Key facts for practice and radiology board exams
The Todani classification sorts a choledochal cyst by where the bile duct is dilated. Type I is a fusiform or saccular cyst of the extrahepatic duct and is the commonest. Type II is a true diverticulum, type III a choledochocele inside the duodenal wall, type IV cysts of both the intrahepatic and extrahepatic ducts, and type V the intrahepatic ducts alone, which is Caroli disease.
- Most type I and type IV cysts carry an anomalous pancreaticobiliary junction, a union of the ducts outside the duodenal wall that leaves a long common channel. There is no official millimetre threshold: the channel measures about 14 mm on MRCP in maljunction against about 7 mm in a merely high confluence [5][6].
- A choledochal cyst follows bile duct signal and attenuation on every modality and does not enhance. Wall thickening, enhancement or a mural nodule means superadded infection or carcinoma, not a simple cyst.
- Types I and IV are treated by complete excision with hepaticojejunostomy. Internal drainage is obsolete: malignancy was still found in 16.4% of patients after internal drainage [8].
- Report the Todani type and the entity name. A choledochocele is managed endoscopically and Caroli disease by liver resection or transplantation, so the type number alone does not tell the surgeon what to do.
Todani classification of choledochal cysts
Todani and colleagues published the scheme in 1977 as a modification of the Alonso-Lej classification, adding the intrahepatic types that cholangiography had begun to show [1]. Five types are recognised. The subtype letters describe the shape and extent of the dilatation, not a different disease.
| Type | What is dilated | Share of cases [2] | Imaging clue | Standard operation |
|---|---|---|---|---|
| I (Ia, Ib, Ic) | Extrahepatic bile duct. Ia saccular, Ib focal segmental, Ic fusiform | 50 to 80% | Cystic structure at the porta continuous with the common bile duct and separate from the gallbladder | Complete cyst excision with Roux-en-Y hepaticojejunostomy |
| II | True diverticulum of the extrahepatic duct on a narrow neck | 2% | Saccular outpouching with a stalk; the rest of the duct is normal calibre | Diverticulectomy, duct preserved |
| III | Intraduodenal segment of the common bile duct, a choledochocele | 1.4 to 4.5% | Cystic mass inside the duodenal wall bulging into the second part, filling on MRCP | Endoscopic sphincterotomy or unroofing |
| IVA | Intrahepatic and extrahepatic ducts | 15 to 35% (IVA and IVB together) | Multiple communicating cysts; look for a stricture at the hilum | Extrahepatic excision with hepaticojejunostomy, plus hepatectomy for lobar disease |
| IVB | Multiple extrahepatic cysts only | part of the type IV share | Beaded extrahepatic duct with normal intrahepatic ducts | Extrahepatic excision with hepaticojejunostomy |
| V | Intrahepatic ducts only, that is Caroli disease | 20% | Saccular intrahepatic dilatation with the central dot sign | Segmental resection if localised, transplantation if diffuse |
Type I is often quoted higher than it should be. The reported range is 50 to 80% [2], and modern adult series sit inside it at 64 to 65% [8]. The one subtype distinction that changes the operation is IVA against IVB, because intrahepatic involvement may need a hepatectomy on top of the extrahepatic excision.
Why types III and V sit awkwardly in the Todani scheme
Types III and V are grouped with the rest because the duct is dilated, not because the disease is the same. Visser and colleagues called the standard classification confusing and misleading, argued that Caroli disease and the choledochocele are unrelated to choledochal cyst and should no longer be thought of as subtypes, and pointed out that even the split between type I and type IV is arbitrary [3]. The objection is practical, not academic.
- Choledochocele, type III. Many are lined by duodenal rather than biliary mucosa, it is typically not associated with a pancreaticobiliary maljunction, and it carries a lower risk of malignancy than the other types [4]. It is treated endoscopically.
