Choledocholithiasis and Cholangitis

A common bile duct stone and acute cholangitis are related findings with different evidential requirements. I can look for the stone on imaging, but recognizing an infected biliary system requires clinical observations that the image does not contain.

Clinical overview

Choledocholithiasis means a stone is present in the common bile duct. It may cause pain, jaundice, pancreatitis, or no obvious symptoms. Acute cholangitis is an infectious syndrome involving the biliary tract, commonly in the setting of impaired drainage, and can progress to organ dysfunction.

Fever, right upper quadrant pain, and jaundice form Charcot’s triad. Its absence does not exclude cholangitis. Waiting for every element would make a memorable clinical pattern into an inappropriate prerequisite for recognition.

I would distinguish the probability that a stone remains in the duct from the urgency of treating the patient. A small or unseen obstructing lesion can accompany serious illness. Conversely, a clearly demonstrated stone does not by itself establish systemic infection or its severity.

Anatomy and pathophysiology

Stone migration creates a changing anatomical problem

Common bile duct stones often migrate from the gallbladder through the cystic duct. Stones can also form within the biliary tree, particularly where stasis and infection alter the local environment. A patient’s gallbladder stone burden therefore does not specify whether a duct stone is currently present.

Obstruction can be partial, intermittent, or persistent. A stone near the ampulla can impede bile drainage and be associated with pancreatitis. If it subsequently passes into the duodenum, the anatomical finding changes while biochemical abnormalities or duct enlargement may persist.

I read this temporal behavior as central to reference construction. Ultrasound, MRCP, and ERCP performed on different days need not be observing the same stone state. Disagreement may reflect passage rather than an erroneous earlier examination.

Infection depends on drainage and host response

Impaired bile drainage facilitates infection, while rising intraductal pressure can promote movement of bacteria and inflammatory products into the systemic circulation. Stones are one cause of this process. Benign strictures, malignancy, and obstructed stents can produce the same clinical emergency.

Duct diameter is an anatomical response, not a direct measurement of bacterial burden or organ dysfunction. Similarly, echogenic duct contents may represent debris but do not independently prove infected bile. I would keep the imaging evidence of obstruction separate from the clinical evidence that establishes the infectious syndrome.

A patient can simultaneously have gallbladder inflammation, a duct stone, pancreatitis, and cholangitis. These processes should not be forced into mutually exclusive disease labels.

Diagnostic workflow and imaging findings

Recognize the syndrome and assess severity early

The Tokyo Guidelines 2018 diagnostic framework uses three domains: systemic inflammation, cholestasis, and imaging evidence of dilatation or an underlying cause. One systemic inflammatory item plus an item from either of the other domains supports suspected cholangitis. A definite diagnosis requires evidence from all three domains.

Systemic inflammation includes fever or inflammatory laboratory abnormalities. Cholestasis includes jaundice or abnormal liver tests. Imaging may demonstrate dilatation, a stone, a stricture, or another relevant cause. Thus, a nondilated duct does not automatically prevent the imaging domain from being satisfied when an obstructing cause is shown.

Severity assessment concerns physiological consequences. Organ dysfunction is central to severe disease, while the moderate category uses specified clinical and laboratory risk features. I would retain the actual observations and guideline version rather than substitute “large stone,” “very dilated duct,” or “underwent ERCP” for a severity grade.

Begin management while anatomical uncertainty is being resolved

Initial assessment includes vital signs, mental status, renal function, inflammatory tests, bilirubin and liver enzymes, and evidence of associated pancreatitis. Blood cultures can guide treatment when obtained without delaying necessary antibiotics.

The Tokyo initial-management guidance emphasizes starting urgent supportive treatment in unstable patients before a definitive diagnosis is complete. Mild disease may respond initially to medical treatment; moderate disease calls for early drainage, and severe disease requires organ support with drainage as soon as feasible after initial stabilization.

I interpret this as a parallel workflow: resuscitation and infection treatment proceed while the team establishes the drainage route. The examination’s value is to clarify an actionable obstruction, not to postpone treatment until every duct segment is visible.

Search for a stone without overinterpreting secondary signs

On transabdominal ultrasound, a duct stone may appear as an echogenic intraluminal focus with posterior shadowing. Small stones may not shadow clearly. The focus should be shown within the duct across planes because adjacent bowel gas, vascular interfaces, and prior intervention can complicate interpretation.

Upstream dilatation supports impaired drainage, but early or intermittent obstruction may leave the duct relatively small. The distal common bile duct can be obscured by bowel gas. A normal visible proximal segment therefore does not establish that the terminal duct is clear.

In a hypothetical examination, ultrasound shows gallbladder stones and a prominent extrahepatic duct, while the distal duct is unseen. This supports further investigation for choledocholithiasis but does not demonstrate a migrated stone. The report should distinguish the observed gallbladder stones from the suspected duct lesion.

