Do Improving Hepatic Biomarkers Predict Spontaneous Stone Passage? A Retrospective Analysis of ERCP Yield in Suspected Choledocholithiasis

BACKGROUND: Choledocholithiasis (CDL) is a common indication for endoscopic retrograde cholangiopancreatography (ERCP). While guidelines use static hepatic biomarker values to estimate pre-test probability, the utility of trending these labs (bilirubin and alkaline phosphatase [ALP]) is unclear. It is unknown if improving biomarker trends reliably indicate stone passage and can help avoid unnecessary procedures. OBJECTIVE: To evaluate whether serial trends in hepatic biomarkers prior to ERCP can predict the presence of suspected CDL in hospitalized patients and guide decisions on ERCP referral. METHODS: We conducted a retrospective study of 198 patients undergoing ERCP for suspected CDL from 2002 to 2018. Patients were categorized based on biomarker trends, primarily into: Group 1 (normalized bilirubin with normalized ALP or ALP falling ≥ 50%) and Group 2 (all other patients). The primary outcome was the diagnostic yield of CDL on ERCP. RESULTS: Among 198 patients, the yield of choledocholithiasis was not significantly different between Group 1 (69.5%) and Group 2 (71.2%) (p = 0.83). Even when redefining Group 1 as patients with "entirely normal hepatic biomarkers," the yield for stones was 55%. The diagnostic accuracy of the final lab values was poor, with an area under the curve of 0.57 for bilirubin and 0.51 for ALP. Patients in Group 1 had a significantly higher rate of post-ERCP pancreatitis (8.5% vs. 0%, p = 0.012) and a lower risk of septic shock. CONCLUSIONS: Normalization or decline in hepatic biomarkers, such as bilirubin and ALP, does not reliably exclude CDL. Over half of the patients with normalized biomarkers were still found to have stones on ERCP. ERCP should be considered when suspicion remains high, regardless of laboratory trends.