surgical endoscopy 3

Bile duct injuries during laparoscopic cholecystectomy

D. Olsen

Baptist Hospital, 200 Church Street, Nashville, TN 37236, USA

Received: 14 May 1996/Accepted: 1 July 1996

AbstractBackground:With the introduction of laparoscopic chole-cystectomy, an increase in the incidence of bile duct injurytwo to three times that seen in open cholecystectomy waswitnessed. Although some of these injuries were blamed onthe ‘‘learning curve,’’ many occurred long after the surgeonhad passed his initial experience. We are still seeing theseinjuries today.Methods:To better understand the mechanism behind theseinjuries, in the hope of reducing the injury rate, 177 cases ofbile duct injury during laparoscopic cholecystectomy werereviewed. All records were studied, including the initialoperative reports and all subsequent treatments. Videotapesof the procedures were available for review in 45 (25%) ofthe cases. All X-ray studies, including interoperative chol-angiograms and ERCPs, were reviewed.Results:The vast majority of the injuries seen in this review(71%) were a direct result of the surgeon misidentifying theanatomy. This misidentification led to ligation and divisionof the common bile duct in 116 (65%) of the cases. Chol-angiograms were performed in only 18% (32 patients) ofcases, and in only two patients was the bile duct injuryrecognized as a result of the cholangiogram. Review of theX-rays showed that in each instance of common bile ductligation and transection in which a cholangiogram was per-formed the impending injury was in evidence on the X-rayfilms but ignored by the surgeon.Conclusions:From this review, several conclusions can bedrawn. First and foremost, the majority of bile duct injuriesseen with laparoscopic cholecystectomy can either be pre-vented or minimized if the surgeon adheres to a simple andbasic rule of biliary surgery; NO structure is ligated or di-vided until it is absolutely identified! Cholangiography willnot prevent bile duct injury, but if performed properly, itwill identify an impending injury before the level of injury

is extended. And lastly, the incidence of bile duct injury isnot related to the laparoscopic technique but to a failure ofthe surgeon to translate his knowledge and skills from hisopen experience to the laparoscopic technique.

Key words: Bile duct injury — Laparoscopic cholecystec-tomy — Common bile duct injury

With the introduction of laparoscopic cholecystectomy [21],the general surgery community was thrust into the world ofminimal invasive surgery. The marked benefits of laparo-scopic cholecystectomy became apparent in a short periodof time [22, 23, 32], driving the procedure to widespread uselong before any clinical trials or studies were available todocument the procedure’s safety. Although there was earlyconcern over the potential dangers of the laparoscopic ap-proach [3], it took several years before there was enoughdata to verify this concern [4–7, 29]. It is now well acceptedthat the incidence of duct injury with the laparoscopic pro-cedure is approximately two to three times that which hasbeen reported for open cholecystectomy [7]. If we assumethat the cholecystectomy performed during a laparoscopicapproach is essentially the same operation that is performedduring an open cholecystectomy, then we have to ask our-selves why this discrepancy occurs. Since the same prin-ciples of surgery are practiced in each approach, the inci-dence of bile duct injury should be no higher for the lapa-roscopic approach as compared to the open procedure. Thisdifference in bile duct injury rates could be related to:

1. An inherent limitation of the laparoscopic approach2. A result of the ‘‘learning curve’’ on the part of the lap-

aroscopic surgeon developing his operative skills3. A failure of the surgeon to translate the basic principles

of performing a cholecystectomy from one approach tothe other

In an attempt to address this issue, a review of 177 bile ductinjuries that occurred during the performance of laparoscop-

Presented at the annual meeting of the Society of American Gastrointes-tinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA,13–17 March 1996

Correspondence to:D. O. Olsen, 300 20th Avenue North, Suite G-3, Nash-ville, TN 37203, USA

Surg Endosc (1997) 11: 133–138

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997

ic cholecystectomy was performed. An effort was made toanswer three individual questions:

1. What is the mechanism of injury responsible for bile ductinjuries seen during laparoscopic cholecystectomy?

2. Are the injuries simply a risk of the procedure, or arethey preventable?

3. Does cholangiography play a role in minimizing or pre-venting the injuries?

Materials and methods

A total of 177 cases of bile duct injuries were collected over a period of 4years from 1990 to 1994. The cases were reviewed as a result of inquiriesregarding the ‘‘appropriateness’’ of care, and none were obtained frompreviously published material. All cases in which either a clear-cut injuryto the duct was recorded or in which a major unexplained bile leak oc-curred were included in the study. The study was retrospective in nature,gathering cases from institutions across the county, comprising both smallcommunity hospitals and major university centers. All records of hospi-talization were reviewed including both the initial hospitalization when thelaparoscopic cholecystectomy was performed and any subsequent hospi-talization for diagnosis or treatment of the complication Cholangiograms,ERCPs, and any other study that was felt to be important to determine thecause of the injury were reviewed. The videotape of the procedure wasviewed in the 45 cases in which a tape was available (25%).

The injuries were divided into five basic types according to commonfeatures suggesting some form of common mechanism of injury among thegroups (Table 1). The three largest groups were further subdivided intosubgroups to further categorize the injuries (Table 2). Data including pa-tient demographics, injury types, whether a cholangiogram was performed,outcome of the cholangiogram, and time of diagnosis o the injury were allrecorded and entered into a database for analysis.

Results

In this study there were 40 males and 137 females whosustained bile duct injuries as a result of laparoscopic cho-lecystectomies. The average age was 45 for the males and39 for the females. Average height for the males was 6’0’’,with an average weight of 198 pounds. The average heightfor the females was 5’5’’, with an average weight of 165pounds.

The most common injury was a type III injury, whichoccurred in 116 (65%) of the 177 cases (Fig. 1). The ma-jority of the type III injuries included loss of bile duct afterthe duct was ligated and divided (type III-B). This subgroupincluded 94 (53%) of the injuries (Fig. 2). This is signifi-cant, since this is perhaps the most devastating of all bileduct injuries. The second most common injury, occurring in30 of the cases (17%), was a type I injury. These injuriesresulted from inadvertent holes, burns, or lacerations thatwere created in the main biliary duct as a result of dissectionor as a result of cauterization around this structure. A typeIV injury was the third most common injury seen, compris-

ing 19 (11%) of the cases. A type IV injury included anyisolated injury to the right hepatic duct system (type IV).The records suggest that these injuries resulted from varia-tions in the ductal architecture that placed the injured struc-ture in harms way during the dissection of the cystic duct.Variations noted included a ‘‘high’’ takeoff of the cysticduct (short common hepatic duct), cystic duct coming off ofthe right hepatic duct (absent common hepatic duct), and alow-lying right segmental duct. There were two (6%) type IIinjuries in the series. The type II injuries were broken downinto two subgroups to set apart the specific circumstancewhen the surgeon misidentified the common bile duct as thecystic duct but recognized the problem with a cholangio-gram prior to division of the duct. Because both the type IIinjuries and the type III injuries result from misidentifyingthe common bile duct as the cystic duct, we can say that thevast majority (71%) of the injuries were a result of thesurgeon misidentifying the anatomy! Type V injuries wereunusual (two cases) and consisted of late strictures seenmonths after an otherwise-uncomplicated operative proce-dure. The causes of the strictures could never be identified,but they are included because they imply an operative injuryto the duct. Major bile leaks as a result of a cystic ductstump leak (type I-B) occurred in only seven (4%) of thecases (Fig. 3), indicating that this was a minor cause ofmajor postoperative morbidity resulting from injury to thebile ducts.

