Laparoscopic
Transcystic Common Bile Duct Exploration John G. Hunter, MD, Salt Lake City. Utah
This study reviews the results of transcystic common bile duct exploration (CBDE) for unsuspected stones found during laparoscopic cholecystectomy by a single surgeon in 150 consecutive patients. Fluoroscopic cholangiograpby was attempted in all but four patients. if the cbolangiogram appeared to show common bile duct (CBD) stones, a 5 Fr, 8mm ureteral stone basket was passed through the cystic duct into the duodenum, opened, and trolled through the CBD. Routine cholangiography was successful in 131 of 144 attempts (90%). An indication for CBDE was found by cholangiogram in seven patients (5%). Two cholangiograms were falsely positive. Stones were removed in five patients. Completion cbolangiograms were normal in all patients. One patient developed mild pancreatitis but was discharged 2 days after laparoscopic cholecystectomy. The remainder were discharged on postoperative day 1. One patient was readmitted on postoperative day 2, possibly having passed a retained stone. Fluoroscopic CBDE was successful in clearing the CBD in all patients in this small series and deserves further evaluation.
From the Department of Surgery, University of Utah Medical Center, Salt Lake City, Utah. Requests for reprints should be addressed to John G. Hunter, MD, University of Utah Medical Center, Department of Surgery, 50 North Medical Drive, Salt Lake City, Utah 84132. Presented at the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orleans, Louisiana, May 20-22, 1991.
W
hile laparoscopic cholecystectomy has achieved consensus approval for the management of symp tomatic cholelithiasis, the management of bile duct stones in these patients is still a subject of debate. Most surgeons and gastroenterologists use endoscopic retrograde cholangiopancreatography (ERCP) to diagnose and remove common bile duct (CBD) stones in conjunction with laparoscopic cholecystectomy [1,2]. Many surgeons leave small stones in situ, hoping that they will pass spontaneously. A third conservative option is to convert to open cholecystectomy with common bile duct exploration (CBDE) when unsuspected stones are found. In order to clear the CBD of stones during the initial procedure and to avoid laparotomy, bile duct exploration through the cystic duct (transcystic CBDE) was attempted in all patients with characteristic filling defects on routine operative cholangiography. With this policy, we were able to clear the CBD in seven of seven patients with unsuspected cholangiographic filling defects. PATIENTS
AND METHODS
One hundred fifty patients had laparoscopic cholecystectomy performed at the University of Utah Medical Center between February 5, 1990, and December 10, 1990, by a single surgeon. Preoperative evaluation included an ultrasound of the gallbladder and bile ducts, a liver profile (alkaline phosphatase, lactate dehydrogenase [ LDH] , asparate aminotransferase [AST] , alanine aminotransferase [ALT], gamma glutamyl transpeptidase [GGTP], total bilirubin), and serum amylase determination. Patients with a dilated CBD (greater than 6 mm), elevated bilirubin, or two other hepatic enzymes twice normal were offered ERCP or open cholecystectomy and were excluded from this review. Laparoscopic cholecystectomy was performed with the four trocar technique described by Olsen [3]. After identifying the gallbladder-cystic duct junction and cleaning 1.5 cm of cystic duct, a cystic-duct cholangiogram was performed through the mid-s&c&al trocar with a 5 French whistle-tipped ureteral catheter (CR Bard, Covington, GA) and a specialized cholangiogram clamp (Karl Storz, Tuttlingen, Germany). Digital C-arm fluoroscopy (OEC Diasonics, Salt Lake City, UT) provided “real-time” imaging of the biliary tree. A coaxial cable to the radiology department allowed immediate review by a staff radiologist. A video record and hard copy spot films were made for later review. When a filling defect was discovered, the surgeon and radiologist determined whether the defect was most likely a stone or an air bubble. If the defect had the characteristics of a bubble, a number of maneuvers, including aspiration of the contrast and positioning the patient in steep reverse Trendelenburg, were performed to remove the
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
53
bubble or float it up into the liver where it generally broke apart in the smaller biliary radicals. When these maneuvers were unsuccessful or when a stone was suspected from the appearance of the first cholangiogram, transcystic CBDE was performed. CBDE was performed with digital fluoroscopic guidance. After cholangiography, the ureteral catheter was removed and replaced with a 4 French ureteral stone basket with tiliform tip (Baxter, V. Mueller, Deerfield, IL). The closed basket was passed well into the CBD, and the C-arm was repositioned (with the laparoscope left in the abdomen). Under fluoroscopy, the floppy tip of the basket was passed into the duodenum. The 8-mm helical basket was opened in the distal CBD and trolled back to the cystic duct orifice where it was closed and removed. Occasionally, it was necessary to crush the entrapped CBD stone with the basket so it could be removed through the cystic ductotomy. When it appeared that all stones had been removed, the cholangiogram was repeated and reviewed with the radiologist. The cystic duct was secured with a pre-tied ligature (Endoloop, Ethicon, Somerville, NJ). The cholecystectomy was completed. Prior to the removal of all trocars, a 10 French round closed suction drain (Zimmer, Dover, OH) was loaded into a 3.5-mm reducer and passed through the most lateral port. The drain was positioned between the liver and abdominal wall. The reducer and trocar were removed, leaving the drain behind. Drains were removed on the first postoperative day if they did not drain bile. A complete blood count, amylase, and liver profile were ob tamed on the first postoperative day and daily thereafter until nearly normal. Postoperative care was otherwise identical to that of patients undergoing laparoscopic cholecystectomy alone. RESULTS During the lo-month study period, 150 cholecystectomies were initiated with laparoscopic exposure. Two patients were denied laparoscopic cholecystectomy because of dilated bile ducts or frank jaundice. These patients underwent open cholecystectomy and CBDE. Cholangiography was attempted in 144 patients, and an adequate study was obtained in 13 1 (90%). Filling defects in the CBD that could not be proven to be air bubbles occurred in seven patients (5%). Stones were removed with irrigation and basketing in four patients. In one patient with a 5-mm CBD, passage of the stone basket pushed two 3-mm stones into the duodenum. In the remaining two patients, no stones were removed from the CBD despite several passes of the basket. Completion cholangiography was normal in all of these patients. Review of the preoperative evaluation of these seven patients showed few clues that would have predicted an abnormal cholangiogram. The CBD was of normal caliber in the six patients who had preoperative ultrasonography. Six patients had normal liver chemistry profiles. The seventh patient had mild elevations (less than twice normal) of the AST and alkaline phosphatase. No patient in this group had a history of fever or jaundice, and none were symptomatic at the time of surgery. 54
THE AMERICAN JOURNAL OF SURGERY