Apply stone-risk criteria to the appropriate decision

The 2019 ASGE guideline identifies high-risk findings that prompt ERCP: a common bile duct stone demonstrated on imaging, ascending cholangitis, or total bilirubin greater than 4 mg/dL together with a dilated common bile duct. For that combined criterion, the guideline uses greater than 6 mm when the gallbladder is present.

The conjunction matters: bilirubin elevation alone is not the same high-risk criterion. Patients with intermediate predictors, such as abnormal liver tests or duct dilatation without a high-risk combination, can undergo EUS, MRCP, or appropriate intraoperative evaluation. These are adult stone-risk criteria, not a universal definition of obstruction or a cholangitis severity score.

EUS can investigate small distal stones near an endoscopic transducer, while MRCP provides a noninvasive ductal map. I would interpret a negative study against its technical quality, the remaining clinical suspicion, and possible interval stone passage.

Distinguish drainage from complete stone treatment

ERCP can decompress the biliary system and remove stones, but these are separable procedural objectives. A stable patient may undergo definitive clearance during the initial intervention. In an unstable patient, decompression alone may be the appropriate immediate goal.

The 2021 ASGE cholangitis guideline suggests endoscopic rather than percutaneous drainage and decompression within 48 hours. It also supports combining drainage with stone treatment unless the patient is too unstable for a more extensive procedure. I would not interpret that interval as a reason to wait in a deteriorating patient.

Differential diagnosis and management context

Acute cholecystitis can produce fever and right upper quadrant pain without infection of the main bile ducts. Pancreatitis can produce systemic inflammation and transient biliary biochemical abnormalities. Hepatitis, medication-related injury, and nonbiliary sepsis can also complicate interpretation.

Cholangitis without stones remains important. An obstructed stent or malignant stricture can provide the relevant anatomical cause. A stone-negative image therefore answers only one part of the differential.

Antimicrobials, resuscitation, and biliary drainage address different components of the illness. Successful stone extraction does not instantly reverse organ dysfunction, while temporary drainage does not necessarily complete treatment of the underlying cause. After stabilization, management must also address residual stones, strictures, or the gallbladder source where appropriate.

I would avoid using improvement after antibiotics as a standalone diagnostic reference. Clinical response can support a case assessment, but it does not uniquely identify the biliary tract as the source of infection.

Implications for medical AI

Build labels around the evidence available at prediction time

I would maintain separate labels for duct stone, obstruction, cholangitis, and severity. Stone references could include reader-confirmed imaging or extraction findings, with dates attached. Cholangitis adjudication would use a prespecified clinical framework and contemporaneous observations.

An image-only model evaluated against cholangitis may be missing fever, blood tests, hypotension, and mental status. I would first ask whether it is intended to detect an anatomical cause or contribute to a multimodal clinical assessment. These are different claims even if both models output a cholangitis-associated score.

If the same ultrasound report contributes to the reference diagnosis, agreement can partly reflect incorporation of the index examination into its own standard. I would record that dependence rather than describe the reference as fully independent.

Test timing and treatment shortcuts

A concrete leakage mechanism occurs when cholangitis cases contribute post-ERCP images containing a stent or pneumobilia, while controls contribute untreated examinations. The model could recognize intervention history instead of pre-treatment evidence relevant to the intended decision.

I would compare performance using only examinations acquired before drainage with performance in the full dataset. Antibiotic timing, prior sphincterotomy, and stent status would remain separate variables. Excluding post-treatment images addresses one pathway without assuming that every procedural feature is irrelevant in every future use case.

Another feasible audit would review patients with a suspected stone on ultrasound but no stone at later ERCP. An adjudicated timeline could separate likely passage, uncertain earlier findings, and inadequate visualization. Forcing every such patient into the negative class would erase clinically meaningful temporal uncertainty.

Evaluate the difficult discordant cases

I would specifically examine cholangitis with little duct dilatation, duct stones without infection, and infected strictures without stones. These groups test whether predictions collapse the clinical syndrome into one familiar image sign.

Within the stone task, sensitivity could be stratified by distal duct visibility and direct versus indirect evidence. Within a multimodal cholangitis task, I would compare image-only, clinical-only, and combined inputs using the same patient split and prediction time.

My immediate research question is whether adding ultrasound contributes useful evidence once contemporaneous clinical findings are available, and which visible findings account for that contribution. This is attackable with linked hepatobiliary examinations and clinical timelines; a plausible heatmap over an enlarged duct would not answer it.

References

• Kiriyama et al., Tokyo Guidelines 2018: diagnostic criteria and severity grading of acute cholangitis (with videos), Journal of Hepato-Biliary-Pancreatic Sciences 2018.
• Miura et al., Tokyo Guidelines 2018: initial management of acute biliary infection and flowchart for acute cholangitis, Journal of Hepato-Biliary-Pancreatic Sciences 2018.
• Buxbaum et al., ASGE guideline on the role of endoscopy in the evaluation and management of choledocholithiasis, Gastrointestinal Endoscopy 2019.
• Buxbaum et al., ASGE guideline on the management of cholangitis, Gastrointestinal Endoscopy 2021.