We read with interest the recent article by Ngu et al.1 that reports a population-based study examining mortality and the risk of hepatobiliary and nonhepatobiliary malignancy in patients with autoimmune liver disease. During the 7-year study period a total of 130 autoimmune hepatitis (AIH), 70 primary biliary cirrhosis (PBC), and 81 primary sclerosing cholangitis patients were identified.
Of those patients with AIH, three developed hepatocellular carcinoma (HCC), while 28 developed an extrahepatic malignancy. Although it is commented that cirrhosis was present in all AIH patients with HCC, no information on the severity of liver disease or fibrosis is given for the study cohort as a whole. We wish to draw the authors' attention to a study conducted by our institution, where 243 patients with AIH were identified and prospectively followed up between 1971 to 2007.2 In this cohort, 15 patients with AIH developed HCC, which equated to an annual incidence of 1.1%. The data demonstrated that cirrhosis was the sine qua non for the development of HCC in AIH, an association that is further supported by the data from Ngu et al. Furthermore, in their study, Ngu et al. report that the risk of HCC is 15-fold greater in AIH patients in comparison to that in a matched general population. However, as a result of the absence of information on liver disease severity it is not possible to determine the proportion of the cohort that were “at risk” of developing HCC. Indeed, the authors comment that the increased incidence of nonhepatic malignancy in the AIH study population may be the result of lead-time bias resulting from incidental diagnoses during HCC surveillance investigations without elaborating on the proportion undergoing surveillance.
Similarly, the influence of liver disease severity on HCC risk may also be reflected in the risk of malignancy observed in patients with PBC. No cases of HCC were reported within their cohort during the study period. As with AIH, cirrhosis is an established risk factor for HCC in PBC,3, 4 and without knowledge of the number of patients with PBC at risk of HCC it is not possible to determine whether the study population size is sufficient to permit firm conclusions as to the incidence of hepatic malignancy in PBC.
Although we commend Ngu et al. for a comprehensive study, we would have welcomed further detailed information on the severity of liver disease within their patient population, given its important and established influence on the risk of hepatobiliary malignancy in patients with autoimmune liver disease.