Postoperative liver profiles were obtained in live patients and were normal in one, mildly elevated in three, and showed frank hyperbilirubinemia (bilirubin: 4.6 mg/dL) in one. This last patient also had hyperamylasemia (450 IU/dL). The bilirubin was declining, and the amylase was normal 36 hours after surgery, so the patient was discharged. The other six patients were discharged the day after surgery. One of these six patients was readmitted 30 hours after discharge with nausea, vomiting, and mildly elevated alkaline phosphatase, LDH, GGTP, ALT, and AST. It was suspected that the patient had passed a retained CBD stone and was discharged the next day with symptoms resolved and laboratory abnormalities approaching normal. The findings at follow-up, 6 to 13 months (mean: 9.4 months) postoperatively, allowed a high degree of satisfaction with laparoscopic cholecystectomy and transcystic CBDE. No patient had recurrence of biliary symptoms, and no patient had required further diagnostic procedures. COMMENTS As laparoscopic cholecystectomy evolved, the general consensus was that the CBD stone was out of the reach of laparoscopic surgery. The experience in this series of seven patients suggests that most unsuspected CBD stones can be removed laparoscopically. Emboldened by this initial success with unsuspected CBD stones, we extended our preoperative indication to include patients with abnormal liver chemistries or a dilated CBD. As of December 1, 1991, laparoscopic transcystic CBDE had been performed in 20 patients. Choledochoscopy as well as completion cholangiography was performed in six of the last 10 patients but in only one of these patients was choledochoscopy necessary to retrieve a stone that could not be retrieved with fluoroscopic imaging. In three patients (including the two previously discussed), the cholangiogram was falsely positive (16%), but these patients were spared open CBDE or postoperative ERCP by a negative laparoscopic transcystic CBDE. In three patients (15%), transcystic CBDE was abandoned. In the first patient, recalcitrant valves of Heister prevented access to the CBD with catheters, baskets, and thin gauged guide wires. In the second patient, cholangiography demonstrated a dilated duct, packed with large stones and a small cystic duct. In the third patient, an impacted distal CBD stone could not be dislodged with balloons and baskets. Endoscopic lithotripsy was not available. In these three patients, CBDE was completed at laparotomy. Stones were removed from the remaining 14 patients with laparoscopic transcystic CBDE. If one considers all positive cholangiograms, the duct clearance rate in this series was 17 of 20 (85%). Including only patients with true positive cholangiograms, the duct clearance rate was 14 of 17 (82%). Of the 17 patients not converted to open CBDE, 13 were discharged on the first postoperative day, 2 were discharged on the second postoperative day, and 2 were discharged on the third postoperative day. Serious complications were rare. Hyperamylasemia was seen in two patients, hyperbilirubinemia in one pa-
VOLUME 163 JANUARY 1992
LAPAROSCOPIC BILE DUCT EXPLORATION
TABLE
I
Laparoscopic Common Bile Duct Exploration (CBDE)
Reference
No. of Laparoscopic Cholecystectomies
Positive Cholangiogram
[41 I51 I61 [71
500 516 342 270
29 35 26 19
Hunter+
252
20 (6)
(%)
(6) (7) (6) (7)
LC Clearance (%) 8 21 19 16
(26)* (60) (73) (64)
17 (65)
Converted to Laparotomy
Postoperative ERCP
3 6 5 2
16 5 5
3
0
1
“Laparoscopic CBDE was not attempted in early patrons of this series. Totals reflect resui?sof study through December 1991.
tient, and both conditions occurred in a fourth patient, but all had normalized within 48 hours. Presumably these patients developed edema or spasm of the sphincter of Oddi during stone extraction procedures that then led to jaundice or hyperamylasemia. In the last eight patients, glucagon has been administered at the time of duct exploration, and there have been no further postoperative enzyme elevations. One of these patients developed a subhepatic abscess related to a difficult cholecystectomy with spillage of infected bile. As previously mentioned, one patient probably passed a small retained stone not visible on completion cholangiography. These results compare favorably to a number of other series utilizing choledochoscopy as the primary imaging modality (Table I) [4-71. The method used most frequently for endoscopic transcystic CBDE requires dilatation of the cystic duct with a 5- to lo-mm balloon dilator (depending upon the size of the stone). A flexible fiberop tic endoscope 2.0 to 3.5 mm in diameter is inserted into the cystic duct and passed into the distal CBD or duodenum. Under direct vision, stones are basketed or grasped and pulled out of the cystic duct with the scope. If the stones are too large to be removed through the dilated cystic duct, they may be fractured with an electrohydraulit lithotriptor or pulsed dye laser. When cystic duct dilatation and choledochoscopy become the primary therapeutic response to a positive cholangiogram, duct clearance rates are 70% to 90% (Table I). The advantages of endoscopic CBDE are that it may be performed without the use of digital fluoroscopy, lithotripsy may be performed under direct vision, and choledochoscopy may result in fewer retained CBD stones than radiologic imaging alone [8]. The disadvantages of the endoscopic approach are that the cystic duct must be dilated in most patients to a ccommodate a therapeutic endoscope, that stones must be removed from the duct one by one, and that baskets small enough to be passed through the endoscope are too flimsy to crush all but the softest stones. In summary, both fluoroscopic and endoscopic imaging allow the removal of the majority of CBD stones. In patients with a tight tortuous cystic duct or a small CBD, the fluoroscopic technique allows stone retrieval without dilating any structure. Choledochoscopy may be best in a more dilated system where “blind basketing” is less likely to be successful. In addition, choledochoscopy is necesTHE AMERICAN
sary when laser or electrohydraulic lithotripsy is performed. Despite the very acceptable results of fluoroscopic and endoscopic CBDE presented here, this technique has not yet come of age. A survey of 8 large laparoscopic centers in Europe showed only 12 successes in 42 attempts (29%) [9]. The major reasons for failure of cystic duct CBDE are that the cystic duct is too small, the angle of access is incorrect, the CBD stones are too large or too numerous to be removed through the cystic duct, or the stones are intrahepatic and beyond the reach of scopes and baskets. The solution to these problems may be laparoscopic choledochotomy. Unfortunately, T-tube placement and secure sutured duct closure is technically demanding and is currently out of reach of the majority of laparoscopic surgeons. Additionally, we must keep in mind that CBDE is not adequate therapy for patients with primary CBD stones. These patients ought to have an endoscopic papillotomy or choledochoduodenostomy to achieve ductal drainage. For surgeons without the technology, training, or patience to perform laparoscopic transcystic CBDE, patients with elevations of liver chemistries or an ultrasonographically dilated CBD will continue to be referred for preoperative ERCP. How often will these noninvasive studies predict the finding of choledocholithiasis on ERCP? How often will the ERCP be necessary? Ultrasound fails to reveal choledocholithiasis in half of the patients proven to have CBD stones by cholangiography and falsely predicts the presence of stones in 50% of patients with normal ERCP [IOJI]. Hepatic chemistries are only slightly more accurate. In two series of more than 100 patients, the likelihood of finding CBD stones was 17% to 22% if one value on the liver profile (alkaline phosphatase, AST, LDH, or bilirubin) was elevated, 32% to 40% if two values were elevated, and 50% to 60% if three or more values were elevated [I2,13]. Thus, the yield of preoperative ERCP will be as low as 20% if performed for elevation of a single liver chemistry and will improve to only 50?&if the threshold is elevation of three values on the liver profile or a dilated CBD by ultrasound. In one large series of laparoscopic cholecystectomy, 64 patients were referred for preoperative ERCP, and only 10 (16%) had choledocholithiasis [2]. Eighty-four percent of preoperative ERCP were unnecessary. In addition to the limited yield of preoperative
JO1JRNAL OF SURGERY
VOLUME
163
JANUARY
1992
55
ERCP, the cost and discomfort level are significantly greater to the patient than removing CBD stones under general anesthesia at the time of laparoscopic cholecystectomy. More importantly, there is a legitimate concern that the morbidity of endoscopic papillotomy and laparoscopic cholecystectomy might be higher than the morbidity of laparoscopic cholecystectomy converted to open cholecystectomy when irretrievable stones are found. No data comparing these two approaches currently exist, but it was recently demonstrated that morbidity is doubled when endoscopic papillotomy and open cholecystectomy are performed rather than open cholecystectomy and CBDE [Z4]. The recommendation that ERCP should be done before laparoscopic cholecystectomy in all patients with elevated liver chemistries and that endoscopic papillotomy should be performed when stones are found needs further prospective evaluation. While elevated liver chemistries are only moderately accurate predictors of choledocholithiasis (specificity: 15% to 50%), normal liver chemistries are reported to be extremely accurate predictors of the absence of stones (specificity: 100%) [ I2,Z 31. This was not our experience, nor has it been the experience of other surgeons. In the initial series of 150 patients, 4 of 133 patients (3%) undergoing successful cholangiography had choledocholithiasis despite normal levels