- Caroli disease, type V. A ductal plate malformation on the autosomal recessive polycystic kidney disease spectrum, not a consequence of pancreatic reflux. The extrahepatic duct is normal, so it never gets an extrahepatic excision.
- Type II is a focal structural anomaly with a normal duct beyond it, which is why a diverticulectomy is enough.
So give the Todani type, because that is how the surgical literature is indexed, and then name the entity. Type V on its own reads as a choledochal cyst needing extrahepatic surgery when what you mean is Caroli disease.
Anomalous pancreaticobiliary junction and the long common channel
An anomalous pancreaticobiliary junction, also called pancreaticobiliary maljunction, is a union of the pancreatic duct and common bile duct outside the duodenal wall, beyond the reach of the sphincter of Oddi. Pancreatic juice refluxes freely into the bile duct, and the duct dilates. It is present in about 83% of adults with a type I cyst [9].
How long does the common channel have to be? The honest answer is that there is no official number. The 2013 criteria of the Japanese Study Group on Pancreaticobiliary Maljunction require an abnormally long common channel or an abnormal union to be shown on direct cholangiography, MRCP, endoscopic ultrasound or drip infusion cholangiography CT, and they state no millimetre threshold at all [5]. The widely repeated figures of 15 mm in adults and 5 mm in children are not in them. What has actually been measured is this:
| Common channel | Direct cholangiography | MRCP |
|---|---|---|
| Pancreaticobiliary maljunction, 184 patients | 16.2 plus or minus 6.9 mm | 13.9 plus or minus 6.2 mm |
| High confluence of pancreaticobiliary ducts, 22 patients | 7.7 plus or minus 1.5 mm | 6.6 plus or minus 1.4 mm |
| Best cut-off to separate the two | not applicable | 9 mm |
- MRCP reads the channel about 2 mm shorter than direct cholangiography, so do not reject the diagnosis on a borderline MRCP number. At or below 9 mm on MRCP, direct cholangiography is needed to confirm it [6].
- A short channel is not a get-out either. The criteria explicitly ask you to confirm, by direct cholangiography, that the papillary sphincter does not reach the junction when the channel is relatively short [5].
The point most often missed: a maljunction without a dilated duct is still a cancer risk, and the cancer is usually in the gallbladder. In the Japanese nationwide survey, biliary cancer occurred in 21.6% of adults with congenital biliary dilatation, of which 62.3% were gallbladder cancer and 32.1% bile duct cancer. In maljunction without biliary dilatation the cancer rate was higher still at 42.4%, and 88.1% of those cancers were in the gallbladder [7]. Prophylactic surgery is recommended once maljunction is established, although the same authors note that the strategy for maljunction without biliary dilatation remains controversial [7]. A normal calibre duct on your MRCP does not make the finding benign, and the gallbladder is the organ to scrutinise.
Imaging of choledochal cysts
MRCP is the test of choice. It shows the extent of the cyst, proves that it communicates with the biliary tree, and demonstrates the common channel in the same study. The single most useful rule is that a choledochal cyst follows bile duct signal and attenuation on every modality: anechoic on ultrasound, fluid attenuation on CT, T1 hypointense and T2 hyperintense, with no wall enhancement.
- Ultrasound. Anechoic cystic lesion continuous with the biliary tree and separate from the gallbladder. That separation is what distinguishes it from a duplicated or distended gallbladder.
- CT. Non-enhancing fluid attenuation structure at the porta contiguous with the duct; an intramural duodenal cyst in type III; multiple communicating cysts in types IV and V. CT is usually where an unsuspected cyst is first seen.
- MRCP. Map the proximal and distal extent, the hilar confluence, any stricture, the segments involved, and the length of the common channel. Hepatobiliary contrast agents or hepatobiliary scintigraphy prove communication when MRCP is equivocal.
- Red flags for carcinoma. Irregular or nodular wall thickening, mural enhancement, an enhancing intraluminal nodule, restricted diffusion, or a new stricture. Suspect cholangiocarcinoma, and look for the double duct sign if the pancreatic duct is dilated too.
