Surgical Management of Perihilar Cholangiocarcinoma: The Bismuth-Corlette Classification in Practice

Dr. George asked to review the actual scan before a specific recommendation, but for a resectable tumour around 3 to 3.5 centimetres, his general advice was not to force a diagnosis upfront: resect with a clear margin and send the specimen for histopathology. At that size, liver function or parenchymal loss is unlikely to be a concern, so surgery can proceed before, rather than after, a biopsy.

Hepatocellular carcinomas develop on a cirrhotic liver, which is like a fertile field: removing one tumour by resection or ablation does not remove the underlying tendency of that liver to produce another. Because the diseased liver stays in place after resection or ablation, these patients carry a higher ongoing risk of new tumours. A transplant removes the whole diseased liver and replaces it with one that does not carry that risk, which is why upfront transplant can be the better option even when a tumour looks resectable.

The first check is whether the background liver is cirrhotic or, from vertical hepatitis B or C transmission, essentially normal; a normal liver allows extended resection with portal vein embolisation to grow the future remnant. On a cirrhotic background, the priority is ruling out disease outside the liver, then grading any portal vein invasion from VP1 (a segmental branch) to VP4 (the main portal vein). Patients with VP1 or VP2 involvement and no extrahepatic disease can still be offered transplant, upfront or after downstaging with TACE, sometimes combined with SBRT for a portal vein tumour thrombus, aiming for 12 weeks of stable disease.

Non-cirrhotic HCC is uncommon in Dr. George's own caseload, around 5 patients in every 100 he sees, but for exactly this group his unit is far more aggressive: extended resections, portal vein embolisation, and even ALPPS, a staged hepatectomy, are all options, because a normal liver can lose as much as 80 percent of its volume and still regenerate enough function from what remains.

HCC has a characteristic imaging signature, so biopsy is reserved for genuine diagnostic dilemmas. An arterially enhancing lesion with venous washout on a properly phased triphasic CT is treated as diagnostic in around 95 percent of cases; MRI is used when the CT is inconclusive. Around 40 percent of HCC patients have an elevated AFP, meaning 60 percent do not, so diagnosis relies on radiology rather than tumour markers.

A Type 1 tumour leaves the junction of the right and left hepatic ducts intact, so removing the bile duct alone may be enough. A Type 2 tumour involves that junction directly and, while often still resectable, may require some surrounding liver parenchyma to come out as well because these tumours spread along the duct's mucosal surface.

A Type 3A tumour extends into the main common hepatic duct, the hilum, the secondary radicals on the right side and part of the left duct, involvement too extensive for anything less than a major resection. The caudate lobe is removed in almost every case, because its bile duct is usually involved too.

When a right trisectionectomy is planned, the remaining left lateral section can be too small to safely support the patient afterwards. Blocking portal flow to the right liver through portal vein embolisation redirects blood flow to the left side, causing it to hypertrophy and creating an adequate future liver remnant before surgery.

Percutaneous transhepatic biliary drainage (PTBD) is used far more often than endoscopic nasobiliary drainage (ENBD) for patients who present with cholangitis, significant debilitation, or who need portal vein embolisation, since draining the liver segment that will be preserved brings down jaundice and helps that segment grow.