of alkaline phosphatase, AST, LDH, and bilirubin. In another series of laparoscopic CBDE, 7 of 14 patients (50%) with choledocholithiasis had normal liver chemistries preoperatively (Sangster, personal communication). These truly unsuspected stones can only be detected by the routine use of operative cholangiography. The management of the unsuspected tilling defect in centers without the capability to perform transcystic CBDE is controversial. In most centers, such patients are referred for postoperative endoscopic papillotomy. While such a practice preserves the “minimally invasive” ap preach, it has certain risks: Papillotomy fails in 5% to 20% of patients (depending on the experience of the endoscopist). Failure may require return to the operating room for an open choledochotomy and CBDE. The morbidity of endoscopic papillotomy is loo/o,and 1% of these complications require operative management (for bleeding or perforation) [15]. Lastly, postoperative ERCP is expensive, uncomfortable, and requires an extra day of hospitalization after laparoscopic cholecystectomy. Another approach advocated is to leave small stones to pass spontaneously. In our community and in Seattle (W. Traverso, personal communication), many cases of pancreatitis have resulted from these small “probably insignificant” bile duct stones left to pass on their own after laparoscopic cholecystectomy. For all these reasons, we have elected to perform open cholecystectomy, CBDE, and choledochoscopy in patients failing transcystic CBDE. While the pain, hospitalization, and rehabilitation may be greater with this approach, the success of duct clearance is 98% [8], and the morbidity may be less than with cholecystectomy and endoscopic papillotomy 1141. The evolution of technique and development of lapa56
THE AMERICAN JOURNAL OF SURGERY
roscopic bile duct exploration skills should markedly reduce the need for preoperative and postoperative ERCP. In this series of 150 laparoscopic cholecystectomies, only 2 patients underwent ERCP; 1 patient who presented with severe cholangitis underwent ERCP preoperatively, and 1 patient underwent ERCP postoperatively after developing cholangitis 3 days after laparoscopic cholecystectomy during which cannulation of the cystic duct had failed. Neither patient had choledocholithiasis at the time of ERCP, and neither patient was included in this review, as there was no indication for transcystic CBDE in either case. A review of Table I demonstrates that, when an aggressive laparoscopic approach to choledocholithiasis is taken, bile duct clearance is most often complete and the use of ERCP diminishes [4-71.
REFERENCES 1. Cotton PB, Baillie J, Pappas TN, Meyers WS. Laparoscopic cholecystectomy and the biliary endoscopist. Gastrointest Endcec 1991; 37: 94-6. 2. Meyers WC, et al. (Southern Surgeons Club). A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 1991; 324: 1073-8. 3. Olsen DO. Laparoscopic cholecystectomy. Am J Surg 1991; 161: 339-44. 4. Spaw AT, Reddick EJ, Olsen DO. Laparoscopic laser cholecystectomy: analysis of 500 procedures. Surg Laparosc Endosc 1991; 1: 2-7. 5. Sackier J, Berci G, Phillips E, Carroll B, Shapiro S, Paz-Partlow M. The role of cholangiography in laparcscopic cholecystectomy. Arch Surg 1991; 126: 1021-6. 6. Petelin JB. Laparoscopic approach to common duct pathology. Surg Laparosc Endosc 1991; 1: 33-41. 7. Swanstrom L, Sangster W. Laparoscopic management of common bile duct stones. Presented at the British Columbia Surgical Society, Vancouver BC. May 10, 1991. 8. Escat J, Fourtanier G, Maigne C, Vaislic C, Fournier D, Prevost F. Choledochoscopy in common bile duct surgery for choledocholithiasis: a must. Am Surg 1985; 51: 166-7. 9. Cuschieri A. Results of transcystic common bile duct exploration: the European experience. Presented at the Royal College of Surgeons, London, March 16, 1991. 10. Cronan JJ. US diagnosis of choledocholithiasis: a reappraisal. Radiology 1986; 161: 133-4. 11. Gross BH, Hatter LP, Gore RM, et al. Ultrasonic evaluation of common bile duct stones: prospective comparison with endoscopic retrograde cholangiopancreatography. Radiology 1983; 146: 471-4. 12. Del Santo P, Kazarian KK, Rogers JF, Bevins PA, Hall JR. Prediction of operative cholangiography in patients undergoing elective cholecystectomy with routine liver function chemistries. Surgery 1985; 98: 7-11. 13. Pemthaler H, Sandbichler P, Schmid Th, Margreiter R. Operative cholangiography in elective cholecystectomy. Br J Surg 1990; 77: 399-400. 14. Neoptolemos JP, Carr-Locke DL, Fossard DP. Prospective random&d study of preoperative endoscopic sphincterotomy versus surgery alone for common bile duct stones. BMJ 1987; 294: 470-4. 15. Cotton PB, Lehman G, Vennes J, et cl. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991; 37: 383-93.
VOLUME 163 JANUARY 1992
LAPAROSCOPIC BILE DUCT EXPLORATION
that when you manipulate the cystic duct, it is important (Jackson, MS): Transcystic fluoro- to consider the means to ligate it at the completion of the scopic basketing of bile duct stones is reliable, and we procedure. Those cystic ducts will be enlarged and friahave used it in 12 or 13 patients. I do emphasize that the ble, and it is important to ligate them with a ligature and dynamic studies obtained with C-arm fluoroscopy offer not just with a clip, because the clip often will not hold. substantially more information than the static films and Finally, do you have to use pharmacologic manipulation obviate the need for expensive radiolucent disposable can- during your cholangiograms? We do not have the luxury nulae (at a cost savings of $425 to $750). We have been of having the fluoroscopic unit that you described. We use quite selective in obtaining operative cholangiograms in static films, and it truly is shooting a “blind” x-ray, with trocars in the way and waiting 25 minutes until they are our series, which now exceeds 800 patients. Forty-nine percent of our patients were judged to be at no risk for developed. So we have performed cholangiograms selechaving CBD stones (common duct diameter less than 5 tively and often use pharmacologic manipulation to obtain optimal films. If the roentgenogram shows only flow mm, normal liver function tests, absence of jaundice, pancreatitis, or acute cholecystitis). None of these pa- distally, we administer intravenous morphine to cause tients underwent operative cholangiograms, and none spasm of the sphincter of Oddi. Conversely, with no flow have had evidence of a retained stone to date. The sensi- into the duodenum, an ampule of intravenous glucagon tivity and specificity of our clinical determination appear may result in relaxation of the sphincter. Thomas R. Gadacz (Baltimore, MD): The concento exceed those obtained by “routine” or “discovery” cholangiography. Furthermore, some of the stones that tration of contrast that you used was 60%. Why wasn’t a you identified may have passed spontaneously (clinical lower concentration used if you wanted to detect these false positives). The combined cost-savings of electrosur- small stones? How much time does the cholangiogram gery (versus laser), reusable instruments, and selective add to your procedure, and how much time does the operative cholangiography are substantial. In a recent cholangiogram plus basket removal add? Michael S. Nussbaum (Cincinnati, OH): Have you audit of 50 patients under age 60, the mean hospital charge for laparoscopic cholecystectomy generated by been able to utilize this technique in handling stones that our approach was only $1,910. What was the charge for are proximal to the cystic-common duct junction? That the procedure at your hospital? What is your policy about seems to be a problem with this technique in that you can only direct the basket down into the CBD but not up the disposable versus nondisposable instruments? common hepatic duct towards the bifurcation. John Sonneland (Spokane, WA): Would you comGerald Larson (Louisville, KY): I commend you for ment on the risk of injury to the cystic duct, when dilated in order to allow the removal of CBD stones by basket? your high success rate in exploring the CBD. I think we Would you comment on an alternative method of han- need to develop our technique so that we can explore the duct as successfully and reliably as we do with open dling CBD stones, namely fragmentation by laser? Greg V. Stiegmann (Denver, CO): I noticed in your cholecystectomy. If you have a patient in the preoperative series that none of these patients appeared to have a high period with a dilated CBD and elevated liver function likelihood of the presence of choledocholithiasis prior to tests and a very high probability of having a CBD