The incidence of cholangiography was 18% in this se-ries (Fig. 3), but in only two out of the 32 cholangiogramsperformed did the surgeon recognize the mistake or injuryfrom the cholangiogram. In one case, the surgeon recog-nized the common bile duct cannulation only after he li-gated, divided, and excised a portion of the common bileduct. This occurred because the surgeon proceeded with theoperation before the cholangiogram was ready for review.In the second case, the surgeon recognized the common bileduct cannulation before the common bile duct was ligatedand divided. The duct was successfully repaired with asimple primary repair of the duct. In reviewing the 32 chol-angiograms performed in this series, 21 of the cholangio-grams were performed during cases that ultimately ended ina common bile duct ligation and transection. The impendingbile duct injury could be seen in retrospect in each of these21 cholangiograms. Accepting an incomplete study as‘‘normal,’’ with no filling of the upper radicals (Fig. 4), wasthe number 1 reason that the surgeon missed the impendinginjury.

In all 177 cases, the injury to the bile duct was identifiedat the time of surgery in only 27% of the cases (Fig. 5). Themajority of patients were discharged within 24 hr of theirprocedure and were admitted at a later time with symptomssuggesting postoperative complications.

Table 1.Classification of injury types

Type I Simple laceration or hole in the duct due to trauma, cauteryburn, etc.

Type II Ligation of duct without transectionType III Ligation of duct with transectionType IV Isolated right hepatic duct injuryType V Stricture of uncertain cause

Table 2. Injury subtypesa

Type I-A Simple laceration or holeType I-B Cystic duct leakType II-A Ligation of duct without transectionType II-B CBD cannulation during C-gramType IIIA Ligation with division of CBDType IIIB Ligation, division, and excision of CBD

a CBD, common bile duct; C-gram, cholangiogram

134

Discussion

Injury to the biliary ducts during the performance of a cho-lecystectomy is not unique to the laparoscopic approach.Bile duct injury has long been recognized as a potentialcomplication during open cholecystectomy [1, 10–12].Through proper training and recognition of the causativefactors [30], the incidence of bile duct injuries has beenreduced to approximately 0.2% for open cholecystectomy[7]. Some reviews report rates as low as 0.07% [1]. Whenwe compare these numbers to laparoscopic cholecystecto-my, we find series of laparoscopic cholecystectomies thatreport as low an incidence of bile duct injury as that seenwith the open technique [2, 27]. But when the experienceacross the United States is examined, we see that the inci-dence of bile duct injuries is higher, with an overall inci-dence of 0.6%, three times that of the open procedure [7]!Although the ‘‘learning curve’’ has been blamed for someof this increase in duct injuries [4], the injuries are stillbeing reported, nearly 6 years after the introduction of lap-aroscopic cholecystectomy. It has been suggested that manyof these injuries are preventable [29], and with proper tech-nique there is no reason that we should not be able to reducethe rate of bile duct injury to that which is seen with theopen procedure.

For open cholecystectomies, injuries to the bile ducthave been attributed to (1) inadequate access, exposure, and

assistance, (2) absence of operative cholangiograms, (3) pa-tient’s obesity, and (4) early dissection of Calot’s triangle[17]. It is further suggested that the vast majority of theseinjuries are avoidable if ‘‘the principles of safe cholecys-tectomy are adhered to’’ [9, 17]. In this review of 177 bileduct injuries that occurred during the performance of lapa-roscopic cholecystectomy, 70% of the injuries occurred be-cause the surgeon mistakenly identified the common bileduct as the cystic duct. The surgeon then proceeded to ligateand divide the common bile duct, and in 53% of the cases,excised a segment of the common bile duct/common hepaticduct. The issue of identification of the anatomy is basic tothe performance of a cholecystectomy and should not behindered by laparoscopic technique. If anything, the abilityto identify the anatomy should be enhanced given the in-creased exposure that is achieved with laparoscopic surgery.When reviewing the videos of the procedures in this review,the surgeons uniformly proceeded to apply clips and dividestructures before any of the key landmarks were identified.A basic principle of surgery is that no structure should beligated or divided until it is identified. All the textbooks onsurgery talk of the importance of identifying the cysticduct/common bile duct junction and of the utility of a ret-rograde dissection to identify the infundibulum of the gall-bladder [9, 26]. The size of a duct or apparent direction ofa duct cannot be used for identification because of theanomalies that are seen in the biliary anatomy [9]. If eitherof these two landmarks is identified (infundibulum or cysticduct/common bile duct junction), an injury of ligation ordivision of the common bile duct cannot occur! During thelaparoscopic procedure, if the surgeon cannot achieve theabove goal, there is always the option to convert to an openprocedure. We can, therefore, conclude from this study thatthe vast majority of bile duct injuries seen during laparo-scopic cholecystectomy are not an inherent limitation withthe procedure but are a result of surgeons not translating thebasic principles of surgery over to their laparoscopic pro-cedures. It has been suggested that these injuries are a resultof the ‘‘learning curve’’ [4]. Although we cannot draw anyconclusions regarding the absolute incidence of duct inju-ries from this review, we can say that the relative incidenceof the types of injury has not changed over the 4 years ofthis study (Fig. 6). Given the short learning curve reportedfor laparoscopic cholecystectomy [4], we should certainlyhave seen a decline in the relative incidence of those injuries

Fig. 1. Bile duct injuries during lap-aroscopic cholecystectomy.

Fig. 2. Injury subtypes.

135

that are related to inexperience, such as misidentification ofthe anatomy.

The injuries that result from a pure misidentification ofthe anatomy should all be preventable if we adhere to thebasic principles of surgery. If we eliminate the 70% of theinjuries that occur as a result of the misidentification ofanatomy, we have a rate of injury that is less than that seenin open cholecystectomy.

The issue of cholangiography has been debated since the1930s when cholangiography was first introduced [8, 13–16, 18, 20, 21, 24, 25, 28, 31]. With the introduction of

laparoscopic cholecystectomy, it has been argued that rou-tine cholangiography can prevent bile duct injury [21, 25].Although we cannot make such a statement as a result ofthis review, it is clear that a properly performed cholangio-gram would have reduced the severity of the injury andallowed the earlier identification of the injury in the major-ity of the cases. This has been suggested in another study,where it was demonstrated that the routine use of cholan-giography reduced the incidence of ‘‘high’’ duct injurieswhen compared to series in which selective cholangiogra-phy was performed [21]. Although the number of duct in-juries was not reduced, the severity of the injury was di-rectly affected by the use of the cholangiogram. The needfor an enteric-biliary anastomosis was zero in the group thathad routine cholangiography, while it was as high as 89% inthe groups that practiced selective cholangiography. Withthe anatomy identified by cholangiography, further dissec-tion of the triangle of Calot can be performed safely usingthe cholangiogram as a road map. This should reduce therisk of injury to an accessory duct or an aberrant right he-patic duct that may be injured in the process of removing thegallbladder despite adequate identification of the cysticduct. In this review, 11% of the injuries were a result ofinjury to the right hepatic duct. Only two cholangiogramswere performed in this group, and when these studies werereviewed in retrospect, little could be said about the status ofthe right hepatic duct because in each study there was poorfilling of the upper radicals. A complete cholangiogramwith clear opacification of at least the right and left hepaticducts should allow some protection against injury to aber-rant ductal structures that can be in ‘‘harms way’’ during thedissection of the triangle of Calot or removal of the gall-bladder. This is especially true when there is a component ofacute inflammation when these structures can easily bepulled into the field of dissection with the upward tractionthat is applied to the gallbladder during this procedure [19].

