Dr Milind Javle explored the intersection of artificial intelligence (AI) and clinical expertise in addressing the challenges of cholangiocarcinoma (CCA) care. His presentation began by recollecting an intriguing experiment he conducted in which he presented the case of a 62-year-old woman with suspected CCA to both a multidisciplinary clinical team and an advanced AI model. The patient presented with abdominal discomfort, elevated liver enzymes, and imaging showing a large liver mass with multiple lesions and a suspicious peripancreatic area. Initial biopsy and immunohistochemistry suggested adenocarcinoma, but the primary tumor origin remained unclear.
When the details of the case were entered into the AI model, the system identified the disease as metastatic periampullary adenocarcinoma and recommended standard treatment regimens such as FOLFIRINOX or FOLFOX. The AI response was rapid and projected confidence based on pattern recognition from the available clinical and imaging data.
In contrast, a multidisciplinary clinical team determined the diagnosis was more consistent with intrahepatic cholangiocarcinoma (after integrating imaging, pathology, and clinical findings). Based on that diagnosis, the team initiated treatment with gemcitabine, cisplatin, and durvalumab.
During the exercise, the stark difference in approach became evident. The AI model demonstrated remarkable efficiency and confidence in analyzing data and generating recommendations. However, as Dr Javle pointed out, AI lacked the nuanced understanding of clinical considerations and the ability to build a rapport with the patient, both of which are critical aspects of effective care.
As treatment progressed, clinical judgment continued to guide care. When the patient experienced side effects that affected daily functioning such as fatigue and mouth sores, AI recommended continuing therapy without adjustment. The clinical team, however, recognized the need to modify the treatment plan to improve the patient's quality of life. Subsequent molecular testing identified an FGFR2 fusion, prompting a transition to an FGFR-targeted therapy within a clinical trial.
This case highlights the complementary roles of AI and clinical expertise. AI offers speed and strong pattern recognition, but it is limited by the data provided and lacks the ability to incorporate nuanced clinical context, evolving patient factors, and real-world decision-making. On the other hand, clinicians synthesize complex information, personalize treatment, and adapt care over time.
Dr Javle emphasized that the future of CCA care lies not in replacing clinicians with AI, but in integrating both, leveraging AI to enhance decision-making while relying on clinical expertise to deliver individualized, patient-centered care.
Source: Javle M. Artificial intelligence vs clinical experience: who will lead the future of CCA care? Presented at: 2026 Annual Cholangiocarcinoma Foundation Conference. May 1-3, 2026; Salt Lake City, UT.
To sign up for our newsletter or print publications, please enter your contact information below.