stone, your laparoscopic operation. How did you find such a would you do a preoperative ERCP and sphincterotomy to clear the duct followed by laparcscopic cholecystectoselected group, or did you perhaps perform prelaparcscopit cholecystectomy retrograde cholangiography on a my? Or, would you proceed directly to the laparoscopic number of these patients prior to your operation? Of your cystectomy assuming that you can remove the stone inseven patients, one seemed to have had elevated amylase traoperatively? I am concerned that perhaps we are being levels afterwards, if not some clinical pancreatitis. That, too liberal in our indications for preoperative sphincteroof course, is one of the major complications of retrograde tomy in young patients with a long life expectancy ahead. cholangiography, sphincterotomy, and so forth. Do you Aaron S. Fink (Cincinnati, OH): I think you introbelieve that with transpapillary manipulation you per- duce another important point when you raise the issue of haps expose these patients to a risk of pancreatitis, and if preoperative versus postoperative ERCP. I think the indiso, can the technique be modified to obviate that risk? cated failure rates of 5% to 20% reflect the fact that both Nathaniel Soper (St. Louis, MO): Some of those the success rate and the complication rate will vary with common duct tilling defects looked so small that either the expertise available in your local institution. I think they would have passed spontaneously or perhaps they that this is a critical factor that must be used when chooswere air bubbles rather than stones. I question the need ing the modality to use in dealing with stones suspected for chasing all of the filling defects that we see, particu- preoperatively or discovered either intraoperatively or larly if there are complications of the basket extraction postoperatively. As we consider these facts, I think it itself. Should we be a little more selective in terms of further urges us to develop more reliable laparoscopic when we perform cholangiogram? If you see small filling techniques, such as you’re beginning to describe. defects, when should these be left alone? My second John G. Hunter (closing): Dr. Voyles, I am glad you question deals with the issue of CBD drainage. Do you did not ask me to address the routine versus selective think it should be used in these patients? If so, how would cholangiography issue as such a discussion could go on for you go about holding the T-tube in place, or would you days. I do not know the exact hospital cost for this proceput it through the cystic duct? My third point would be dure, but I can tell you that our mean hospital charges DISCUSSION C. Randle Voyles
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
163
JANUARY
1992
57
were $3,900 with routine cholangiography. This charge is $900 less than the charge for open cholecystectomy and a 3-day hospital stay in the year before I started performing laparoscopic cholecystectomy. We use mostly reusable trocars to save expense, but it is difficult to keep them sharp and I believe the safety shield may offer some small advantage. Dr. Sonneland, cystic duct dilatation is necessary in 45% to 50% of patients who are undergoing choledochoscopy because the smallest adequate choledochoscope that one can use laparoscopically is close to 3 mm in diameter and unless the passage of stones has dilated the cystic duct, it is necessary to use a balloon or rigid dilator. We have not used cystic duct dilatation routinely. We have only been doing completion choledochoscopy after fluoroscopic stone extraction when we could pass the scope readily through the undilated cystic duct. I am a little concerned that dilatation might rupture or excessively thin out the cystic duct and increase the risk of a postoperative bile leak, when the gain of choledochoscopy is not yet clear. Dr. Stiegmann, I have performed a preoperative ERCP in only one patient in 1% years. While I enjoy performing this procedure, I have been tempered by our inability to accurately predict who will have stones preoperatively and by the additional morbidity of adding endoscopic sphincterotomy to cholecystectomy. In this early experience, I refused to perform laparoscopic cholecystectomy on two patients for ultrasonographically visible CBD stones. All the patients reported in this series had truly unsuspected stones. A completion cholangiogram of the patient who developed hyperamylasemia showed no tilling of the duodenum. Such is the case after open bile duct exploration 20% of the time, and rarely is this related to a retained stone. I suspect my patient had some spasm or edema of the ampulla, which contributed to the pancreatitis and jaundice. This begs the question as to whether we should be draining the CBD when we cannot demonstrate duodenal filling on the completion cholangiogram. Dr. Soper, if I could identify stones that were going to pass spontaneously and not cause pancreatitis or cholangitis in passing, I think it would be reasonable to leave them behind. However, I have much anecdotal evidence from around our community and elsewhere that small stones left behind in the hope that they will pass are causing significant pancreatitis and cholangitis. I think the main reason for doing routine cholangiography is not to diagnose choledocholithiasis but to identify the anatomy of the region and to become proficient in the technique. The impact of a false negative laparoscopic transcystic CBDE is much less than doing an unnecessary open bile duct exploration. There are no tubes coming out of the patient’s side, and none of the patients with negative duct explorations have yet had any sequelae of their
58
THE AMERICAN JOURNAL OF SURGERY
exploration. I believe bile duct drainage with a small caliber latex drain, secured with a pre-tied catgut ligature (Endoloop, Ethicon, Somerville, NJ) , might be appropriate for decompression, imaging, and possibly stone extraction in patients with nonfilling of the duodenum on completion cholangiography. Using fluoroscopy, we have been able to obtain a complete cholangiogram in every patient without pharmacologic manipulation, Sometimes it is necessary to put the patients head down to fill the intrahepatic ducts, and often we need to rotate the head of the C-arm in order to avoid the overlying shadow of metal trocars. In all patients where cystic duct cannulation has been successful, dynamic fluoroscopic cholangiography has provided an adequate study. Dr. Gadacz, our radiologist asked me to start using 60% Conray in order that clear ductal definition would be provided despite the pneumoperitoneum and fluoroscopic imaging. If one starts injection slowly and watches the duct fill, there will be sufficient dilution of contrast by bile so small stones are seen. On the other hand, the patient who may have had a retained stone because of excessive opacification of the bile duct makes a good argument for using a more dilute contrast agent. Initially, the cholangiogram took 20 minutes but now takes 5 to 10 minutes. Initially, duct exploration took 60 to 90 minutes, because we were somewhat disorganized and had difficulty lccating baskets and guide wires. We now have it fairly streamlined so that it takes from 20 to 40 minutes to perform this procedure. Dr. Nussbaum, most of the patients in this series had a few small stones, and, with the patient in a steep reverse Trendelenburg, most stones were located in the distal CBD. In this series, we did not have a filling defect proximal to the cystic duct. With guide wires, it may be possible to get into the proximal bile duct but such access will be limited by cystic duct anatomy. Choledochotomy may be required to remove many of these stones. Dr. Larson, I think preoperative ERCP with sphincterotomy is being over-utilized. However, I think that a patient with a markedly dilated duct who is jaundiced may have a periampullary cancer. If the ERCP demonstrates choledolithiasis, this patient will most likely need a duct drainage procedure that cannot be provided laparostopically. An ERCP with sphincterotomy may provide that necessary duct drainage and allow the surgeon to proceed with laparoscopic cholecystectomy. In nonjaundiced patients with a biochemical indication of CBD stones but no ductal dilatation, I have proceeded with laparoscopic cholecystectomy as the initial procedure. The patient is aware that a failure of laparoscopy will result in conversion to open bile duct exploration. Dr. Fink, there is no doubt that the use of perioperative ERCP will vary by community, but I believe many surgeons will be able to remove the majority of CBD stones at the time of laparoscopic cholecystectomy.
