HBV-related cirrhosis in the Chronic Hepatitis Cohort Study (CHeCS)
Gordon SC, Rupp LB, Boscarion JA, Schmidt MA, Trinacty CM, Lamerato L, Oja-Tebbe N, Lu M. HBV-related cirrhosis in the Chronic Hepatitis Cohort Study (CHeCS). Hepatology 2015; 62(S1):980A.
PURPOSE: The overall spectrum of liver disease among Americans with chronic hepatitis B (CHB) infection remains unknown, and staging liver biopsy has become increasingly uncommon. We used four different methods, including liver biopsy, to identify potential cirrhosis among CHB patients enrolled in the Chronic Hepatitis Cohort Study (CHeCS), an ongoing observational study at four large, integrated health systems in the United States.
METHODS: We included patients who met predefined inclusion criteria for CHB whose case status had been confirmed through chart abstraction. We excluded patients coinfected with hepatitis C and liver transplant recipients. Data were collected through 2012 from liver biopsy reports, lab results, and diagnosis/procedure codes contained in the electronic health record. We calculated FIB-4 scores based on most recently available aminotransferase and platelet count values collected when a patient was not on antiviral therapy. Cirrhosis was identified via four methods: (1) liver biopsy reports, (2) presence of a diagnosis code for cirrhosis (ICD-9 diagnosis codes 571.2 and 571.5), (3) presence of a diagnosis or procedure code for end stage liver disease (ESLD), and (4) FIB-4 score >5.17, a cut-off value previously validated as predictive of cirrhosis in this cohort. We compared the extent to which cirrhosis was indicated by each method individually and by multiple methods.
RESULTS: Among the 2,731 CHB patients, 24% were <40 and 26% >60 years old, 54% were male, and 53% were Asian. Median follow-up time was 6.0 years. Of the total, 564 (21%) patients had at least one biopsy report during follow-up, and of those, 85 (15% of those who had a biopsy) had a recent biopsy less than two years old. Overall, cirrhosis was indicated for 520 (19%) patients by at least one of the four methods: by liver biopsy for 66 patients (2% of patients overall, and 12% among those with any liver biopsy report); by FIB-4 score for 281 (10%); by diagnosis code for cirrhosis for 318 (12%); and by diagnosis or procedure code for ESLD for 216 (8%), 145 of whom had a diagnosis code for both cirrhosis and ESLD. Of the 281 patients with FIB-4 score >5.17, only 26 (9%) had been diagnosed with cirrhosis via biopsy, 135 (48%) via ICD-9 code for cirrhosis, and 113 (40%) via diagnosis or procedure code indicating ESLD.
CONCLUSION: An estimated 19% of CHB patients among our cohort had indication of cirrhosis at some point in their follow-up through 2012. The use of additional parameters, including a previously validated biomarker as a surrogate for more advanced liver disease, potentially identifies additional unrecognized cirrhosis in untreated CHB patients.