What else looks like this
| Mimic | What separates it from a choledochal cyst |
|---|---|
| Obstructive biliary dilatation | The duct tapers to a cause at the distal end and the dilatation is proportionate; a type I cyst is disproportionately dilated with a normal duct beyond it |
| Duplicated or distended gallbladder | Continuous with the cystic duct, not with the common bile duct; trace the connection |
| Pancreatic pseudocyst | History of pancreatitis, no communication with the bile duct, and it displaces rather than replaces the duct |
| Cystic biliary atresia in infancy | A small hilar cyst with non-dilated intrahepatic ducts, an abnormal or absent gallbladder and no visible distal duct; a choledochal cyst is larger with a patent duct [10] |
| Hepatic or peribiliary cyst | No communication on MRCP or on hepatobiliary contrast imaging |
The complications are worth carrying as the 3Cs: calculi, carcinoma and cyst rupture with biliary peritonitis, to which cholangitis and pancreatitis can be added. The overall malignancy risk is usually quoted as 10 to 15% and rises steeply with age, from about 2.3% at 20 to 30 years to 75% at 70 to 80 years [2]. A systematic review of 1337 adults in the West found 10.9% [8], and a modern series of 100 adults with type I cysts found cholangiocarcinoma in 6%, five of those six having been diagnosed in childhood and never resected or only bypassed [9]. That age gradient is the whole argument for excising a cyst when it is found rather than watching it.
Frequently asked questions
References and further reading
- [1] Todani T, Watanabe Y, Narusue M, et al. Congenital bile duct cysts: Classification, operative procedures, and review of thirty-seven cases including cancer arising from choledochal cyst. Am J Surg. 1977;134(2):263-269. PMID: 889044.
- [2] Singham J, Yoshida EM, Scudamore CH. Choledochal cysts: part 1 of 3: classification and pathogenesis. Can J Surg. 2009;52(5):434-440. PMID: 19865581.
- [3] Visser BC, Suh I, Way LW, et al. Congenital choledochal cysts in adults. Arch Surg. 2004;139(8):855-860. PMID: 15302695.
- [4] Law R, Topazian M. Diagnosis and treatment of choledochoceles. Clin Gastroenterol Hepatol. 2014;12(2):196-203. PMID: 23660418.
- [5] Kamisawa T, Ando H, Hamada Y, et al. Diagnostic criteria for pancreaticobiliary maljunction 2013. J Hepatobiliary Pancreat Sci. 2014;21(3):159-161. PMID: 24307541.
- [6] Itokawa F, Kamisawa T, Nakano T, et al. Exploring the length of the common channel of pancreaticobiliary maljunction on magnetic resonance cholangiopancreatography. J Hepatobiliary Pancreat Sci. 2015;22(1):68-73. PMID: 25234051.
- [7] Kamisawa T, Kuruma S, Tabata T, et al. Pancreaticobiliary maljunction and biliary cancer. J Gastroenterol. 2015;50(3):273-279. PMID: 25404143.
- [8] Bloomfield GC, Nigam A, Calvo IG, et al. Characteristics and malignancy rates of adult patients diagnosed with choledochal cyst in the West: a systematic review. J Gastrointest Surg. 2024;28(1):77-87. PMID: 38353080.
- [9] Walia S, Saad M, Butler J, et al. Type I choledochal cysts in adults: natural history and implications for management. Surg Endosc. 2026;40(9):7827-7835. PMID: 42432235.
- [10] He F, Guan X, Yang B, Liu H. Key postnatal magnetic resonance characteristics for differentiating cystic biliary atresia from choledochal cyst. Eur Radiol. 2024;34(11):7471-7480. PMID: 38753194.
To attend live, join our Telegram group to get regular updates for these webinars:
More Radiology videos:
Image credits: I, Drriad / CC BY-SA
















