Liver cirrhosis : epidemiology, prognosis, and cancer

Abstract

Liver cirrhosis is a major risk factor for hepatocellular carcinoma (HCC). Patients with liver cirrhosis also have a high risk to develop infections leading to deterioration of liver function and increased mortality. In this PhD-thesis, our aim was to improve the ability to predict risk of developing liver cancer and infections in patients with cirrhosis.

In study I, all patients diagnosed with HCC at the Karolinska University Hospital between 2004 and 2018 were included. Patients with HCC and underlying non-alcoholic fatty liver disease (NAFLD) were characterized in detail to investigate their prognosis compared to that of other patients with HCC. In study II, we included randomly selected patients with an international classification of disease (ICD)-code corresponding to cirrhosis and cirrhosis complications registered in the national patient registry (NPR) between 2000 and 2016 to investigate the positive predictive value (PPV) of liver-related ICD-codes. In study III, all patients with cirrhosis registered in the outpatient part of the NPR were included to investigate rate and risk of HCC in cirrhosis. In study IV, we included patients with cirrhosis seen at the Hepatology clinic at the Karolinska University Hospital and obtained a blood test from the participants analyzed for fractions of mucosal-associated invariant T (MAIT) cells. Patient were followed prospectively for risk of bacterial infection and hepatic decompensation.

In study I, we included 1,562 patients with HCC, and 225 (14%) of these had NAFLD. We report that NAFLD is a growing cause of HCC. One third of the patients with NAFLD-HCC had no clinical signs of cirrhosis. NAFLD patients were older than non-NAFLD patients, and non-cirrhotic NAFLD patients were even older than NAFLD patients with cirrhosis. Survival was similar between patients with NAFLD and non-NAFLD and between patients with cirrhotic and non-cirrhotic NAFLD. In study II, we found that ICD-10 codes for cirrhosis and esophageal varices had a PPV above 90%, whereas HCC had a PPV of 84%. Ascites had an unsatisfactorily low PPV of 43% for liver-related ascites, but when combined with a code indicating chronic liver disease, the PPV increased to 91%. In study III we included 15,215 individuals with cirrhosis and report that the rate of HCC in cirrhosis is 23/1,000 person-years with a lower-than-expected cumulative risk at five and ten years of 8.3% and 12.2% respectively. The cancer risk varied significantly depending on sex, age, and etiology of liver disease. In study IV, we included 106 patients with cirrhosis and found that relatively preserved MAIT cell fractions were associated with a higher risk of bacterial infections in patients with cirrhosis.

In conclusion, we describe NAFLD HCC-patients with and without cirrhosis and found that patients with non-cirrhotic NAFLD are older. We suggest that any surveillance attempts in this patient group should take age into account. ICD-codes for cirrhosis and esophageal varices have a high PPV, but when using ICD-10 code for ascites to identify patients with cirrhosis, we recommend adding another code for chronic liver disease to obtain a PPV above 90%. In study III, we report that the incidence for HCC in cirrhosis and the cumulative risk at five and ten years highly depends on sex, age, and type of liver disease, indicating that HCC-surveillance should be individually tailored. In study IV, the association of bacterial infections and a relatively preserved MAIT cell fraction is an interesting finding that needs to be investigated further.