Introduction 

Cholangiocarcinoma, or CCA, is a type of cancer that grows in the bile ducts, the tubes that carry bile from the liver to the intestines. CCA is often deadly and aggressive because it stays silent and shows no symptoms in the beginning, with symptoms appearing only after the disease has advanced and become harder to treat. To overcome this challenge, advanced imaging techniques are essential for timely identification, staging and treatment of cholangiocarcinoma. These procedures not only provide necessary information related to tumour findings, size, blood vessel involvement, and possible spread to other parts but also help with treatment decisions and monitoring.^1,2

Understanding cholangiocarcinoma: types and symptoms

As per the site of origin, cholangiocarcinoma (CCA) is categorised into three main types, each with specific diagnostic methods and the AJCC staging system.^3,4

Intrahepatic cholangiocarcinoma (iCCA)

A cancer that is found inside the smaller bile ducts of the liver, which becomes a mass or a solid lump.^1,3 Symptoms associated with iCCA are vague stomach aches and weight loss, while jaundice can be found in around 15% of affected people, especially when the tumour grows towards the hilum (opening where the liver connects to the bile ducts).^3,5

Perihilar cholangiocarcinoma (pCCA)

It develops in the area where the smaller bile ducts join near the liver and continues till the cystic duct (which connects to the gallbladder).³ It grows by spreading around or within the ducts.¹ Symptoms such as jaundice, itching and changes in the colour of stools and urine appear later when the biliary drainage system gets blocked.⁵

Distal cholangiocarcinoma (dCCA)

Cancer can develop anywhere, starting from the cystic duct to the ampulla of Vater.³ This type often grows by spreading around or maturing inside bile ducts.¹ The symptoms are quite similar to those of pCCA.⁵

Conventional imaging for cholangiocarcinoma

Computed tomography (CT)

A CT is the most commonly employed technique for investigating and measuring how far the cancer has spread, including outside the liver. Oftentimes, CT is performed with a dye or contrast (CECT) for more detailed imaging. It is particularly helpful in finding whether the cancer involves lymph nodes or blood vessels, especially the portal vein.^1,6 For iCCA, CT scans detect the cancer approximately 70% of the time. Whereas a CT scan provides higher detection accuracy (about 79% to 92%) for cancers that have spread outside the liver and is generally more efficient in showing how far pCCA has spread, with about 92% accuracy.¹

Magnetic resonance imaging (MRI) and magnetic resonance cholangiopancreatography (MRCP)

When used together, MRI and MRCP provide better visualisation of affected areas such as the bile ducts, blood vessels, and liver tissue to help identify cancerous growth. This combination is excellent for checking the bile ducts both above and below a stricture and for detecting abnormal masses inside the liver.⁸ They generate high-quality images without being invasive or using radiation and can therefore replace the need for invasive studies.^2,4,7

Advanced magnetic resonance techniques

Diffusion-weighted imaging (DWI)

A special type of MRI scan that highlights tumours by detecting the restricted water movements between the tightly packed cells. The apparent diffusion coefficient (ADC), a measurement derived from DWI, helps in understanding tumour severity and the outcomes.⁹

Hepatobiliary-specific contrast agents

During MRI scans, specific contrast agents are used to brighten both the liver cells and bile ducts, making healthy liver tissue appear brighter than the abnormal cells, like cancer. One such agent is gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA), which makes it easier to see how much the tumour has spread and to find small nearby spots called satellite lesions by generating a highlighting effect on MRI images.^3,8

Nuclear medicine imaging

Fluorodeoxyglucose Positron Emission Tomography (FDG-PET) / PET-CT

PET scans are primarily useful in identifying advanced tumours that have spread from their origin, described as distant metastases. FDG is a sugary substance that is absorbed in excess by the cancer cells, causing tumours to light up on the images. This method is mainly reliable for diagnosing CCA, especially for the peripherally located tumours (pCCA), with 90% sensitivity and accuracy.⁸ 

Endoscopic imaging techniques

Endoscopic ultrasonography (EUS)

A minimally invasive procedure that connects a thin, flexible tube with an ultrasound camera, which is passed through the mouth into the stomach and the first part of the intestine, which is located near the liver and bile ducts. This technique is being utilised in diagnosing and staging CCA because of its ability to identify cancerous lymph nodes, as nodal involvement can change the staging, prognosis and treatment of the cancer. During EUS, small tissue samples are collected using fine needles (FNA or FNB) to confirm the diagnosis. Overall, its detection rate for CCA is about 94%, with even greater success for extrahepatic CCA (at 100%), compared to 83% for iCCA.³

Single-operator cholangioscopy (SOC)

A special procedure where doctors put a tiny camera to view directly inside the bile ducts, which allows them to have a closer look at tumours and other unusual changes, especially for cancer outside the liver (eCCA). This additional information helps guide whether the patient needs minimally invasive surgery, a different type of surgery or no surgery at all. SOC visual accuracy is confirmed by testing tissue samples post-surgery and is found to be as high as 93% for establishing treatment plans.¹⁰