Conclusion

The most common cause of injury to the bile ducts duringlaparoscopic cholecystectomy is misidentification of theanatomy. Whether we blame the injuries of misidentifica-

Fig. 3. Incidence of cholangiogra-phy in cases of bile duct injury: 18%cholangiography overall.

Fig. 4. Typical cholangiogram from a type III injury, in which case thesurgeon interpreted the study as ‘‘normal’’ and proceeded to ligate, divide,and excise a segment of common duct.

136

tion on poor exposure, absence of operative cholangio-grams, patient’s obesity, acute inflammation, or poor tech-nique in retraction of the triangle of Calot, we can sayadamantly that they ultimately occurred because the sur-geon violated one of the most basic and simple of all sur-gical principles; NO structure is ligated or divided until it isidentified. In biliary tract surgery this is of even greaterimportance, because we are trained to recognize the variousanomalies and variations that occur in the biliary tree. Fur-thermore, we understand the significance of a mistake inthis area and the morbidity that occurs if the wrong structureis divided. If the laparoscopic surgeon cannot make a posi-tive identification, he is obligated to convert to an openprocedure so that gallbladder can be taken down in a fun-dus-first technique.

We can conclude from this review that the vast majorityof the bile duct injuries that occur in laparoscopic chole-cystectomy are avoidable if the ‘‘principles of safe chole-cystectomy are adhered to.’’ Although the learning curveundoubtedly plays a role in the overall incidence of ductinjuries, it does not seem to have played a role in the typeof injury that occur. This is hopeful in the sense that if westrive to follow these ‘‘basic’’ principles, we stand to lowerthe incidence of duct injury even more.

If an injury is going to occur because of misidentifica-tion, cholangiography will not prevent the injury. But aproperly performed cholangiogram will minimize the extent

of the injury, which would otherwise turn out to be a majorduct excision without the aid of a cholangiogram. In orderfor a cholangiogram to be of benefit, a complete studyshowing the upper biliary tree has to be achieved. Preven-tion of just one major duct injury can justify the expense ofall the routine cholangiograms that a surgeon performs inhis career.

References

1. Andren-Sandberg A, Alinder G, Bengmark S (1985) Accidental le-sions of the common bile duct at cholecystectomy. Ann Surg 201:328–332

2. Baird DR, Wilson JP, Mason EM, Duncan TD, Evans JS, Luke JP,Ruban DM, Lucas GW (1992) An early review of 800 laparoscopiccholecystectomies at a university-affiliated community teaching hos-pital. Am Surg 58: 206–210

3. Cushieri A, Berci G, McSherry C (1990) Laparoscopic cholecystec-tomy. Am J Surg 159: 273

4. Davidof AM, Pappas TN, Murray EA, Hilleren DJ, Johnson RD,Baker ME, Newman GE, Cotton PB, Meyers WC (1992) Mechanismsof major biliary injury during laparoscopic cholecystectomy. Ann Surg215: 196–202

5. Deziel DJ (1994) Complications of cholecystectomy; incidence, clini-cal manifestation and diagnosis. Surg Clin North Am 74(4): 809–823

6. Deziel DJ, Millikan KW, Economou SG, Doolas A, Ko ST, Airan MC(1993) Complications of laparoscopic cholecystectomy: a national sur-vey of 4,292 hospitals and an analysis of 77,604 cases. Am J Surg 165:9–14

Fig. 5. Identification of bile duct in-juries at the time of surgery.

Fig. 6. Injury types by year.

137

7. Gallstones & Laparoscopic Cholecystectomy. NIH Consensus State-ment 1992 Sept 14–16: 10(3): 1–26

8. Gerber A (1986) A requiem for routine operative cholangiogram. SurgGynecol Obstet 163: 363

9. Glen F, Cameron JL (1981) Complications following operations on thebiliary tract. In: Hardy’s (ed) Complications in surgery and their man-agement. WB Saunders, Philadelphia, pp 512–538

10. Glenn F (1978) Iatrogenic injuries to the biliary ductal system. SurgGynecol Obstet 146: 430–434

11. Hillis T, Westbrook K, Caldwell F, Read F (1977) Surgical injury tothe common bile duct. Am J Surg 134: 712–716

12. Kune G (1979) Bile duct injury during cholecystectomy: cause, pre-vention and surgical repair in 1979. Aust NZ J Surg 49: 35–40

13. Lillemoe K, Yeo C, Talamini M, Wang BH, Pitt HA, Gadacz TR(1992) Selective cholangiography: current role in laparoscopic chole-cystectomy. Ann Surg 215: 669–676

14. Lorimer J, Fairfull-Smith R (1995) Intraoperative cholangiography isnot essential to avoid duct injuries during laparoscopic cholecystecto-my. Am J Surg 169: 344–347

15. Mirizzi PL (1932) Operative cholangiography. Surg Gynecol Obstet65: 702–710

16. Mofti AB, Ahmed I, Tandon RC, Al-Tameem MM, Al-Khudairy NN(1986) Routine use of selective preoperative cholangiography. Br JSurg 73: 548–550

17. Moossa AR, Mayer AD, Stabile B (1990) Iatrogenic injury to the bileduct. Arch Surg 125: 1028–1031

18. Nauta R (1989) Selective vs routine use of interoperative cholangiog-raphy. Arch Surg 124: 1041

19. Olsen D, Asbun H, Reddick E, Spaw A (1991) Laparoscopic chole-cystectomy for acute cholecystitis. Probl Gen Surg 8: 426–431

20. Pasquale M, Nauta R (1989) Selective vs routine use of interoperativecholangiography. Arch Surg 124: 1041–1042

21. Phillips E (1993) Routine versus selective intraoperative cholangiog-raphy. Am J Surg 165: 505–507

22. Reddick EJ, Olsen DO (1989) Laparoscopic cholecystectomy: a com-parison with mini-lap cholecystectomy. Surg Endosc 3: 131–133

23. Reddick EJ, Olsen DO, Daniell JF, Saye WB, McKernen B, Muller W,Hoback M (1989) Laparoscopic laser cholecystectomy. Laser MedSurg News 7: 38–40

24. Rolfsmeyer E, Bubrick M, Kollitz PR, Onstad GR, Hitchcock CR(1982) The value of operative cholangiography. Surg Gynecol Obstet154: 369–371

