D0197 - Assessment of the Accuracy of a Clinical Risk Prediction (Kunzmann) Score 5 Years Prior to Barrett’s Esophagus and Esophageal Adenocarcinoma Diagnosis: Results From a Large Population-Based Database

Karan Sachdeva, MBBS¹, Alex Liu, MD², Don C. Codipilly, MD¹, Erin Gibbons, BS¹, Ramona Lansing, BSN¹, Kenneth K. Wang, MD, FACG², Prasad G. Iyer, MD, FACG^1
1Mayo Clinic, Rochester, MN; ²Mayo Clinic Rochester, Rochester, MN

Introduction: Barrett’s esophagus (BE) is the only known precursor lesion of esophageal adenocarcinoma (EAC), a malignancy with poor 5-year survival. Screening for BE is endorsed in those with risk factors. Assessing BE/EAC risk remains challenging. The Kunzmann score is a tool to identify patients at risk of BE/EAC. We assessed the ability of this tool to predict BE/EAC risk 5 years prior to BE/EAC diagnosis in a large population-based database.



Methods: The Rochester Epidemiology Project contains medical records of over 90% of SE Minnesota residents since 1977. Utilizing appropriate ICD-9 and 10 codes, we identified all BE/EAC patients from 1977-2020. Endoscopic evidence of at least 1 cm of salmon colored mucosa in the tubular esophagus and presence of intestinal metaplasia on endoscopic biopsies were assessed to confirm BE diagnosis. We also identified non-BE/EAC controls, and endoscopic reports were reviewed to exclude BE/EAC in these patients. The Kunzmann score is a composite score based on age, gender, smoking history, presence of esophageal conditions (such as heartburn), and BMI. It has previously been reported to have a sensitivity and specificity for EAC of 77.5% and 70.5%, respectively, utilizing a cut off score of 8. We compared risk prediction scores between BE patients and non-BE controls at datapoints obtained 5 (±1) years prior to BE diagnosis.

Results: Data to calculate the Kunzmann score were abstracted from 916 cases (684 BE,232 EAC) 5 years prior to diagnosis and from 100 endoscopy negative controls. The mean Kunzmann score 5 years prior to diagnosis was significantly higher in the EAC group (10.6; SD: 2.0) compared to those with baseline BE (9.7; SD: 2.5) and controls (10.2; SD: 2.5; p < 0.01). Furthermore, the percentage of patients with a Kunzmann score greater than 8 at 5 years prior to diagnosis was highest in the EAC group (90.9%) compared to the baseline BE (78.8%) and control (83.0%; p< 0.01) groups. Utilizing a cut-off score of 8, the Kunzmann score at 5-years prior to diagnosis demonstrated a sensitivity of 84.0% and specificity of 18.1% for the diagnosis of BE/EAC, and demonstrated a sensitivity of 79.5% and specificity of 9.1% for the diagnosis of EAC alone.



Discussion: In this large population-based database, the Kunzmann score demonstrated reasonable sensitivity but low specificity to predict BE/EAC at 5 years prior to diagnosis. Its utility for predicting BE/EAC risk needs to be further evaluated.



Disclosures:


Karan Sachdeva, MBBS¹, Alex Liu, MD², Don C. Codipilly, MD¹, Erin Gibbons, BS¹, Ramona Lansing, BSN¹, Kenneth K. Wang, MD, FACG², Prasad G. Iyer, MD, FACG¹. D0197 - Assessment of the Accuracy of a Clinical Risk Prediction (Kunzmann) Score 5 Years Prior to Barrett’s Esophagus and Esophageal Adenocarcinoma Diagnosis: Results From a Large Population-Based Database, ACG 2022 Annual Scientific Meeting Abstracts. Charlotte, NC: American College of Gastroenterology.