The Temporal Trends in Incidence of Intra and Extrahepatic Cholangiocarcinoma in the United States and the Impact of Klatskin Tumor Classification

Abstract

Background and Aims: Previous studies of cholangiocarcinoma have reported an increasing incidence of intrahepatic (ICC) and a decreasing incidence of extrahepatic (ECC) in the United States. Hilar cholangiocarcinomas (Klatskin tumors) are anatomically defined as ECC and could have affected these trends. In version 1 of the International Classification of Diseases for Oncology (ICD-O) (1973-1991), Klatskin tumors were not assigned a unique ICD-O code and may have been coded as either ICC or ECC. In ICD-O version 2 (19922000), they were assigned a unique histology code (8162/3) which was cross-referenced to ICC. ICD-O version 3 (2000-present) cross-referenced Klatskin tumors to either ICC or ECC. Thus, Klatskin tumors may have been misclassified as ICC under all versions of ICDO. We conducted this study to determine 1) the secular trends of ICC and ECC from 19922007 and 2) the potential impact of misclassification bias on these trends. Methods: Using data from the Surveillance, Epidemiology and End Results (SEER) program, age-adjusted incidence rates of ICC and ECC were calculated in two consecutive time periods (19922000, 2001-2007). Following definitions were used: ICC: topography codes C22.0 (Liver) and C22.1 (Intra-hepatic bile duct) and histology codes 8140, 8160, 8161, 8020, and 8010. ECC: topography code C24.0; histology codes 8010, 8020, 8041, 8070, 8140, 8144, 8160, 8161, 8260, 8310, 8480, 8490, and 8560. Klatskin: histology code 8162/3. We calculated the annual percent change (APC) in incidence rates, utilizing a weighted least squares method. Results: We identified 4297 patients with ICC and 2679 patients with ECC during 1992-2007. The age adjusted rates (per 100,000) for ICC excluding Klatskin tumors (i.e., true ICC rate) decreased from 1.5 during 1992-2000 to 1.4 in 2001-2007 with an APC of -0.3% (95% C.I.: 1.5, 0.8), whereas true ECC rates, including Klatskin tumors significantly increased from 0.8 to 1.1 during the same time periods with an APC of 3.3% (95% C.I.: 2.0, 4.5). The increased incidence of ECC was more pronounced in males (APC 3.6%, 95% C.I.: 2.0, 5.1), Asians and pacific islanders (APC 4.5%) and in the 65+ age group (APC 3.9%). During 1992-2000, 208/224 (93%) cases of Klatskin tumors were misclassified as ICC's. This proportion dropped to 60/144 (42%) in 2001-2007. However, even with misclassification the incidence rate for ICC (including Klatskin tumors) declined from 1.6 to 1.4 per 100,000 during the two time periods with an APC of -0.9% (95% C.I -2.1, 0.3). Conclusions: Contrary to recent data, the incidence rates of ICC in the US have remained stable during the period 1992-2007, whereas ECC rates have risen. Improved classification of Klatskin tumors may explain some but not all of these trends.