VOLUME 163 JANUARY 1992
Transcystic Common Bile Duct Exploration John G. Hunter, MD, Salt Lake City. Utah
This study reviews the results of transcystic common bile duct exploration (CBDE) for unsuspected stones found during laparoscopic cholecystectomy by a single surgeon in 150 consecutive patients. Fluoroscopic cholangiograpby was attempted in all but four patients. if the cbolangiogram appeared to show common bile duct (CBD) stones, a 5 Fr, 8mm ureteral stone basket was passed through the cystic duct into the duodenum, opened, and trolled through the CBD. Routine cholangiography was successful in 131 of 144 attempts (90%). An indication for CBDE was found by cholangiogram in seven patients (5%). Two cholangiograms were falsely positive. Stones were removed in five patients. Completion cbolangiograms were normal in all patients. One patient developed mild pancreatitis but was discharged 2 days after laparoscopic cholecystectomy. The remainder were discharged on postoperative day 1. One patient was readmitted on postoperative day 2, possibly having passed a retained stone. Fluoroscopic CBDE was successful in clearing the CBD in all patients in this small series and deserves further evaluation.
From the Department of Surgery, University of Utah Medical Center, Salt Lake City, Utah. Requests for reprints should be addressed to John G. Hunter, MD, University of Utah Medical Center, Department of Surgery, 50 North Medical Drive, Salt Lake City, Utah 84132. Presented at the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orleans, Louisiana, May 20-22, 1991.
W
hile laparoscopic cholecystectomy has achieved consensus approval for the management of symp tomatic cholelithiasis, the management of bile duct stones in these patients is still a subject of debate. Most surgeons and gastroenterologists use endoscopic retrograde cholangiopancreatography (ERCP) to diagnose and remove common bile duct (CBD) stones in conjunction with laparoscopic cholecystectomy [1,2]. Many surgeons leave small stones in situ, hoping that they will pass spontaneously. A third conservative option is to convert to open cholecystectomy with common bile duct exploration (CBDE) when unsuspected stones are found. In order to clear the CBD of stones during the initial procedure and to avoid laparotomy, bile duct exploration through the cystic duct (transcystic CBDE) was attempted in all patients with characteristic filling defects on routine operative cholangiography. With this policy, we were able to clear the CBD in seven of seven patients with unsuspected cholangiographic filling defects. PATIENTS
AND METHODS
One hundred fifty patients had laparoscopic cholecystectomy performed at the University of Utah Medical Center between February 5, 1990, and December 10, 1990, by a single surgeon. Preoperative evaluation included an ultrasound of the gallbladder and bile ducts, a liver profile (alkaline phosphatase, lactate dehydrogenase [ LDH] , asparate aminotransferase [AST] , alanine aminotransferase [ALT], gamma glutamyl transpeptidase [GGTP], total bilirubin), and serum amylase determination. Patients with a dilated CBD (greater than 6 mm), elevated bilirubin, or two other hepatic enzymes twice normal were offered ERCP or open cholecystectomy and were excluded from this review. Laparoscopic cholecystectomy was performed with the four trocar technique described by Olsen [3]. After identifying the gallbladder-cystic duct junction and cleaning 1.5 cm of cystic duct, a cystic-duct cholangiogram was performed through the mid-s&c&al trocar with a 5 French whistle-tipped ureteral catheter (CR Bard, Covington, GA) and a specialized cholangiogram clamp (Karl Storz, Tuttlingen, Germany). Digital C-arm fluoroscopy (OEC Diasonics, Salt Lake City, UT) provided “real-time” imaging of the biliary tree. A coaxial cable to the radiology department allowed immediate review by a staff radiologist. A video record and hard copy spot films were made for later review. When a filling defect was discovered, the surgeon and radiologist determined whether the defect was most likely a stone or an air bubble. If the defect had the characteristics of a bubble, a number of maneuvers, including aspiration of the contrast and positioning the patient in steep reverse Trendelenburg, were performed to remove the
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
53
bubble or float it up into the liver where it generally broke apart in the smaller biliary radicals. When these maneuvers were unsuccessful or when a stone was suspected from the appearance of the first cholangiogram, transcystic CBDE was performed. CBDE was performed with digital fluoroscopic guidance. After cholangiography, the ureteral catheter was removed and replaced with a 4 French ureteral stone basket with tiliform tip (Baxter, V. Mueller, Deerfield, IL). The closed basket was passed well into the CBD, and the C-arm was repositioned (with the laparoscope left in the abdomen). Under fluoroscopy, the floppy tip of the basket was passed into the duodenum. The 8-mm helical basket was opened in the distal CBD and trolled back to the cystic duct orifice where it was closed and removed. Occasionally, it was necessary to crush the entrapped CBD stone with the basket so it could be removed through the cystic ductotomy. When it appeared that all stones had been removed, the cholangiogram was repeated and reviewed with the radiologist. The cystic duct was secured with a pre-tied ligature (Endoloop, Ethicon, Somerville, NJ). The cholecystectomy was completed. Prior to the removal of all trocars, a 10 French round closed suction drain (Zimmer, Dover, OH) was loaded into a 3.5-mm reducer and passed through the most lateral port. The drain was positioned between the liver and abdominal wall. The reducer and trocar were removed, leaving the drain behind. Drains were removed on the first postoperative day if they did not drain bile. A complete blood count, amylase, and liver profile were ob tamed on the first postoperative day and daily thereafter until nearly normal. Postoperative care was otherwise identical to that of patients undergoing laparoscopic cholecystectomy alone. RESULTS During the lo-month study period, 150 cholecystectomies were initiated with laparoscopic exposure. Two patients were denied laparoscopic cholecystectomy because of dilated bile ducts or frank jaundice. These patients underwent open cholecystectomy and CBDE. Cholangiography was attempted in 144 patients, and an adequate study was obtained in 13 1 (90%). Filling defects in the CBD that could not be proven to be air bubbles occurred in seven patients (5%). Stones were removed with irrigation and basketing in four patients. In one patient with a 5-mm CBD, passage of the stone basket pushed two 3-mm stones into the duodenum. In the remaining two patients, no stones were removed from the CBD despite several passes of the basket. Completion cholangiography was normal in all of these patients. Review of the preoperative evaluation of these seven patients showed few clues that would have predicted an abnormal cholangiogram. The CBD was of normal caliber in the six patients who had preoperative ultrasonography. Six patients had normal liver chemistry profiles. The seventh patient had mild elevations (less than twice normal) of the AST and alkaline phosphatase. No patient in this group had a history of fever or jaundice, and none were symptomatic at the time of surgery. 54
THE AMERICAN JOURNAL OF SURGERY
