Chapter 53A. Perspective on Biliary Surgery

There is little question that laparoscopic surgery in general surgery has had the greatest impact in the treatment of cholelithiasis and cholecystitis with over 80% of cholecystectomies being performed laparoscopically and laparoscopic cholecystectomy being one of the most common general surgical procedures performed. Although laparoscopic cholecystectomy has led to more frequent surgical treatment of gallbladder disease and a reduction in postoperative pain and hospital stay, as well as earlier return of bowel function and full activity, and decreased cost, it has not led to a reduction in the incidence of common bile duct (CBD) injuries. The most recent data cite an estimated incidence of 0.4–1.3% bile duct injuries after laparoscopic cholecystectomy compared to 0.2% after open cholecystectomy with significant patient morbidity and a high risk of litigation.1,2 It has always seemed odd that, faced with a bile duct injury, no more than 15% of surgeons refer these patients to subspecialty hepatobiliary surgeons for repair.3 The risk of bile duct injury had been thought to be greater earlier in a surgeon's experience with a 1.7% incidence of bile duct injury in the first case and a 0.17% chance of bile duct injury in the 50th case.4 However, other studies have shown that there is no reduction in risk as surgeons do more cases.2,5

Anatomic variability is common in the biliary tree and methods to reduce the incidence of bile duct injury have included routine cholangiography, identifying the boundaries of Calot's triangle, and identifying the “critical view,”6 all appropriately emphasized by Auyang and Soper.1 However, it has been my practice to emphasize to the residents to “stay on the gallbladder, stay on the gall bladder, and stay on the gosh darn gallbladder” as the most effective way to avoid injury to either the CBD or right hepatic artery. This prevents transection of the CBD and, if adhered to as the gallbladder is dissected free of the gallbladder fossa adjacent to the porta hepatis medially, will prevent injury to a low-lying right or right posterior hepatic duct. If this is not easily accomplished due to acute cholecystitis or other abnormalities, consideration should be given to converting to an open procedure. This principle still holds true for an open cholecystectomy as well. Late conversion is often associated with bile duct injuries. Although cystic duct leaks, transection or clipping (partial or complete) of the CBD or common hepatic duct have been the most commonly described injuries; increasingly we are seeing injuries to the right hepatic duct or right posterior hepatic duct which may be difficult to characterize as they are not in communication with the common duct and may be missed on endoscopic retrograde cholangiopancreatography (ERCP).7 Only with percutaneous transhepatic cholangiography (PTC) or injection of a subhepatic drain, along with a high index of suspicion, will they be identified.

Choledocholithiasis

Because most patients with suspected choledocholithiasis undergo preoperative ERCP, there are fewer laparoscopic surgeons who have extensive experience or are comfortable with laparoscopic CBD exploration and stone extraction. In addition, as Yoo and Pappas8 point out, the success rate of ERCP in clearing CBD stones is 90–95% compared to a laparoscopic success rate of 70–90%. di Carlo and McFadden9 point out there are many adjunctive procedures that can be used to extract and break up stones endoscopically including the use of a Dormia basket, mechanical lithotripsy, and intraductal shock wave lithotripsy using a cholangioscope. The method of choice will depend on the expertise available and the specific patient circumstances.

However, because of the known complications of ERCP, particularly pancreatitis and cholangitis, and a relatively high incidence of negative ERCPs, it is important to try preoperatively to identify the presence of choledocholithiasis noninvasively which can best be done with magnetic resonance cholangiopancreatography (MRCP) or endoscopic ultrasound (EUS) with a high degree of specificity and sensitivity.9 Patients with a positive MRCP or EUS can then be treated appropriately with laparoscopic CBD exploration or ERCP, and the number of negative ERCPs reduced. ERCP or intraoperative cholangiography remain, however, the diagnostic gold standards and are still indicated in clinical situations where the suspicion for choledocholithiasis is high and the MRCP or EUS is negative. PTC can be used in patients in whom anatomic considerations preclude the performance of an ERCP.

Patients who are found to have stones following cholecystectomy are usually treated with ERCP; PTC or operative exploration may be indicated if the CBD cannot be accessed endoscopically. The need for surgical drainage procedures such as Roux-en-Y hepaticojejunostomy for recurrent CBD stones has been almost eliminated by the other noninvasive techniques for stone removal. Choledochoduodenostomy should be avoided if possible because of a higher incidence of obstruction, cholangitis, and anastomotic strictures. Patients with intrahepatic stones and strictures are best managed with surgical resection in order to rule out malignancy and because of a lower incidence of recurrent stones.10

Choledochal Cysts

Parikh and Lillemoe2 have emphasized the association between type I and type IV choledochal cysts and cholangiocarcinoma and the importance of complete cyst excision and reconstruction with a Roux-en-Y hepaticojejunostomy. The association with type II and type III (choledochoceles) is less clear, but adult patients still often present with jaundice, cholangitis, or pancreatitis and these cysts should also be resected. An important diagnostic tool in patients with choledochal cysts is biliary cytology, but the authors do not describe the potential screening value of the tumor-associated antigen CA19-9,11 that has also been shown to be useful in patients with primary sclerosing cholangitis who are at risk for developing cholangiocarcinoma. The fluorescence in situ hybridization (FISH) cytological test is a new screening tool that holds additional promise for early identification of cholangiocarcinoma.12 The authors note that unilobar Caroli's disease (type V) should be managed with hepatic resection because of the risk of cholangiocarcinoma and this has been pointed out to be true in children as well where they are often seen in association with type I or type IV choledochal cysts and concomitant cyst excision and lobectomy should be performed.

