icon-folder.gif   Conference Reports for NATAP  
  The International Liver Congress™
EASL - European Association for the
Study of the Liver
June 23-26 2021
Back grey_arrow_rt.gif
About 2-Fold Higher Risk of MI,
HCC, CKD When NASH Patients Have Cirrhosis

  EASL International Liver Congress, June 23-26, 2021
Mark Mascolini
People with nonalcoholic steatohepatitis (NASH) in a large US database had about 2-fold higher incidence of myocardial infarction, hepatocellular carcinoma, and chronic kidney disease if they also had cirrhosis [1]. The cirrhosis group also had twice higher mortality. The narrow confidence intervals in these 93,000-person estimates-even for rare outcomes-bolster growing evidence of the baleful clinical impact of advanced liver fibrosis with NASH.
Researchers from Optum Life Sciences and Intercept Pharmaceuticals noted that research has yielded scant longitudinal data on progression and outcomes with NASH, with either fibrosis or cirrhosis. This is especially true for rare outcomes, like HCC and liver transplant, which require a large number of individuals for meaningful analysis.
The Optum/Intercept team proposed that available electronic health records could provide "useful real-world data" for studying the epidemiology of NASH because they contain structured data, semistructured data, and unstructured data. They used Optum's EHR (electronic health record) Research Database [2] for this analysis. The database includes structured information (diagnosis, procedures, prescriptions, administrations, lab results, clinical observations), semistructured data (clinical concepts, attributes, notes), and unstructured full-text clinical notes.
The analysis included people identified with NASH between 2016 and 2019 by structured and semistructured data. Researchers classified people with NASH as having cirrhosis or fibrosis based on information from the year before they entered the study cohort. Evidence of fibrosis or cirrhosis could include: ICD-10 codes, natural language processing data, and elevated fibrosis score (by APRI, nonalcoholic fatty liver disease (NAFLD) fibrosis score (NFS), Fib-4, or FibroSure derived from natural language processing data. The investigators estimated outcome incidence rates in the subset of people at risk for that outcome at the start of follow-up, after excluding people who already had that outcome. Researchers counted 93,204 people with NASH, including 19,347 (21% of 93,204) with no fibrosis or cirrhosis, 45,331 (49%) with fibrosis, and 28,526 (31%) with cirrhosis. For the entire group, the no-fibrosis/cirrhosis group, the fibrosis group, and the cirrhosis group, age at cohort entry averaged 58, 53, 57, and 64. Respective proportions of whites were 81%, 79%, 80%, and 84%, of blacks 6%, 5%, 6%, and 6%, of Hispanics 8%, 9%, 8%, and 6%, and of Asians 2%, 2%, 2%, and 1%. Women made up 60% of each of the four groups.
Overall, at cohort entry the group had an average body mass index of 35.6 kg/m2, 49% had type 2 diabetes, 75% had hypertension, and 67% had hyperlipidemia. In the whole NASH group, the no-fibrosis/cirrhosis group, the fibrosis group, and the cirrhosis group, proportions who already had liver biopsy were 3.5%, 2.6%, 3.0%, and 4.8%; median Fib-4 stood at 1.43, 0.74, 1.28, and 2.57, median NFS at -0.29, -2.34, -0.67, and 1.39, and median APRI at 0.38, 0.24, 0.35, and 0.59.
Researchers reported incidence as rate per 1000 person-years. The liver transplant rate was almost 3-fold higher with cirrhosis (2.75) than in the all-NASH group (0.92) and almost 20-fold higher than in the fibrosis group (0.14). Incidence of variceal bleeding was more than 3-fold higher in the cirrhosis group (8.59) than in the all-NASH group (2.60), compared with 0.09 in the fibrosis group. Similar incidence rate differences in the all-NASH group, the fibrosis group, and the cirrhosis group could be seen for hepatic encephalopathy (8.28, 0.95, 26.15), esophageal varices (10.21, 1.23, 33.15), portal hypertension (14.21, 2.26, 45.18), ascites (20.01, 7.58, 44.74), and liver failure (21.87, 5.00, 72.81). Incidence of hepatocellular carcinoma was more than twice higher with cirrhosis (7.64) than in the all-NASH group (3.03) versus 1.00 in the fibrosis group.
Certain major nonliver outcomes had about twice higher incidence rates with cirrhosis than in the all-NASH group, including chronic kidney disease (216.93 vs 124.96 per 1000 person-years), acute kidney injury (105.16 vs 54.23), and myocardial infarction (27.40 vs 16.14). Overall mortality proved more than twice greater with cirrhosis (71.13) than in the all-NASH group (29.31). Other major comorbidity outcomes had moderately higher incidence with cirrhosis: type 2 diabetes, hyperlipidemia, and stroke or transient ischemic attack.
The Optum/Intercept team cautioned that they did not assess the positive predictive value of their algorithms for NASH, fibrosis, or cirrhosis, and they stressed that more than 80% of participants were white. But the analysis involved a big, geographically diverse US population including people with private insurance, public insurance, and no insurance, and with recorded race/ethnicity.
The researchers stressed that the higher risk of various liver and nonliver outcomes with more advanced fibrosis "add to the evolving body of literature on the impact of hepatic fibrosis in patients with NASH." They hope these findings will give clinicians and researchers "additional context to better differentiate between the effects of pharmacotherapy and the underlying disease process."
1. Bertoia M, Ness E, Capozza T, Titievsky L, Seeger JD. Incidence rates of select outcomes among patients with nonalcoholic steatohepatitis (NASH) and evidence of fibrosis or cirrhosis. EASL International Liver Congress, June 23-26, 2021. Abstract OS-1780.
2. Optum. Optum HER Database. https://www.optum.com/business/solutions/life-sciences/real-world-data/ehr-data.html