Postoperative liver profiles were obtained in live patients and were normal in one, mildly elevated in three, and showed frank hyperbilirubinemia (bilirubin: 4.6 mg/dL) in one. This last patient also had hyperamylasemia (450 IU/dL). The bilirubin was declining, and the amylase was normal 36 hours after surgery, so the patient was discharged. The other six patients were discharged the day after surgery. One of these six patients was readmitted 30 hours after discharge with nausea, vomiting, and mildly elevated alkaline phosphatase, LDH, GGTP, ALT, and AST. It was suspected that the patient had passed a retained CBD stone and was discharged the next day with symptoms resolved and laboratory abnormalities approaching normal. The findings at follow-up, 6 to 13 months (mean: 9.4 months) postoperatively, allowed a high degree of satisfaction with laparoscopic cholecystectomy and transcystic CBDE. No patient had recurrence of biliary symptoms, and no patient had required further diagnostic procedures. COMMENTS As laparoscopic cholecystectomy evolved, the general consensus was that the CBD stone was out of the reach of laparoscopic surgery. The experience in this series of seven patients suggests that most unsuspected CBD stones can be removed laparoscopically. Emboldened by this initial success with unsuspected CBD stones, we extended our preoperative indication to include patients with abnormal liver chemistries or a dilated CBD. As of December 1, 1991, laparoscopic transcystic CBDE had been performed in 20 patients. Choledochoscopy as well as completion cholangiography was performed in six of the last 10 patients but in only one of these patients was choledochoscopy necessary to retrieve a stone that could not be retrieved with fluoroscopic imaging. In three patients (including the two previously discussed), the cholangiogram was falsely positive (16%), but these patients were spared open CBDE or postoperative ERCP by a negative laparoscopic transcystic CBDE. In three patients (15%), transcystic CBDE was abandoned. In the first patient, recalcitrant valves of Heister prevented access to the CBD with catheters, baskets, and thin gauged guide wires. In the second patient, cholangiography demonstrated a dilated duct, packed with large stones and a small cystic duct. In the third patient, an impacted distal CBD stone could not be dislodged with balloons and baskets. Endoscopic lithotripsy was not available. In these three patients, CBDE was completed at laparotomy. Stones were removed from the remaining 14 patients with laparoscopic transcystic CBDE. If one considers all positive cholangiograms, the duct clearance rate in this series was 17 of 20 (85%). Including only patients with true positive cholangiograms, the duct clearance rate was 14 of 17 (82%). Of the 17 patients not converted to open CBDE, 13 were discharged on the first postoperative day, 2 were discharged on the second postoperative day, and 2 were discharged on the third postoperative day. Serious complications were rare. Hyperamylasemia was seen in two patients, hyperbilirubinemia in one pa-
VOLUME 163 JANUARY 1992
LAPAROSCOPIC BILE DUCT EXPLORATION
TABLE
I
Laparoscopic Common Bile Duct Exploration (CBDE)
Reference
No. of Laparoscopic Cholecystectomies
Positive Cholangiogram
[41 I51 I61 [71
500 516 342 270
29 35 26 19
Hunter+
252
20 (6)
(%)
(6) (7) (6) (7)
LC Clearance (%) 8 21 19 16
(26)* (60) (73) (64)
17 (65)
Converted to Laparotomy
Postoperative ERCP
3 6 5 2
16 5 5
3
0
1
“Laparoscopic CBDE was not attempted in early patrons of this series. Totals reflect resui?sof study through December 1991.
tient, and both conditions occurred in a fourth patient, but all had normalized within 48 hours. Presumably these patients developed edema or spasm of the sphincter of Oddi during stone extraction procedures that then led to jaundice or hyperamylasemia. In the last eight patients, glucagon has been administered at the time of duct exploration, and there have been no further postoperative enzyme elevations. One of these patients developed a subhepatic abscess related to a difficult cholecystectomy with spillage of infected bile. As previously mentioned, one patient probably passed a small retained stone not visible on completion cholangiography. These results compare favorably to a number of other series utilizing choledochoscopy as the primary imaging modality (Table I) [4-71. The method used most frequently for endoscopic transcystic CBDE requires dilatation of the cystic duct with a 5- to lo-mm balloon dilator (depending upon the size of the stone). A flexible fiberop tic endoscope 2.0 to 3.5 mm in diameter is inserted into the cystic duct and passed into the distal CBD or duodenum. Under direct vision, stones are basketed or grasped and pulled out of the cystic duct with the scope. If the stones are too large to be removed through the dilated cystic duct, they may be fractured with an electrohydraulit lithotriptor or pulsed dye laser. When cystic duct dilatation and choledochoscopy become the primary therapeutic response to a positive cholangiogram, duct clearance rates are 70% to 90% (Table I). The advantages of endoscopic CBDE are that it may be performed without the use of digital fluoroscopy, lithotripsy may be performed under direct vision, and choledochoscopy may result in fewer retained CBD stones than radiologic imaging alone [8]. The disadvantages of the endoscopic approach are that the cystic duct must be dilated in most patients to a ccommodate a therapeutic endoscope, that stones must be removed from the duct one by one, and that baskets small enough to be passed through the endoscope are too flimsy to crush all but the softest stones. In summary, both fluoroscopic and endoscopic imaging allow the removal of the majority of CBD stones. In patients with a tight tortuous cystic duct or a small CBD, the fluoroscopic technique allows stone retrieval without dilating any structure. Choledochoscopy may be best in a more dilated system where “blind basketing” is less likely to be successful. In addition, choledochoscopy is necesTHE AMERICAN
sary when laser or electrohydraulic lithotripsy is performed. Despite the very acceptable results of fluoroscopic and endoscopic CBDE presented here, this technique has not yet come of age. A survey of 8 large laparoscopic centers in Europe showed only 12 successes in 42 attempts (29%) [9]. The major reasons for failure of cystic duct CBDE are that the cystic duct is too small, the angle of access is incorrect, the CBD stones are too large or too numerous to be removed through the cystic duct, or the stones are intrahepatic and beyond the reach of scopes and baskets. The solution to these problems may be laparoscopic choledochotomy. Unfortunately, T-tube placement and secure sutured duct closure is technically demanding and is currently out of reach of the majority of laparoscopic surgeons. Additionally, we must keep in mind that CBDE is not adequate therapy for patients with primary CBD stones. These patients ought to have an endoscopic papillotomy or choledochoduodenostomy to achieve ductal drainage. For surgeons without the technology, training, or patience to perform laparoscopic transcystic CBDE, patients with elevations of liver chemistries or an ultrasonographically dilated CBD will continue to be referred for preoperative ERCP. How often will these noninvasive studies predict the finding of choledocholithiasis on ERCP? How often will the ERCP be necessary? Ultrasound fails to reveal choledocholithiasis in half of the patients proven to have CBD stones by cholangiography and falsely predicts the presence of stones in 50% of patients with normal ERCP [IOJI]. Hepatic chemistries are only slightly more accurate. In two series of more than 100 patients, the likelihood of finding CBD stones was 17% to 22% if one value on the liver profile (alkaline phosphatase, AST, LDH, or bilirubin) was elevated, 32% to 40% if two values were elevated, and 50% to 60% if three or more values were elevated [I2,13]. Thus, the yield of preoperative ERCP will be as low as 20% if performed for elevation of a single liver chemistry and will improve to only 50?&if the threshold is elevation of three values on the liver profile or a dilated CBD by ultrasound. In one large series of laparoscopic cholecystectomy, 64 patients were referred for preoperative ERCP, and only 10 (16%) had choledocholithiasis [2]. Eighty-four percent of preoperative ERCP were unnecessary. In addition to the limited yield of preoperative
JO1JRNAL OF SURGERY
VOLUME
163
JANUARY
1992
55
ERCP, the cost and discomfort level are significantly greater to the patient than removing CBD stones under general anesthesia at the time of laparoscopic cholecystectomy. More importantly, there is a legitimate concern that the morbidity of endoscopic papillotomy and laparoscopic cholecystectomy might be higher than the morbidity of laparoscopic cholecystectomy converted to open cholecystectomy when irretrievable stones are found. No data comparing these two approaches currently exist, but it was recently demonstrated that morbidity is doubled when endoscopic papillotomy and open cholecystectomy are performed rather than open cholecystectomy and CBDE [Z4]. The recommendation that ERCP should be done before laparoscopic cholecystectomy in all patients with elevated liver chemistries and that endoscopic papillotomy should be performed when stones are found needs further prospective evaluation. While elevated liver chemistries are only moderately accurate predictors of choledocholithiasis (specificity: 15% to 50%), normal liver chemistries are reported to be extremely accurate predictors of the absence of stones (specificity: 100%) [ I2,Z 31. This was not our experience, nor has it been the experience of other surgeons. In the initial series of 150 patients, 4 of 133 patients (3%) undergoing successful cholangiography had choledocholithiasis despite normal levels of