Discussion:
CDF is a rare clinical entity, typically associated with chronic inflammatory pathologies, malignancies, and trauma. (5,6). It is well-established that chemotherapy for gastrointestinal neoplasms can be associated with complications such as haemorrhage and perforation, and, in some instances, fistula formation (2,5,6,7). A comprehensive review of the literature using the terms “choledochoduodenal fistula”, “pneumobilia”, “lymphoma”, “neoplasms” and “Non-Hodgkin’s Lymphoma” across the databases Ovid Medline, PubMed, Scopus and Google Scholar identified two documented cases in which chemotherapy treatment of duodenal lymphoma and methotrexate-induced lymphoproliferative disease led to CDF formation. However, no documented cases of CDF arising from chemotherapy for pancreatic lymphoma were identified (2,5).
The pathogenesis of fistula formation secondary to chemotherapy is thought to stem from the rapid regression in tumour and lymph node size, as well tumour tissue destruction (2,5). In cases of duodenal lymphoma, fistula development is postulated to arise from transmural invasion by the tumour, establishing a contiguous tissue bridge between the duodenum and CBD. As transmural invasion progresses, normal tissue is progressively replaced by the developing tumour. With chemotherapy, the tumour in the common wall established between the two organs becomes necrotic and is destroyed, resulting in fistula formation. (2) Similarly, in a patient with methotrexate-associated lymphoproliferative disorder, the development of their fistula was thought to be a consequence of the chemotherapy regimen of rituximab, brentuximab and nivolumab causing rapid regression of lymph nodes adjacent to both the duodenal wall and common bile duct, leading to CDF formation (5). In our case, the CDF likely developed from inflammation, necrosis, and the formation of a distal biliary stricture, following chemotherapy for pancreatic lymphoma.
The diagnosis of CDF can be challenging due to the non-specific nature of symptoms. In some instances, patients may present asymptomatically, further complicating early detection. In our case, the only symptoms reported were chronic lethargy and weight loss, with chronic LFT derangement on biochemical investigation. The suspicion of CDF development in this patient from his initial presentation was understandably low given the rarity of this condition and the non-specific clinical presentation. Suspicion of CDF was raised when pneumobilia was identified on CT, and a formal diagnosis of CDF was confirmed following ERCP. In a review article of CDF cases in China, most patients presented with epigastric pain (80.91%), cholangitis (54.26%) and fever (50.69%). In the same review, ERCP was determined to be the most reliable diagnostic modality, with 475 of 728 cases confirmed using this method, while ultrasound and CT were of limited utility for CDF diagnosis (6). As such, the non-specific nature and scarcity in documented cases of CDF, combined with the necessity for invasive procedures such as ERCP to establish a formal diagnosis, make clinical recognition and diagnosis very challenging.