Cholangiocarcinoma (CCA) is a relatively rare hepatobiliary malignancy, with an incidence of <1 case per 100,000 individuals, which accounts for 3% of all gastrointestinal malignancies worldwide (1,2). Known risk factors for the development of CCA include primary sclerosing cholangitis, choledochal cysts and hepatolithiasis, as well as liver flukes in Asian populations (2).

Patients with extra-hepatic cholangiocarcinoma often present with symptoms of biliary obstruction, including painless jaundice, pale stools, dark urine and pruritis. Intra-hepatic disease may present differently, often with less specific symptoms, such as malaise, abdominal pain and weight loss (3).

Commonly used diagnostic imaging modalities for CCA include computed tomography (CT) and magnetic resonance cholangio-pancreatography. Endoscopic ultrasound and endoscopic retrograde cholangio-pancreatography are useful for imaging distal extra-hepatic lesions, as these techniques allow stenting to relieve biliary obstruction and cytological brushings to be obtained (3,4). Diagnosis may also be supported by elevated levels of the tumor marker, carcinoembryonic antigen 19–9 (4).

Surgical resection remains the standard treatment for cholangiocarcinoma, however 5-year postoperative survival rates remain poor at 27–37% for patients with extra-hepatic disease and 23–42% for patients with intra-hepatic disease and clear surgical margins (3,4).

In the current study, two cases of CCA, in patients who had previously undergone external beam radiotherapy (EBRT) for the treatment of non-Hodgkin's lymphoma (NHL), are presented. The clinical data of the patients are presented and the medical records of the patients are retrospectively reviewed. The study details the clinical history, investigations and outcome of these two cases of this rarely encountered phenomenon, which may ultimately be considered an unusual but important variant of CCA. We hypothesize that the biliary epithelium is particularly sensitive to EBRT and review the current literature to investigate the proposed etiology of secondary malignancies following EBRT.

Modern cancer therapies have led to increases in the quality of life and prognosis of cancer patients, allowing for the observation of long-term sequelae of cancer therapies. The risk of secondary malignancies following delivery of EBRT has been identified previously in a number of studies (7–10). Known risk factors for this include a younger age at diagnosis, high radiation dose and being of the female gender (11).

In 1948, Cahan and Woodard (12) proposed certain classic criteria that were required for the diagnosis of secondary malignancies due to prior radiotherapy treatment. These criteria remain widely accepted as a conservative guide and stipulate that affected patients must possess the following: i) A prior history of radiation treatment; ii) an asymptomatic latent period of several years; iii) occurrence of a second malignancy within the previously irradiated field; and iv) histopathology of a secondary malignancy distinct from that of the primary malignancy (12).

Primary cancers that have been consistently associated with a high risk of EBRT-associated secondary malignant neoplasm (SMN), include breast cancer, prostate cancer and lymphoma (including Hodgkin's and non-Hodgkin's subtypes) (11). The most commonly described SMNs vary with regard to primary cancer site and radiation field. In patients treated for NHL, the most common SMNs include leukaemia and lung cancer (7–9,13) as well as melanoma in Australian populations (14).

At present, the association between ionizing radiation exposure and hepatobiliary malignancy remains unclear. Recently, Dores et al (15) reported an increased risk of pancreatic malignancy in patients treated with moderate to high doses of EBRT for Hodgkin's lymphoma. Furthermore, the study demonstrated that this risk was higher in patients treated with both EBRT and chemotherapy, which indicates a possible interaction between the two treatment modalities in contributing to SMN. Similarly, Morton et al (16) reported an increased risk of stomach cancer with EBRT, and showed that this risk was associated with the sub-diaphragmatic radiation dose.

Radiation exposure associated with hepatobiliary malignancy has previously been reported within the context of Thorotrast (17,18) and in survivors of atomic bomb fallout exposure in Nagasaki and Hiroshima (19,20).

In total, 5 cases of EBRT-associated CCA have been previously reported in the literature; 3 of these cases occurred 18–23 years after EBRT for urogenital carcinoma (21), and two cases occurred following abdominal EBRT for Hodgkin's lymphoma and teratocarcinoma, respectively (22). Radiation-induced liver disease is a known complication of EBRT (23), and there have also been reports of patients developing hepatocellular carcinoma (24) and benign biliary strictures (25,26) following EBRT.

The molecular mechanism of EBRT-induced biliary carcinogenesis remains unclear. However, it has been postulated that chronic inflammation secondary to biliary epithelial injury, and the accompanying disruption to bile flow, are important factors in the development of carcinogenesis (27–29). Persistent inflammation is hypothesized to damage DNA mismatch repair genes and tumor suppressor genes, thus promoting carcinogenesis (27). Cheng et al (23) previously suggested that hepatocytes may be more susceptible to radiation injury than the biliary epithelium, however, the susceptibility of these cell-types to radiation has not yet been reported.

In the present study, in the two cases reported, the histopathology was consistent with primary CCA. Considering the reported occurrence of CCA and other hepatobiliary complications (21–26,30,31), we hypothesize that chronic inflammatory reactions and resultant changes to the biliary epithelium following abdominal EBRT were fundamental to the development of CCA. However, it is unclear whether this represents a radiation-associated variant of the disease.

CCA continues to exhibit a poor prognosis. Thus, risk factors must be closely assessed to improve patient prognosis. Abdominal EBRT must be recognized as a risk factor of SMNs and thus, a heightened index of suspicion and more intensive surveillance of patients may facilitate an earlier diagnosis.