alkaline phosphatase, AST, LDH, and bilirubin. In another series of laparoscopic CBDE, 7 of 14 patients (50%) with choledocholithiasis had normal liver chemistries preoperatively (Sangster, personal communication). These truly unsuspected stones can only be detected by the routine use of operative cholangiography. The management of the unsuspected tilling defect in centers without the capability to perform transcystic CBDE is controversial. In most centers, such patients are referred for postoperative endoscopic papillotomy. While such a practice preserves the “minimally invasive” ap preach, it has certain risks: Papillotomy fails in 5% to 20% of patients (depending on the experience of the endoscopist). Failure may require return to the operating room for an open choledochotomy and CBDE. The morbidity of endoscopic papillotomy is loo/o,and 1% of these complications require operative management (for bleeding or perforation) [15]. Lastly, postoperative ERCP is expensive, uncomfortable, and requires an extra day of hospitalization after laparoscopic cholecystectomy. Another approach advocated is to leave small stones to pass spontaneously. In our community and in Seattle (W. Traverso, personal communication), many cases of pancreatitis have resulted from these small “probably insignificant” bile duct stones left to pass on their own after laparoscopic cholecystectomy. For all these reasons, we have elected to perform open cholecystectomy, CBDE, and choledochoscopy in patients failing transcystic CBDE. While the pain, hospitalization, and rehabilitation may be greater with this approach, the success of duct clearance is 98% [8], and the morbidity may be less than with cholecystectomy and endoscopic papillotomy 1141. The evolution of technique and development of lapa56
THE AMERICAN JOURNAL OF SURGERY
roscopic bile duct exploration skills should markedly reduce the need for preoperative and postoperative ERCP. In this series of 150 laparoscopic cholecystectomies, only 2 patients underwent ERCP; 1 patient who presented with severe cholangitis underwent ERCP preoperatively, and 1 patient underwent ERCP postoperatively after developing cholangitis 3 days after laparoscopic cholecystectomy during which cannulation of the cystic duct had failed. Neither patient had choledocholithiasis at the time of ERCP, and neither patient was included in this review, as there was no indication for transcystic CBDE in either case. A review of Table I demonstrates that, when an aggressive laparoscopic approach to choledocholithiasis is taken, bile duct clearance is most often complete and the use of ERCP diminishes [4-71.
REFERENCES 1. Cotton PB, Baillie J, Pappas TN, Meyers WS. Laparoscopic cholecystectomy and the biliary endoscopist. Gastrointest Endcec 1991; 37: 94-6. 2. Meyers WC, et al. (Southern Surgeons Club). A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 1991; 324: 1073-8. 3. Olsen DO. Laparoscopic cholecystectomy. Am J Surg 1991; 161: 339-44. 4. Spaw AT, Reddick EJ, Olsen DO. Laparoscopic laser cholecystectomy: analysis of 500 procedures. Surg Laparosc Endosc 1991; 1: 2-7. 5. Sackier J, Berci G, Phillips E, Carroll B, Shapiro S, Paz-Partlow M. The role of cholangiography in laparcscopic cholecystectomy. Arch Surg 1991; 126: 1021-6. 6. Petelin JB. Laparoscopic approach to common duct pathology. Surg Laparosc Endosc 1991; 1: 33-41. 7. Swanstrom L, Sangster W. Laparoscopic management of common bile duct stones. Presented at the British Columbia Surgical Society, Vancouver BC. May 10, 1991. 8. Escat J, Fourtanier G, Maigne C, Vaislic C, Fournier D, Prevost F. Choledochoscopy in common bile duct surgery for choledocholithiasis: a must. Am Surg 1985; 51: 166-7. 9. Cuschieri A. Results of transcystic common bile duct exploration: the European experience. Presented at the Royal College of Surgeons, London, March 16, 1991. 10. Cronan JJ. US diagnosis of choledocholithiasis: a reappraisal. Radiology 1986; 161: 133-4. 11. Gross BH, Hatter LP, Gore RM, et al. Ultrasonic evaluation of common bile duct stones: prospective comparison with endoscopic retrograde cholangiopancreatography. Radiology 1983; 146: 471-4. 12. Del Santo P, Kazarian KK, Rogers JF, Bevins PA, Hall JR. Prediction of operative cholangiography in patients undergoing elective cholecystectomy with routine liver function chemistries. Surgery 1985; 98: 7-11. 13. Pemthaler H, Sandbichler P, Schmid Th, Margreiter R. Operative cholangiography in elective cholecystectomy. Br J Surg 1990; 77: 399-400. 14. Neoptolemos JP, Carr-Locke DL, Fossard DP. Prospective random&d study of preoperative endoscopic sphincterotomy versus surgery alone for common bile duct stones. BMJ 1987; 294: 470-4. 15. Cotton PB, Lehman G, Vennes J, et cl. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991; 37: 383-93.
VOLUME 163 JANUARY 1992
LAPAROSCOPIC BILE DUCT EXPLORATION
that when you manipulate the cystic duct, it is important (Jackson, MS): Transcystic fluoro- to consider the means to ligate it at the completion of the scopic basketing of bile duct stones is reliable, and we procedure. Those cystic ducts will be enlarged and friahave used it in 12 or 13 patients. I do emphasize that the ble, and it is important to ligate them with a ligature and dynamic studies obtained with C-arm fluoroscopy offer not just with a clip, because the clip often will not hold. substantially more information than the static films and Finally, do you have to use pharmacologic manipulation obviate the need for expensive radiolucent disposable can- during your cholangiograms? We do not have the luxury nulae (at a cost savings of $425 to $750). We have been of having the fluoroscopic unit that you described. We use quite selective in obtaining operative cholangiograms in static films, and it truly is shooting a “blind” x-ray, with trocars in the way and waiting 25 minutes until they are our series, which now exceeds 800 patients. Forty-nine percent of our patients were judged to be at no risk for developed. So we have performed cholangiograms selechaving CBD stones (common duct diameter less than 5 tively and often use pharmacologic manipulation to obtain optimal films. If the roentgenogram shows only flow mm, normal liver function tests, absence of jaundice, pancreatitis, or acute cholecystitis). None of these pa- distally, we administer intravenous morphine to cause tients underwent operative cholangiograms, and none spasm of the sphincter of Oddi. Conversely, with no flow have had evidence of a retained stone to date. The sensi- into the duodenum, an ampule of intravenous glucagon tivity and specificity of our clinical determination appear may result in relaxation of the sphincter. Thomas R. Gadacz (Baltimore, MD): The concento exceed those obtained by “routine” or “discovery” cholangiography. Furthermore, some of the stones that tration of contrast that you used was 60%. Why wasn’t a you identified may have passed spontaneously (clinical lower concentration used if you wanted to detect these false positives). The combined cost-savings of electrosur- small stones? How much time does the cholangiogram gery (versus laser), reusable instruments, and selective add to your procedure, and how much time does the operative cholangiography are substantial. In a recent cholangiogram plus basket removal add? Michael S. Nussbaum (Cincinnati, OH): Have you audit of 50 patients under age 60, the mean hospital charge for laparoscopic cholecystectomy generated by been able to utilize this technique in handling stones that our approach was only $1,910. What was the charge for are proximal to the cystic-common duct junction? That the procedure at your hospital? What is your policy about seems to be a problem with this technique in that you can only direct the basket down into the CBD but not up the disposable versus nondisposable instruments? common hepatic duct towards the bifurcation. John Sonneland (Spokane, WA): Would you comGerald Larson (Louisville, KY): I commend you for ment on the risk of injury to the cystic duct, when dilated in order to allow the removal of CBD stones by basket? your high success rate in exploring the CBD. I think we Would you comment on an alternative method of han- need to develop our technique so that we can explore the duct as successfully and reliably as we do with open dling CBD stones, namely fragmentation by laser? Greg V. Stiegmann (Denver, CO): I noticed in your cholecystectomy. If you have a patient in the preoperative series that none of these patients appeared to have a high period with a dilated CBD and elevated liver function likelihood of the presence of choledocholithiasis prior to tests and a very high probability of having a CBD stone, your