25. Sackier J, Berci G, Phillips E, Carroll B, Shapiro S, Paz-Partlow M(1991) The role of cholangiography in laparoscopic cholecystectomy.Arch Surg 126: 1021–1026

26. Schwartz S (1984) Gallbladder and extrahepatic biliary system. In:Schwartz S, Shires G, Spencer F, Storer E (eds) Principles of surgery.McGraw-Hill, New York, pp 1307–1343

27. Spaw AT, Reddick EJ, Olsen DO (1991) Laparoscopic laser chole-cystectomy: analysis of 500 procedures. Surg Laparosc Endosc 1: 2–7

28. Soper N, Dunnegan D (1992) Routine versus selective interoperativecholangiography during laparoscopic cholecystectomy. World J Surg16: 1133–1140

29. Way LW (1992) Bile duct injury during laparoscopic cholecystecto-my. Ann Surg 215: 195

30. Wilder J (1964) Common methods of surgical injury to the commonduct. In: Atlas of general surgery. CV Mosby, St Louis, pp 98

31. Wilson TG, Hall JC, McWatts J (1986) Is operative cholangiographyalways necessary? Br J Surg 73: 637–640

32. Zucker KA, Bailey RW, Gadacz TR, Imbembo AL (1991) Laparo-scopic guided cholecystectomy: a plea for cautious enthusiasm. Am JSurg 161: 36–44

138

AbstractsSurgical

Endoscopy© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 189–218

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

Pneumothorax as a complication of laparoscopic inguinal hernia repair

G. S. Ferzli, T. Kiel, J. B. Hurwitz, P. Davidson, B. Piperno, M. A. Fiorillo, N. E. Hayek, L. L. H. Riina, P. Sayad

Department of Laparoendoscopic Surgery, Staten Island University Hospital, 78 Cromwell Avenue, Staten Island, NY 10304, USA

Received: 1 April 1996/Accepted: 13 June 1996

Abstract. Pneumothorax was identified as a complicationof endoscopic hernia repair in two patients with insufflationpressures of 15 mmHg and operating times exceeding 2 h.These patients also showed intraoperative perturbations inboth oxygen saturation and end-tidal CO2 production. Aprospective study was undertaken to determine whethersimilar complications would arise if preperitoneal insuffla-tion pressures were limited to 10 mmHg. Postoperativechest x-rays were obtained on all patients to check for pneu-mothoraces, even clinically occult ones. Fifty patients werestudied, with average operating times of 67 min. No patientdemonstrated any hemodynamic or ventilatory changes, andnone had any evidence of pneumothorax on x-ray. We con-clude that these complications were not present when insuf-flation pressure was maintained at 10 mmHg and that rou-tine x-ray is not warranted. Larger randomized trials ofinsufflation pressures are needed.

Key words: Pneumothorax — Laparoscopy — Inguinalherniorrhaphy

Successful surgical correction of inguinal hernias has beenpossible since the late 19th century [1]. Endoscopic extra-peritoneal herniorrhaphy (EEPH) has also been well de-scribed [4, 7, 13] and is considered an excellent alternativeto the traditional open procedure [11]. While complicationscan arise in both techniques, however, pneumothorax isunique to the endoscopic approach. This study discusses theetiology, treatment, and prevention of this condition, as wellas the necessity of routine postoperative chest x-rays.


Two sentinel cases involving dramatic ventilation abnormalities were iden-tified at our institution in early 1995. Details follow.

Case report 1

A 38-year-old male, ASA I, was admitted for elective bilateral inguinalhernia repair by endoscopic extraperitoneal technique. The induction ofanesthesia was uneventful. At the onset of surgery, O2 saturation (SO2) was97%, blood pressure was 130/70, heart rate was 80, and end-tidal carbondioxide (ETCO2) was 36 mmHg. At the completion of the case, the patientdesaturated to 84% and the ETCO2 increased to 48 mmHg. A chest x-rayobtained in the operating room revealed bilateral pneumothoraces as wellas pneumomediastinum. The patient remained intubated for 20 min, atwhich time a repeat x-ray showed a decrease in the extent of the pneumo-thoraces. He was extubated, and a third x-ray done in the recovery roomshowed complete resolution of the pneumothoraces. The patient was ad-mitted to the hospital for observation and discharged 24 h later. The du-ration of the procedure was 135 min, and the insufflation pressure withinthe preperitoneal space was 15 mmHg.

Case report 2

A 40-year-old male, ASA II, was admitted for elective repair of bilateralinguinal hernias using the endoscopic extraperitoneal technique. Anes-thetic induction was uneventful. At the onset of surgery, SO2 was 99%,blood pressure 140/80, heart rate 80, and ETCO2 27 mmHg. The preperi-toneal working space was maintained at a pressure of 15 mmHg. As thecase progressed, ETCO2 rose steadily and reached a peak of 49 mmHg.SO2 at this time was 92%. The patient was noted to have massive subcu-taneous emphysema, and a postoperative chest x-ray revealed pneumome-diastinum and a 50% right pneumothorax. He was observed for 3 h in therecovery room, after which a repeat chest x-ray showed resolution of thepneumothorax. He was admitted overnight for observation and dischargedthe next morning. The duration of surgery was 145 min.

The respiratory abnormalities described in the two cases above arepresented graphically in Fig. 1. Both patients were young male nonsmokerswithout any evidence of pulmonary disease. There was also no evidence ofdiaphragmatic or hiatal hernia. No break in the peritoneum was notedduring either procedure. The recognition that pneumothorax occurred inthese two patients with preperitoneal insufflation pressures of 15 mmHgprompted a prospective study involving 50 consecutive patients whosepressures were set at 10 mmHg. The study was performed between July 1and December 31, 1995. A total of 56 type II and III Nyhus-classifiedhernias were repaired. Patients ranged in age from 18 to 75 years (median42). There were six bilateral hernias. The endoscopic extraperitoneal ap-proach previously described by us was used exclusively [5]. All patientsreceived the same type of anesthesia. The following measurements wererecorded intra- and postoperatively (where applicable): end-tidal CO2, oxy-gen saturation, blood pressure, and heart rate. A postoperative chest x-raywas performed in all patients.Correspondence to:G. S. Ferzli

Surg Endosc (1997) 11: 152–153

SurgicalEndoscopy


Results

In this prospective study, 50 consecutive patients underwentsuccessful EEPH with an average operating time of 67 min.There was no mortality and no morbidity. No patient dem-onstrated any significant fluctuation in either ETCO2 or SO2intraoperatively, and none exhibited any postoperative de-compensation in SO2. Vital signs were likewise stable bothintra- and postoperatively. Of the 50 patients, eight experi-enced episodes of vomiting and retching in the recoveryroom. Postoperative chest x-rays showed no evidence ofpneumothorax or of pneumomediastinum. There was nopostoperative subcutaneous emphysema. All patients wereseen by the operating surgeon at 1, 3, and 6 weeks postop-eratively, and were without complications or recurrence.