Artificial intelligence (AI) and radiomics applications

AI provides powerful, rapid and less invasive methods to locate and diagnose CCA, leading to earlier treatment and better outcomes. It works by analysing samples from the body. For example, Raman spectroscopy uses light to find chemical changes in blood or tissue; when combined with AI methods like support vector machines (SVM) it can identify cancer with fair accuracy. Mass spectrometry is another technique that studies molecules in samples; when combined with AI tools like SVM and random forest (RF), it has proven almost 99% accurate in detecting cancer.⁵

Machine learning (ML) in imaging applications

A subset of AI that trains computers in learning from large data sets and finding key patterns. ML can quickly analyse the radiological findings and highlight the location of CCA, helping radiologists speed up the diagnostic process.⁵

Radiomics and texture analysis

Radiomics is an emerging technique that uses computers to analyse medical scans like MRI or CT and extract detailed information about tumours and tissues, including patterns, textures, and shapes. These details, called radiomics features, may not be visible to the human eye. 

In the diagnosis and staging of CCA, radiomics, when combined with AI, can help with tumour grading. For example, in iCCA, when radiomic features from the portal phase of preoperative CT scans (taken after injecting a dye) are assessed, especially those from both the tumour and the small rim of surrounding tissue, they can help doctors better predict the severity of cancer and if it has reached small blood vessels.¹¹

Molecular imaging and radiogenomics

Molecular imaging is a promising diagnostic tool that allows visualisation of tissues without any invasion or surgery. It also helps determine what’s happening on a much deeper level, such as inside cells and molecules. By understanding the biological changes, such as cell behaviours, doctors can diagnose cancer earlier and assess its pattern of growth and energy consumption.² 

Radiogenomics combines information from medical images with data about a tumour’s cellular, molecular and hereditary patterns. By connecting the imaging findings with what’s happening inside cells and genes, it becomes easier to classify liver tumours and predict their behaviour more precisely.¹¹

Emerging  molecular imaging techniques

Molecular imaging methods used in early research to study CCA include nuclear medicine imaging (PET and SPECT), MRI, optical imaging and multimodal imaging.²

• Nuclear medicine imaging: PET and single-photon emission computed tomography (SPECT) show the body’s functioning by tracking the energy from tiny amounts of radioactive substances called tracers²
• MR molecular imaging: a powerful method that combines molecular imaging with MRI to produce high-resolution 3D images of soft tissue without using radiation²
• Optical imaging: utilises light to provide visualisation of cancer cells without radiation and at a low cost. However, light can only reach shallow depths, so this method works best during endoscopy and during surgery²
• Interventional molecular imaging: an expansion of molecular imaging techniques that can find very small tumours, guide the delivery of imaging or therapeutic agents and make targeted therapies more effective – all in real time. For example, when a high-resolution MRI is used, it can track the delivery of a drug called motexafin gadolinium (MGd) in the bile duct wall through imaging²

Challenges and future directions

Despite the range of available techniques, confirming cholangiocarcinoma (CCA) is still challenging, especially when differentiating between harmless (benign) and cancerous (malignant) narrowing of the bile ducts or detecting small, spreading tumours. Further studies are needed to compare the potency of different imaging methods, especially newer techniques that visualise the bile ducts, as well as to create standard guidelines for identifying cancerous features on images.¹²

Combining emerging technologies such as radiomics with other detailed biological data about tumours, such as molecular, tissue and genetic information, is helping to introduce better models for classifying liver tumours more accurately.¹³ Working together with different techniques in sequence, such as starting with ultrasound, then advancing to CT and MRI/MRCP scans, and sometimes supplementing with PET/CT and cholangioscopy, is essential for diagnosing and staging cholangiocarcinoma. Continued advancements in molecular and enhanced imaging procedures may enable earlier detection and the design of personalised treatment plans for CCA.^1,2,7,13

Summary

Cholangiocarcinoma (CCA) is the second most common type of liver cancer, forming in the lining of the bile ducts. It often remains symptom-free in the initial stages and is therefore diagnosed at an advanced stage. Treatment options are very limited, making accurate preoperative staging necessary to determine whether the tumour can be resected or treated with a liver transplant. CCA imaging relies on a combination of imaging techniques. CT is a traditional method commonly used for initial workup and to check the involvement of blood vessels with accuracy. MRI, when combined with MRCP, provides excellent visualisation of bile ducts, blood vessels and liver tissues without radiation. 

Advanced techniques like diffusion-weighted imaging (DWI) and specialised contrast agents allow tumour identification and detection of small lesions. Nuclear medicine imaging, such as FDG-PET, detects tumours with increased metabolic activity and distant metastases. Endoscopic methods give minimally invasive options for determining cancerous lymph nodes and directly viewing bile duct tumours. Artificial intelligence, machine learning, and radiomics are emerging tools that analyse imaging and molecular data to enhance tumour detection, grading and prediction of cancer spread, revealing details that may be invisible to the human eye. Combinational use of radiomics, radiogenomics and AI systems may establish new methods for early diagnosis and personalised treatment strategies for conditions affecting the liver and biliary system.