laparoscopic operation. How did you find such a would you do a preoperative ERCP and sphincterotomy to clear the duct followed by laparcscopic cholecystectoselected group, or did you perhaps perform prelaparcscopit cholecystectomy retrograde cholangiography on a my? Or, would you proceed directly to the laparoscopic number of these patients prior to your operation? Of your cystectomy assuming that you can remove the stone inseven patients, one seemed to have had elevated amylase traoperatively? I am concerned that perhaps we are being levels afterwards, if not some clinical pancreatitis. That, too liberal in our indications for preoperative sphincteroof course, is one of the major complications of retrograde tomy in young patients with a long life expectancy ahead. cholangiography, sphincterotomy, and so forth. Do you Aaron S. Fink (Cincinnati, OH): I think you introbelieve that with transpapillary manipulation you per- duce another important point when you raise the issue of haps expose these patients to a risk of pancreatitis, and if preoperative versus postoperative ERCP. I think the indiso, can the technique be modified to obviate that risk? cated failure rates of 5% to 20% reflect the fact that both Nathaniel Soper (St. Louis, MO): Some of those the success rate and the complication rate will vary with common duct tilling defects looked so small that either the expertise available in your local institution. I think they would have passed spontaneously or perhaps they that this is a critical factor that must be used when chooswere air bubbles rather than stones. I question the need ing the modality to use in dealing with stones suspected for chasing all of the filling defects that we see, particu- preoperatively or discovered either intraoperatively or larly if there are complications of the basket extraction postoperatively. As we consider these facts, I think it itself. Should we be a little more selective in terms of further urges us to develop more reliable laparoscopic when we perform cholangiogram? If you see small filling techniques, such as you’re beginning to describe. defects, when should these be left alone? My second John G. Hunter (closing): Dr. Voyles, I am glad you question deals with the issue of CBD drainage. Do you did not ask me to address the routine versus selective think it should be used in these patients? If so, how would cholangiography issue as such a discussion could go on for you go about holding the T-tube in place, or would you days. I do not know the exact hospital cost for this proceput it through the cystic duct? My third point would be dure, but I can tell you that our mean hospital charges DISCUSSION C. Randle Voyles
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
163
JANUARY
1992
57
were $3,900 with routine cholangiography. This charge is $900 less than the charge for open cholecystectomy and a 3-day hospital stay in the year before I started performing laparoscopic cholecystectomy. We use mostly reusable trocars to save expense, but it is difficult to keep them sharp and I believe the safety shield may offer some small advantage. Dr. Sonneland, cystic duct dilatation is necessary in 45% to 50% of patients who are undergoing choledochoscopy because the smallest adequate choledochoscope that one can use laparoscopically is close to 3 mm in diameter and unless the passage of stones has dilated the cystic duct, it is necessary to use a balloon or rigid dilator. We have not used cystic duct dilatation routinely. We have only been doing completion choledochoscopy after fluoroscopic stone extraction when we could pass the scope readily through the undilated cystic duct. I am a little concerned that dilatation might rupture or excessively thin out the cystic duct and increase the risk of a postoperative bile leak, when the gain of choledochoscopy is not yet clear. Dr. Stiegmann, I have performed a preoperative ERCP in only one patient in 1% years. While I enjoy performing this procedure, I have been tempered by our inability to accurately predict who will have stones preoperatively and by the additional morbidity of adding endoscopic sphincterotomy to cholecystectomy. In this early experience, I refused to perform laparoscopic cholecystectomy on two patients for ultrasonographically visible CBD stones. All the patients reported in this series had truly unsuspected stones. A completion cholangiogram of the patient who developed hyperamylasemia showed no tilling of the duodenum. Such is the case after open bile duct exploration 20% of the time, and rarely is this related to a retained stone. I suspect my patient had some spasm or edema of the ampulla, which contributed to the pancreatitis and jaundice. This begs the question as to whether we should be draining the CBD when we cannot demonstrate duodenal filling on the completion cholangiogram. Dr. Soper, if I could identify stones that were going to pass spontaneously and not cause pancreatitis or cholangitis in passing, I think it would be reasonable to leave them behind. However, I have much anecdotal evidence from around our community and elsewhere that small stones left behind in the hope that they will pass are causing significant pancreatitis and cholangitis. I think the main reason for doing routine cholangiography is not to diagnose choledocholithiasis but to identify the anatomy of the region and to become proficient in the technique. The impact of a false negative laparoscopic transcystic CBDE is much less than doing an unnecessary open bile duct exploration. There are no tubes coming out of the patient’s side, and none of the patients with negative duct explorations have yet had any sequelae of their
58
THE AMERICAN JOURNAL OF SURGERY
exploration. I believe bile duct drainage with a small caliber latex drain, secured with a pre-tied catgut ligature (Endoloop, Ethicon, Somerville, NJ) , might be appropriate for decompression, imaging, and possibly stone extraction in patients with nonfilling of the duodenum on completion cholangiography. Using fluoroscopy, we have been able to obtain a complete cholangiogram in every patient without pharmacologic manipulation, Sometimes it is necessary to put the patients head down to fill the intrahepatic ducts, and often we need to rotate the head of the C-arm in order to avoid the overlying shadow of metal trocars. In all patients where cystic duct cannulation has been successful, dynamic fluoroscopic cholangiography has provided an adequate study. Dr. Gadacz, our radiologist asked me to start using 60% Conray in order that clear ductal definition would be provided despite the pneumoperitoneum and fluoroscopic imaging. If one starts injection slowly and watches the duct fill, there will be sufficient dilution of contrast by bile so small stones are seen. On the other hand, the patient who may have had a retained stone because of excessive opacification of the bile duct makes a good argument for using a more dilute contrast agent. Initially, the cholangiogram took 20 minutes but now takes 5 to 10 minutes. Initially, duct exploration took 60 to 90 minutes, because we were somewhat disorganized and had difficulty lccating baskets and guide wires. We now have it fairly streamlined so that it takes from 20 to 40 minutes to perform this procedure. Dr. Nussbaum, most of the patients in this series had a few small stones, and, with the patient in a steep reverse Trendelenburg, most stones were located in the distal CBD. In this series, we did not have a filling defect proximal to the cystic duct. With guide wires, it may be possible to get into the proximal bile duct but such access will be limited by cystic duct anatomy. Choledochotomy may be required to remove many of these stones. Dr. Larson, I think preoperative ERCP with sphincterotomy is being over-utilized. However, I think that a patient with a markedly dilated duct who is jaundiced may have a periampullary cancer. If the ERCP demonstrates choledolithiasis, this patient will most likely need a duct drainage procedure that cannot be provided laparostopically. An ERCP with sphincterotomy may provide that necessary duct drainage and allow the surgeon to proceed with laparoscopic cholecystectomy. In nonjaundiced patients with a biochemical indication of CBD stones but no ductal dilatation, I have proceeded with laparoscopic cholecystectomy as the initial procedure. The patient is aware that a failure of laparoscopy will result in conversion to open bile duct exploration. Dr. Fink, there is no doubt that the use of perioperative ERCP will vary by community, but I believe many surgeons will be able to remove the majority of CBD stones at the time of laparoscopic cholecystectomy.
VOLUME 163 JANUARY 1992
