Communication and Surveillance Recommendations on Gastric Intestinal Metaplasia by US Gastroenterologists Before and After the 2020 American Gastroenterological Association Guidelines: Data From a Tertiary Care Center

Introduction: Gastric intestinal metaplasia (GIM) is a precursor to non-cardia gastric cancer (NCGC). Studies showed variability in gastroenterologists’ (GE) knowledge and management of GIM, attributed to limited guidance from US GI societies. In 2020, the American Gastroenterological Association (AGA) published GIM guidelines, recommending against routine surveillance and advising physician/patient shared decision-making on surveillance based on risk factors. We hypothesize awareness of risk factors for GIM progression and surveillance recommendations have increased since the guidelines. We hence sought to compare GE’s communication trends of GIM finding as well as surveillance recommendations before and after 2020 at a tertiary care center. Methods: A sample of patients with GIM on biopsies from upper endoscopies (EGD) performed between 1/2018 and 12/2022 were included in this retrospective cohort study. Exclusions were history of gastric cancer or surgery, cardia-only GIM and hereditary cancer syndromes. Cohort A included GIM patients diagnosed < 2020 and cohort B >2020. Data were manually collected from the electronic medical record. GEs’ communication of a GIM finding was deemed present if there was a communication about GIM via phone, MyChart, letter, office visit or other. Mapping recommendation to determine GIM extent was defined if repeat endoscopy was recommended within 1 year of baseline EGD and surveillance recommendation if EGD advised >1 year. Pearson's Chi-square and Fisher’s exact tests were used for categorical variables and Wilcoxon rank sum test for continuous variables. Analyses were performed with R software. Results: 347 patients were included: 175 in cohort A and 172 in cohort B. Median age was 65.7 [56.0, 73.4] and 54.5% were females. Patients’ demographic, clinical, endoscopic and pathology characteristics are presented in Table 1. GEs’ communication to patients about GIM finding increased from 24.6% < 2020 to 50% >2020 (P < 0.001). Similarly, there was a significant increase in mapping and surveillance recommendations after 2020 (5.71% vs 12.8%, P = 0.036 and 41.3% >2020 vs 20% < 2020, P < 0.001 respectively) (Figure 1). Conclusion: While communicating a GIM diagnosis to patients was overall low, the rates of communication, surveillance and mapping recommendations increased after 2020, indicating more awareness, likely in the setting of the 2020 AGA guidelines on GIM. More studies are needed to determine if mapping and surveillance are offered to appropriate high-risk patients.Figure 1.: Communication about GIM finding, mapping and surveillance recommendations before and after 2020 GIM: Gastric Intestinal Metaplasia. Table 1. - Patients’ demographics and characteristics GIM: gastric intestinal metaplasia MALT: mucosa-associated lymphoid tissue All (Total N 347)* N (%) Cohort A (< 2020)(Total N 175)*N (%) Cohort B ( > 2020)(Total N 172)* N (%) P-value Demographics and clinical characteristics Age, Median [25th;75th] 65.7 [56.0;73.5] 65.5 [54.9;72.3] 65.8 [57.5;73.6] 0.498 Sex 0.695 Female 189 (54.5) 93 (53.1) 96 (55.8) Ethnicity 0.150 Hispanic/Latino 47 (13.5) 21 (12.0) 26 (15.1) Non-Hispanic/Latino 297 (85.6) 154 (88.0) 143 (83.1) Unknown 3 (0.86) 0 (0.00) 3 (1.74) Race 0.007 Asian 16 (4.61) 13 (7.43) 3 (1.74) Black/African American 63 (18.2) 36 (20.6) 27 (15.7) Caucasian/White 245 (70.6) 112 (64.0) 133 (77.3) Other/ Unknown 23 (6.63) 14 (8.00) 9 (5.23) Smoking history 0.796 Never 170 (49.0) 84 (48.0) 86 (50.0) Former 141 (40.6) 74 (42.3) 67 (39.0) Current 36 (10.4) 17 (9.71) 19 (11.0) Alcohol use history 192 (55.3) 91 (52.0) 101 (58.7) 0.250 Previous history of Helicobacter pylori 63 (18.2) 42 (24.0) 21 (12.2) 0.007 Helicobacter pylori eradication confirmation 42 (68.9) 28 (68.3) 14 (70.0) 1.000 Family history of gastric cancer 33 (9.51) 11 (6.29) 22 (12.8) 0.060 History of pernicious anemia/autoimmune atrophic gastritis 56 (16.1) 25 (14.3) 31 (18.0) 0.424 Endoscopy and pathology characteristics Use of Sydney protocol 0.497 Yes 164 (47.5) 85 (49.1) 79 (45.9) No 118 (34.2) 54 (31.2) 64 (37.2) Probable 44 (12.8) 22 (12.7) 22 (12.8) unknown 19 (5.51) 12 (6.94) 7 (4.07) GIM extent based on location 0.023 Limited 217 (62.5) 110 (62.9) 107 (62.2) Extensive 54 (15.6) 28 (16.0) 26 (15.1) Oxyntic only 45 (13.0) 15 (8.57) 30 (17.4) Fundic only 5 (1.44) 3 (1.71) 2 (1.16) No location specified 26 (7.49) 19 (10.9) 7 (4.07) GIM severity on pathology 0.002 Focal 152 (44.1) 78 (45.1) 74 (43.0) Extensive/severe 37 (10.7) 28 (16.2) 9 (5.23) Not specified 156 (45.2) 67 (38.7) 89 (51.7) GIM subtype < 0.001 Complete 12 (3.48) 1 (0.58) 11 (6.40) Incomplete 2 (0.58) 0 (0.00) 2 (1.16) Mixed 3 (0.87) 0 (0.00) 3 (1.74) Not specified 328 (95.1) 172 (99.4) 156 (90.7) Dysplasia on pathology note 0.176 Low-grade 5 (1.45) 4 (2.30) 1 (0.58) High-grade 2 (0.58) 2 (1.15) 0 (0.00) Autoimmune gastritis suspected on pathology 40 (11.5) 16 (9.14) 24 (14.0) 0.217 Adenocarcinoma on pathology 6 (1.73) 1 (0.57) 5 (2.91) 0.119 MALT lymphoma on pathology 2 (0.58) 2 (1.15) 0 (0.00) 0.499 *Unless otherwise specified Use of Sydney protocol: Yes if 2 biopsies taken from each of the antrum and the corpus (with or without incisura) and placed in separately labeled bottles; Probable if pathology report includes both oxyntic and antral mucosa types; No if does not meet criteria for former and latter; Not specified: biopsy site and mucosa type not specified in endoscopy and pathology reports respectively. Extent: Limited if GIM limited to antrum and/or incisura; Extensive if GIM in antrum and/or incisura and another stomach part other than cardia.