Discussion

Pneumothorax is a known complication of laparoscopic sur-gery [2, 6, 10, 12]. Suspected etiologies include increasedintra-abdominal pressure as a result of vomiting or retching[3], direct diaphragmatic injury, an infiltration of insufflatedgas between the pleural or pericardial sacs, and spontaneousrupture of a preexisting bulla. It is also reported that incertain individuals potential channels exist between theperitoneal cavity and the pleural or pericardial sacs [9].Another hypothesis postulates that gas may enter the retro-peritoneal space and then migrate along tissue planes intothe mediastinum or pleural spaces [12].

Pneumothorax has not been previously reported duringEEPH. However, a recent article in the anesthesia literature[8] describes endoscopic extraperitoneal hernia repair hav-ing been complicated by a sudden increase in ETCO2 andthe development of massive subcutaneous emphysemawithout hemodynamic instability. This was felt to be due inpart to insufflation pressures that were kept between 14 and18 mmHg.

In both cases we have reported here, pressures of 15mmHg were maintained. In addition, both operating timesexceeded 2 h, whereas no operation in the prospecitve groupexceeded 90 min. In the prospective patients whose insuf-flation pressures were not allowed to exceed 10 mmHgthere were no hemodynamic or ventilatory fluctuations in-tra- or postoperatively, and postoperative chest x-raysshowed no pneumothoraces. It is our premise that pneumo-thorax as a complication of EEPH is most likely related tohigh insufflation pressures, as well as, possibly, to thelength of the procedure. A larger randomized trial of insuf-flation pressures would help to clarify this issue, particularlysince operating times cannot be controlled. We can con-clude with greater certainty, however, that changes shouldnot be anticipated in either SO2 or ETCO2 when pressuresare maintained at 10 mmHg or less, and that routine post-operative chest x-ray should be considered needless. Ifpneumothorax does develop, conservative treatment withmonitoring of SO2 is all that is indicated; the condition canbe expected to resolve without the placement of a chesttube.

References

1. Bassini E (1887) Suala Crura radicale dell’ernia inguinale. Arch SocItal Chir 4: 380

2. Batra MS, Driscoll JJ, Coburn WA, Marks WM (1983) Evanescentnitrous oxide pneumothorax after laparoscopy. Anesth Analg 62:1121–1123

3. Bremner WG, Kumar CM (1993) Delayed surgical emphysema, pneu-momediastinum and bilateral pneumothoraces after postoperativevomiting. Br J Anaesth 71: 296–297

4. Corbitt JD (1991) Laparoscopic herniorrhaphy. Surg Laparosc Endosc1: 23–25

5. Ferzli GS, Massaad AA, Dysarz F, Kopatsis A (1993) A study of 101patients treated with extraperitoneal endoscopic laparoscopic hernior-rhaphy. Am Surg 59: 707–708

6. Gabbott DA, Dunkley AB, Roberts FL (1992) Carbon dioxide pneu-mothorax occurring during laparoscopic cholecystectomy. Anaesthesia47: 587–588

7. Ger R, Mishrick A, Hurwitz J, Romero C, Oddsen R (1993) Manage-ment of groin hernias by laparoscopy. World J Surg 1: 46

8. Klopfenstein CE, Gaggero G, Mammie C, Morel P, Foster A (1995)Laparoscopic extraperitoneal inguinal hernia repair complicated bysubcutaneous emphysema. Can J Anaesth 42(6): 523–525

9. Laws HL, Fox LS, Younger JB (1977) Bilateral catamenial pneumo-thorax. Arch Surg 112: 627–628

10. MacFadyen B, Arregui M, Corbitt J, Filipi C, Fitzgibbons R, FranklinM, McKernan B, Olsen D, Phillips E, Rosenthal D, Schultz L, SewellR, Smoot R, Spaw A, Toy F, Waddell R, Zucker K (1993) Compli-cations of laparoscopic hernia. Surg Endosc 7: 155–158

11. Massaad AA, Fiorillo MA, Hallak A, Ferzli GS (1996) Endoscopicextraperitoneal herniorrhaphy in 316 patients. J Laparoendosc Surg6(1): 13–15

12. Murray DP, Rankin RA, Lackey C (1984) Bilateral pneumothoracescomplicating peritoneoscopy. Gastrointest Endosc 30: 45–46

13. Nyhus LM, Pollack R, Bruceck CT, Donahue PE (1988) The preperi-toneal approach and prosthetic buttress repair for recurrent hernia. AnnSurg 208: 783

Fig. 1. Ventilatory changes in two cases of pneumothorax during EEPH.

153

Laparoscopic surgery for lower gastrointestinal fistulas

J. S. Joo, F. Agachan, S. D. Wexner

Department of Colorectal Surgery, Cleveland Clinic Florida, 3000 W. Cypress Creek Road, Fort Lauderdale, FL 33309, USA

Received: 30 April 1996/Accepted: 5 July 1996

AbstractBackground: Increased experience and improved instru-mentation have lead to a reduction in morbidity and a com-mensurate increase in the spectrum of laparoscopic indica-tions. The purpose of this study was to assess the feasibilityof laparoscopic surgery in patients with gastrointestinal fis-tulas.Methods:Between March 1993 and March 1995, patientswith gastrointestinal fistulas who were laparoscopicallytreated were analyzed for age, gender, diagnosis, type ofprocedure, operative time, conversion rate, length of post-operative hospitalization, time until oral intake and return ofbowel function, morbidity, and mortality.Results:Ten patients (five females; five males) with a meanage of 49.7 (range 20–86) years were preoperatively diag-nosed as having the following fistulas: colocutaneous fistuladue to diverticulitis (one), enterocolic fistula (two)—due toCrohn’s ileocolitis (one) and due to diverticulitis (one)—pouchvaginal fistula after restorative proctocolectomy forfamilial adenomatous polyposis (two), colofallopian fistuladue to diverticulitis (one), rectourethral fistula due toCrohn’s disease (one), high transsphincteric fistula due toperianal Crohn’s disease (one), enteroenteric fistula due toCrohn’s disease (one), and colovesical fistula due to diver-ticulitis (one). Procedures performed consisted of sigmoid-ectomy with coloproctostomy (four), ileocolic resection(two), small-bowel resection with ileostomy (one), and di-verting loop ileostomy (three). A complex jejunal enterot-omy was noted in one (10%) patient. The mean operativetime was 195 (range 75–360) min and mean postoperativehospital stay was 6.1 (range 3–12) days. Two additionalcases were converted to open procedures for extensive dis-ease (one) and adhesions (one). The patients started oralintake after a mean of 2.2 (range 1–5) days and bowelfunction returned after a mean of 3.4 (range 2–7) days. One

patient required laparotomy on postoperative day 7 for amalrotated loop ileostomy.Conclusions:Laparoscopic colorectal surgery is feasible inpatients with simple lower gastrointestinal fistulas. Themorbidity rate of 10% and length of hospitalization of 6days are similar to results after laparoscopic procedures for‘‘simpler’’ colorectal pathology. However, the 30% conver-sion rate is higher, attesting to the challenging nature ofthese conditions.

