Test for General Population Barrett's Esophagus Suspected Perform Endoscopy Barrett's Esophagus Confirmed Risk Stratification with Biomarker and Clinical Risk Panel Very Low Risk of Progression No Further Evaluation Low-Intermediate Risk of Progression Chemoprophylaxis and Surveillance No Further Evaluation High Risk of Progression Endoscopic Eradication Therapy No Barrett's Suspected No Further Evaluation Barrett's excluded Muthusamy & Sharma, GI Endoscopy Clinics N. Am, 2010 Conclusions A cost-effective screening method to identify patients with Barrett s esophagus is needed New strategies to increase the area sampled via surveillance should be developed. Confirmed Barrett s with dysplasia has a significant rate of progression and are amenable to endoscopic eradication therapies. Highly effective, safe, and durable endoscopic eradication techniques are available Clinical and biomarker risk factors that predict which non-dysplastic patients will progress to dysplasia and would benefit from treatment remain The Holy Grail of Barrett s Esophagus