Uniformity Between Serological Biomarker Test, Esophago-gastro-duodenoscopy and Biopsy Histology in Triage of Upper Abdominal Symptoms in Gastroscopy Referral Patients.

[BACKGROUND/AIM] The use of serological biomarker testing for diagnosing upper abdominal symptoms and screening for gastric cancer (GC) risk conditions is increasing worldwide, raising questions about its diagnostic accuracy. This study evaluated the diagnostic concordance between a serological biomarker panel, gastroscopy (EGD), and biopsy histology in patients referred for gastroscopy.

[PATIENTS AND METHODS] A total of 522 patients referred from primary care to Oulu University Hospital (Finland) underwent GastroPanel (GP) testing, EGD, and gastric biopsies classified using the Updated Sydney System (USS). Three gastroenterology experts independently assessed cases with discrepant results to determine preferred management. Agreement between diagnostic methods was measured using kappa statistics.

[RESULTS] Of the GP test-negatives (n=447), 60% had normal findings on EGD, 36% (n=161) were diagnosed as inflammatory lesions and 4% (n=19) were classified as having possible or definite atrophic gastritis (AG). Biopsies revealed moderate AG in only three GP-negative patients (0.007%). GP and USS showed excellent agreement (weighted k=0.861) while agreement between GP/EGD and EGD/USS was moderate (weighted k=0.458 and k=0.480, respectively). Of the 175 discrepant diagnoses, over 90% were ranked by all gastroenterologists as benign conditions with no need for therapy or EGD follow-up.

[CONCLUSION] GP, EGD, and USS do not always provide uniform diagnoses; however, after a negative GP test, the probability of detecting moderate/severe AG using EGD is extremely low (3/447; 0.007%). In the first-line diagnosis of upper abdominal symptoms, substantial cost savings (90%) could be achieved by preserving EGD only for those (<10%) patients whose GP test indicates AG.