Key words: Laparoscopy — Colovesical fistula — Colo-cutaneous fistula — Pouchvaginal fistula — Enteroentericfistula — Colofallopian fistula — Rectourethral fistula

Laparoscopy is likely to have an increasing impact on sur-gical procedures that have previously required laparotomy.Since the first laparoscopic cholecystectomy, the scope oflaparoscopic surgery has increased as have experience andinstrumentation [5, 10, 12, 14, 16]. Laparoscopic colon andrectal surgery is widely practiced [3, 6–8, 19, 23–26]. How-ever, most of the cases reported in these series includedeither ‘‘simple’’ benign or malignant pathology rather than‘‘complex’’ problems such as enteric fistulas [4]. Some casereports detailing the laparoscopic management of fistulashave included cholecystoduodenal fistulas [20], appendo-vesical fistulas [1], colovesical fistulas [18], vesicovaginalfistulas [13], and lymphatic fistulas [11]. This current reportassesses the feasibility, safety, and outcome of laparoscopicsurgery in a series of patients with lower gastrointestinalfistulas.

Materials and methodsBetween March 1993 and March 1995, all patients with simple lowergastrointestinal fistulas without disseminated sepsis were treated by a lap-aroscopic approach. Exclusion criteria were sepsis with free intraabdomi-nal purulence, generalized peritonitis, systemic sepsis, or an emergencypresentation. Preoperative evaluation included fistulography in cases ofenterocutaneous or colocutaneous fistula, contrast enema for coloenteric orcolovesical fistula, pouchogram and vaginogram for pouch vaginal fistula,and cystogram and cystoscopy for colovesical fistula. In addition, abdomi-nal computerized tomography (CT) scan was used in all patients exceptthose with pouch-vaginal fistulas. Exclusions for complex fistulas were


Correspondence to:S. D. Wexner

Surg Endosc (1997) 11: 116–118

SurgicalEndoscopy


based upon these results. A simple fistula was defined as a single commu-nication to the intestinal tract and a single external opening or a single tractbetween two intestinal segments.

All patients underwent conventional preoperative mechanical oralbowel preparation and received both oral and parenteral antibiotic prophy-laxis. Urinary catheters were utilized in cases of a CT-scan-proven phleg-mon adjacent to the ureter or in cases of colovesical fistulas. All patientshad an intraoperative indwelling urinary catheter and all patients wereoperated upon in the modified Lloyd-Davies position. A standard proce-dure was developed following the establishment of pneumoperitoneumwith either a conventional Veress needle or a Hassan technique. A peri-umbilical 10/12-mm disposable port was used as the initial camera site andthe remaining two to four 10/12-mm ports were inserted according to thepatient’s body habitus and the site of the surgery. An alternative techniquefor initial port placement was the Visiport (US Surgical Corporation Nor-walk, CT). Before approaching the site of pathology, adhesiolysis andconfirmation of the anatomy were undertaken. Dissection and mobilizationwere accomplished using a mixture of sharp and blunt dissection with a10-mm-diameter electrocautery scissors. Great care was taken to maintainan excellent video image and optimal hemostasis throughout the dissectionand mobilization. These goals were fulfilled by using an experienced sur-gical team. When the location of the fistula tract was on the right side ofthe intestine, a 3–5-cm trans- or periumbilical incision was made from thecamera port site and the diseased bowel was extracorporealized for resec-tion and anastomosis after full dissection around the main disease siteincluding the fistula tract. When the disease site was on the left, completeintracorporeal sigmoid colectomy and double-stapled coloproctostomywere undertaken, as previously described [9]. In the case of colovesicalfistula, the tract was pinched off with clamps, and, as in the ‘‘open’’setting, the bladder was drained with a catheter for 7 days; the bladderdefect was not closed. When the fistula was a rectourethral or pouchvaginal fistula, stoma construction was performed as a definitive or adjunctprocedure as described elsewhere [15]. The procedure was converted to alaparotomy if any single phase failed to adequately progress within 30–60min or if the entire procedure failed to progress within 2 h.

Results

Ten patients underwent laparoscopic treatment of lowergastrointestinal fistulas; the demographic characteristics, di-agnosis, and procedures are shown in Table 1. The meanage of these patients was 49.7 (range 20–86) years. Therewere five females and five males. Pathology included onecolocutaneous fistula due to diverticulitis, one enterocolicfistula due to Crohn’s ileocolitis and one due to diverticu-

litis, two pouch-vaginal fistulas after restorative procto-colectomy for familial adenomatous polyposis, one colofal-lopian fistula due to diverticulitis, one rectourethral fistuladue to Crohn’s disease, one high transsphincteric fistula dueto perianal Crohn’s disease, one enteroenteric fistula due toCrohn’s disease and one colovesical fistula due to diver-ticulitis. Procedures consisted of sigmoidectomy and colo-proctostomy (four), ileocolic resection (two), small-bowelresection with ileostomy (one), and diverting loop ileosto-my (three).

The mean operating time was 195 (range 75–360) minand intraoperative complications were limited to one ente-rotomy described below (10%). There was no postoperativemortality. Patients commenced oral intake after a mean of2.2 (range 1–5) days, resolved ileus at a mean of 3.4 (range2–7) days, and were discharged at a mean of 6.1 (range3–12) days. Discharge was often delayed because of in-dwelling bladder catheterization or requirement for enter-ostomal education. The overall conversion rate was 30%.The reasons for conversion to open surgery included a com-plex jejunal enterotomy made with use of the Visiport in apatient with severe adhesions, after restorative proctocolec-tomy with an ileal J pouch due to familial adenomatouspolyposis in one patient and an undissectable inflammatorymass due to Crohn’s ileocolitis with complex enterocolicand enteropsoas fistula in a second patient. The first patienthad recently undergone a single-stage restorative procto-colectomy for familial adenomatous polyposis and an asso-ciated mesenteric desmoid tumor. On the 7th postoperativeday she was noted to have a pouchvaginal fistula. This is theonly patient in whom the Visiport was used. Unfortunatelya complex jejunal enterotomy was created during directcamera-guided placement of that port. The second patientrequired laparotomy on postoperative day 7 due to a mal-rotated loop ileostomy. This patient had developed a pouch-vaginal fistula 7 days after a single-stage restorative proc-tocolectomy for familial adenomatous polyposis. Extensiveadhesions at the time of laparoscopic ileostomy creation onpostoperative day 7 precluded optimal confirmation of limb

Table 1. Demographic characteristics of 10 patients

Age(years) Sex Indication Underlying disease Procedures Comments

78 F Colocutaneous fistula Diverticulitis Sigmoid colectomy Success51 M Rectourethral fistula Crohn’s disease Loop ileostomy Success41 M Complex enterocolic and Crohn’s disease Ileocolic resection Converteda

enteropsoas fistulas32 F Pouch-vaginal fistula Familial adenomatous

polyposisLoop ileostomy with extensive enterolysis Reoperatedb

50 F Colofallopian fistula Diverticulitis Sigmoid colectomy Successc

86 F Enterocolic fistula Diverticulitis Sigmoid colectomy, small bowel resection Convertedd

with loop ileostomy34 F Pouch-vaginal fistula Familial adenomatous

polyposisSmall-bowel resection with loop ileostomy Convertedc

39 M High transsphinctericfistula

Perianal Crohn’sdisease

Loop ileostomy Success

20 M Enteroenteric fistula Crohn’s disease Ileocolic resection Success58 M Colovesical fistula Diverticulitis Sigmoid colectomy with loop ileostomy Success

a Converted due to extensive purulenceb Patient reoperated upon on postoperative day 7 for malrotation of ileostomy limb (see text)c Convert due to iatrogenic complex jejunal enterotomy with the Visiport (United States Surgical Corp., Norwalk, CT)d Converted due to extensive fibrosis associated with an enterocolic fistula and a phlegmon; unable to maintain optimal visualization

117

orientation. After laparoscopic stoma creation, subsequentrecovery was uneventful.

