Degree and Duration of Alkaline Phosphatase and Total Bilirubin Elevation Below Treatment Thresholds Predict Clinical Outcomes in Primary Biliary Cholangitis

Abstract

Research Question: Among individuals diagnosed with PBC, do chronic elevations in ALP and TB below presently accepted guidelines correlate with an increased risk of adverse clinical outcomes (including time to the first occurrence of hospitalization, hospitalization for hepatic decompensation, liver transplant, or death) compared to PBC patients whose ALP and TB values align with or exceed accepted guidelines?

Background and Significance: Primary biliary cholangitis (PBC) is an uncommon liver condition that mainly impacts women aged 50 and above. Existing models that evaluate the risk of disease advancement using biomarker values typically rely on individual baseline measurements. This study aimed to assess the influence of persistent elevations in alkaline phosphatase (ALP) and total bilirubin (TB) on clinical outcomes, even at levels lower than those indicated in current treatment guidelines.

Materials and Methods: The population was patients diagnosed with PBC in the Komodo Health claims linked to two national laboratory databases. The primary analysis was a Cox regression treating proportion of time ALP exceeded specific thresholds of Upper Limit of Normal (ULN 1×; 1.2×; 1.5×; 1.67×; 2.0×) and TB thresholds below and above ULN (0.6×; 0.8×; 1.0×; 2.0×) on time to first occurrence of death, liver transplant or hospitalization for hepatic compensation.

Results: 2,379 patients met all inclusion/exclusion criteria. Patients were predominantly female (85.3%) with a mean age of 63.3 years (SD=13.1). Median follow-up time was 2.2yrs (IQR: 1.1 – 3.9 years) with a mean of 7.2 ALP measures and 8.4 TB measures. For both ALP and TB, greater proportion of time spent above each threshold, and greater divergence from ULN for ALP and 0.6x ULN for TB, were associated with increased risk of a composite outcome event. A 10% increase in proportion of time ALP>ULN was associated with a statistically significant increased risk of event (HR=1.06; 95% CI=1.03, 1.09; p<0.001) as was a 10% increase in proportion of time of TB>0.06xULN (HR=1.09; 95% CI=1.06, 1.12: p<0.001). There was more than 4-fold increase in the risk of an event in patients with both ALP>ULN and TB>0.06xULN 100% of the time.

Conclusion: These findings propose a shift in the management of PBC, moving away from a binary endpoint approach (e.g., <1.67×ULN or >1.67×ULN) to a more nuanced understanding of the continuous relationship between biomarkers and outcomes. Physicians should evaluate both the absolute level of a biomarker and the duration it remains at that level when assessing the risk of disease progression and determining treatment strategies. Further research is essential to explore persistent elevations in alternative biomarkers, as well as in measures of fibrosis and inflammation.