In the authors' discussion of bile leaks after laparoscopic cholecystectomy, they note that hepatobiliary iminodiacetic acid (HIDA) scans are not sensitive enough to identify the source of the leak. However, in a patient with a suspected bile duct injury, the HIDA scan can help in determining whether the patient should undergo an ERCP or PTC. If the HIDA scan shows the radionuclide passing into the duodenum, then an ERCP would be the appropriate initial diagnostic test, whereas if no radionuclide passes into the duodenum the most appropriate initial study would be a PTC since the ERCP would likely only show an obstructed, clipped, or transected duct. The role of injecting a subhepatic drain, placed to drain a biloma, in diagnosing the source of leak cannot be overemphasized if the PTC or ERCP do not show the source of the leak. Drain injection can demonstrate a transected right posterior duct that is not in communication with the CBD. Finally, the authors suggest for waiting 6–8 weeks after drainage to proceed with repair. My practice has been to wait about 2 weeks after drainage from above (PTC) and below (CT or ultrasound-guided drainage) before proceeding with Roux-en-Y hepaticojejunostomy which provides more than adequate time for the inflammatory process to abate and permits safe repair. I also leave the transhepatic catheter in place for 3 months, obtaining a cholangiogram at 5 days, 1 month, and 3 months.

The authors argue that the management of recurrent pyogenic cholangitis involves primarily treating biliary strictures with biliary drainage utilizing a Roux-en-Y hepaticojejunostomy. However, just as with percutaneous or endoscopic management, drainage procedures are associated with a high risk of recurrence and hepatic resection of unilobar disease is curative and is the treatment of choice.10

Gallbladder Carcinoma and Cholangiocarcinoma

Whang et al13 point out the important changes in the tumor, node, metastasis (TNM) staging system for gallbladder carcinoma that reflect data showing differences in survival in patients with N1 (metastasis to cystic duct, CBD, hepatic artery, and/or portal vein lymph nodes) or N2 (metastasis to periaortic, pericaval, superior mesenteric, and/or celiac artery lymph node) nodal disease and in patients with locally unresectable T4 malignancies. Incidentally found gallbladder carcinoma with invasion into the muscularis is not uncommon and requires gallbladder fossa (segment V) resection, portal lymph node dissection, and possible CBD resection with Roux-en-Y hepaticojejunostomy if the cystic duct margin is positive.

In the discussion of tumor markers the authors indicate that CA19-9 is not sensitive or specific enough to be applicable for screening or diagnosis in the general population, which is to a degree true, but in patients with risk factors for the development of cholangiocarcinoma, for example primary sclerosing cholangitis (PSC), it can be useful, especially when combined with ERCP, EUS, and cytology or the newer FISH assay for detecting early malignancy.12 In patients where CA19-9 has been elevated it can be followed as a useful marker for recurrence postresection. Sadly, imaging of cholangiocarcinoma is still not able to separate unresectable from resectable tumors reliably compared to imaging in pancreatic cancer and usually underestimates the degree of disease, which is why the resection rate is still so low. Preoperatively placed percutaneous transhepatic catheters can be helpful in determining the extent of involvement of the left and right hepatic ducts and guide the surgeon as to which liver lobe to resect to facilitate negative margins on the opposite side (>75% of hilar cholangiocarcinoma resections include a concomitant hepatic resection).14

Laparoscopic Procedures

Yoo and Pappas15 discuss the controversy around the utility of routine versus selective intraoperative cholangiography and suggest that it is probably not cost-effective for the number of significant common duct stones identified that need removal (most small stones will pass spontaneously) and whether it prevents common duct injuries remains controversial. I use the selective approach and perform intraoperative cholangiography if the anatomy is unusual or confusing or if there is a clinical indication (eg, elevated liver function tests [LFTs] or preoperative imaging that suggests common duct stones). Their review of laparoscopic procedures illustrates the many options available to surgeons and patients for laparoscopy in biliary surgery, but also points out the difficulties of some of these procedures and the need for advanced training and acquisition of the necessary skills to optimize patient care. These advanced procedures (eg, Roux-en-Y hepaticojejunostomy) still take much longer than their open counterparts and provide marginal patient benefit at this time, but this may change over time as surgeons, techniques, and equipment all improve.