Discussion

Laparoscopic surgery has changed the practice of surgery[21]. Both the indications for laparoscopic colorectal sur-gery and the surgical procedures performed are continu-ously evolving as the experience of the surgeons and so-phistication of the available instrumentation develop. Dur-ing our first 66 cases of laparoscopic surgery for colorectaldisease, lower gastrointestinal fistulas were considered as acontraindication. However, as our experience increased andinstrumentation improved, we began to apply this procedureto these indications [2].

The experience described in this review of our prelimi-nary experience in the use of laparoscopic technology tofacilitate surgery for lower gastrointestinal fistulas has con-vinced us of the feasibility of this approach. Proceduresperformed ranged from ‘‘simpler’’ ones such as creation ofa stoma to complex laparoscopic colorectal procedures suchas mobilization and resection of an enteric fistula sur-rounded by inflammation. The difficulty of laparoscopicprocedures for these cases is well recognized. These proce-dures were converted to laparotomy due to injury, adhe-sions, and sepsis. Nonetheless, the procedure was success-fully undertaken using laparoscopic techniques in 70% ofcases with a 10% incidence of reoperation and no mortality.Morbidity rates and lengths of hospitalization after laparo-scopic colectomy for nonfistulous conditions ranged from 0to 45% and from 4 to 12 days, respectively [22].

Some case reports of the laparoscopic management offistulas exist. Puente et al. [18] reported two patients withcolovesical fistulas due to diverticulitis treated by laparo-scopic surgery with a mean hospital stay of 5 days and amean procedure duration of 5.5 h. Prasad and Foley [17]also reported one successful laparoscopic cholecystectomyfor a cholecystocolic fistula, as did Afifi et al. [1] with onecase of appendovesical fistula. Sharma et al. [20] reportedfive cases of cholecystoduodenal fistula with a mean hos-pital stay of 7.5 days and mean procedure duration of 2.9 h,similar to the current report. The mean 6 days of hospital-ization also confirms the expectation that laparoscopic tech-niques allow relatively short hospitalization. In fact, hospi-talization was often prolonged to allow indwelling bladdercatheterization or stoma education and management. Be-cause of the short postoperative stay, an important consid-eration is the necessity for stoma therapists to intensivelytrain within 48 h after surgery to expedite discharge fromhospital.

The 30% conversion rate is much higher than is ouroverall conversion rate of 11% for noninflammatory disor-ders [25] but will hopefully decrease as our experience inlaparoscopic management of these challenging conditionsincreases.

Conclusion

In conclusion, laparoscopic colorectal surgery is feasible inpatients with lower gastrointestinal fistulas. An experiencedteam providing optimal video imaging and hemostasis areimperative to attain these results. The morbidity rate of 10%

and length of hospitalization of 6 days are similar to thosefigures cited after laparoscopic procedures for ‘‘simpler’’colorectal pathology. However, the 30% conversion rate ishigher, attesting to the challenging nature of these condi-tions.

References1. Afifi AY, Fusia TJ, Feucht K, Paluzzi MW (1994) Laparoscopic treat-

ment of appendicovesical fistula: a case report. Surg Laparosc Endosc4: 320–324

2. Agachan F, Gilliand R, Joo JS, Sher M, Wexner SD (1996) The impactof experience on the outcome of laparoscopic colorectal surgery overfour years. (Abstract). Surg Endosc 10: 219

3. Agachan F, Joo JS, Weiss EG, Wexner SD (1996) Intraoperative lap-aroscopic complications: what is the learning curve? Presented at the96th Annual meeting of the American Society of Colon and RectalSurgeons. Seattle, Washington, 9–14 June 1996

4. Agachan F, Joo JS, Weiss EG, Wexner SD (1996) Role of laparoscopyin colorectal surgery: a prospective evaluation of 172 cases. Presentedat the 37th annual meeting of the Society for Surgery of the AlimentaryTract. San Francisco, California 19–22 May 1996

5. Cuschieri A, Shimi S, Nathanson LK (1992) Laparoscopic reduction,crural repair and fundoplication of large hiatal hernia. Am J Surg 163:425–430

6. Falk PM, Beart RW Jr, Wexner SD, Thorson AG, Jagelman DG,Lavery IC, Johansen OB, Fitzgibbons RJ Jr (1993) Laparoscopic col-ectomy: a critical appraisal. Dis Colon Rectum 36: 28–34

7. Guillou PJ, Darzi A, Monson JR (1993) Experience with laparoscopiccolorectal surgery for malignant disease. Surg Oncol 2(Suppl): 43–49

8. Jacobs M, Verdeja JC, Goldstein HS (1991) Minimally invasive colonresection. Surg Laparosc Endosc 1: 144–150

9. Jager R, Wexner SD (1996) Laparoscopic colorectal surgery. Chur-chill-Livingstone, New York, pp 187–190

10. Kathouda N, Mouiel J (1991) A new technique of surgical treatment ofchronic duodenal ulcer without laparotomy by videocoelioscopy. AmJ Surg 161: 361–364

11. Khauli RB, Mosenthal AC, Caushaj PF (1992) Treatment of lympho-cele and lymphatic fistula following rectal transplantation by laparo-scopic peritoneal window. J Urol 147: 1353–1355

12. Nezhat CR, Burrell MO, Nezhat FR, Benigno BB, Welander CE(1992) Laparoscopic radical hysterectomy with para-aortic and pelvicnode dissection. Am J Obstet Gynecol 166: 864–865

13. Nezhat CH, Nezhat F, Rottenberg H (1994) Laparoscopic repair of avesicovaginal fistula: a case report. Obstet Gynecol 83: 899–901

14. Nyhus LM (1992) Laparoscopic hernia repair: a point of view. ArchSurg 127: 137

15. Oliveira L, Reissman P, Nogueras JJ, Wexner SD (1996) Laparoscopiccreation of stomas (Abstract). Surg Endosc 10: 209

16. Paget GW (1992) Laparoscopic repair of inguinal hernia. Med J Aust156: 508–510

17. Prasad A, Foley RJ (1994) Laparoscopic management of cholecystolicfistula. Br J Surg 81: 1789–1790

18. Puente I, Sosa JL, Desai U, Sleeman D, Hartmann R (1994) Laparo-scopic treatment of colovesical fistulas: technique and report of twocases. Surg Laparosc Endosc 4: 157–160

19. Schlinkert RT (1991) Laparoscopic assisted right hemicolectomy. DisColon Rectum 34: 1030–1031

20. Sharma A, Sullivan M, English H, Foley R (1994) Laparoscopic repairof cholecystoduodenal fistulae. Surg Laparosc Endosc 4: 433–435

21. Tompkins RK (1990) Laparoscopic cholecystectomy: threat or oppor-tunity? Arch Surg 125: 1245

22. Teoh T-A, Wexner SD (1996) Laparoscopic surgery in colorectal can-cer. In: Williams N (ed) Colorectal cancer. Churchill-Livingstone, Ed-inburgh, pp 103–121

23. Wexner SD, Reissman P (1994) Laparoscopic colorectal surgery: aprovocative critique. Int Surg 79: 235–239

24. Wexner SD, Cohen SM (1995) Port site metastases after laparoscopiccolorectal surgery for cure of malignancy. Br J Surg 82: 295–298

25. Wexner SD, Cohen SM, Johansen OB, Nogueras JJ, Jagelman DG(1993) Laparoscopic colorectal surgery: a prospective assessment andcurrent perspective. Br J Surg 80: 1602–1605

26. Wexner SD, Reissman P, Pfeifer J, Berstein M, Geron N (1996) Lap-aroscopic colorectal surgery: analysis of 140 cases. Surg Endosc 10:133–136

118

Total bilateral laparoscopic adrenalectomy in patients with Cushing’ssyndrome and multiple endocrine neoplasia (IIa)

L. Fernandez-Cruz,1 A. Saenz,1 G. Benarroch,1 L. Sabater,1 P. Taura2

1 Department of Surgery, Hospital Clinıć, University of Barcelona, C/ Villarroel, 170, 08036-Barcelona, Spain2 Department of Anaesthesiology, Hospital Clinic, University of Barcelona, C/ Villarroel, 170, 08036-Barcelona, Spain

Received: 22 March 1996/Accepted: 7 July 1996

AbstractBackground:The benefit of simultaneous bilateral laparo-scopic adrenalectomy in patients with Cushing’s syndromeand pheochromocytoma associated with multiple endocrineneoplasia (MEN) is unknown.Methods:Ten patients underwent laparoscopic adrenalecto-my (LpA) with CO2 pneumoperitoneum for Cushing’s syn-drome. One MEN patient underwent simultaneous bilaterallaparoscopic adrenalectomy with helium pneumoperito-neum for bilateral pheochromocytoma. A comparison wasmade between unilateral LpA and simultaneous bilaterallaparoscopic adrenalectomy in patients with Cushing’s syn-drome. Plasma catecholamines were correlated with hemo-dynamic changes in the patient with pheochromocytoma.Results:Simultaneous bilateral laparoscopic adrenalectomyin the patient with pheochromocytoma lasted 330 min. Thesubstantial increase in plasma catecholamines was not as-sociated with cardiovascular instability. Operative time(270 ± 3 vs 120 ± 4min), blood loss (365 ± 1 vs 210 ± 1ml), hospital stay (7.6 ± 1.5 vs 4.6 ± 1 days), and normalactivity (19.3 ± 2 vs10.4 ± 4.4 days) were, in patients withCushing’s syndrome, significantly (p < 0.05) higher aftersimultaneous bilateral laparoscopic adrenalectomy than af-ter unilateral LpA; the differences were not significant inthe analgesic requirements (7.6 ± 1 vs 4.6 ± 1doses). Onepatient with unilateral LpA was converted to open surgery.Conclusion:Simultaneous bilateral laparoscopic adrenalec-tomy is safe, and associated with short hospital stay andlessening of the time needed to achieve normal activity.

Key words: Bilateral laparoscopic adrenalectomy — Cush-ing’s syndrome — Pheochromocytoma

Adrenalectomy as a surgical procedure is ideally suited tominimally invasive surgery based on the well-defined bloodsupply of the adrenal and the characteristics of the tumors,which are usually small and benign. However, the partiallyretrocaval and retrohepatic location of the right gland andthe para-aortic retropancreatic and justa-splenic position ofthe left, mean dissection is not always easy, and sometimesthe distinction between the perirenal and the periadrenal fatand the adrenal gland may be difficult to determine on thebasis of color differences. With all these advantages anddisadvantages in mind, Gagner et al. in 1992 [4], and sub-sequently other groups, reported clinical success in the sur-gical management of a variety of benign nonfunctioningand functioning adrenal tumors [2, 5, 6, 10, 13, 14]. In themajority of the reports the surgical experience has beenconcentrated on unilateral adrenal tumors, and few havereported the laparoscopic management of bilateral adrenaldisorders [2, 5].

The aim of this report is to describe simultaneous bilat-eral laparoscopic adrenalectomy in patients with Cushing’ssyndrome after having failed hypophysectomy. This groupwill be compared with patients with Cushing’s adenomaundergoing unilateral laparoscopic adrenalectomy (LpA).Also, we include in this report the use of simultaneousbilateral laparoscopic adrenalectomy in one patient with bi-lateral pheochromocytoma associated with multiple endo-crine neoplasia (MEN) type IIa. Particular attention wasgiven to the influence of pneumoperitoneum and the lengthysurgical procedure on catecholamine release and subsequentintraoperative haemodynamic changes.

Patients and methods

Between June 1992 and February 1996, 37 patients underwent 42 laparo-scopic adrenalectomies. From this group there were 10 patients with Cush-ing’s syndrome: four women, mean age 44.8 ± 20.3 years, ranging from 34to 72 years, had Cushing’s disease after having once failed hypophysec-tomy, and six patients (four women and two men), mean age 47.4 ± 24.7years, ranging from 31 to 74 years, had Cushing’s adenoma. Eleven pa-Correspondence to:L. Fernandez-Cruz

Surg Endosc (1997) 11: 103–107

SurgicalEndoscopy


tients (mean age 39.3 ± 7.2 years, ranging from 25 to 50 years) hadpheochromocytoma; 10 had unilateral tumors, and one 26-year-old womanhad bilateral tumors associated with MEN type IIa.

Preoperative management in patients with Cushing’s syndrome in-cluded the administration of 100 mg of corticosteroids prior to surgery, andin patients with pheochromocytoma the consumption of phenoxybenza-mine (20 mg/24 h) for 5 days and propranolol (20 mg/12 h) for 1 day priorto surgery.

Plasma catecholamines were correlated with intraoperative hemody-namic changes in the patient with bilateral pheochromocytoma. CO2 pneu-moperitoneum was used in all patients except in patients with pheochro-mocytoma, where helium was used. A straight telescope (0°) attached to avideo camera or a 30°-angle laparoscope was inserted through a 10-mmcannula.

The patient with bilateral pheochromocytoma was in the supine posi-tion, and exaggerated lordosis was produced with the legs apart and thetable extended. A firm pillow was placed beneath the dorsal region, and theoperating table was turned to the right or to the left to remove the tumorsin each side. Six trocars were then inserted as shown in Fig. 1. The generallaparoscopic steps that were performed in this patient were: (1) Operativeexposure of the right adrenal gland with retraction of the right lobe of theliver anteriorly. (2) The overlying peritoneum lateral to the duodenum wasincised and the anterior and right side of the vena cava was freed. (3) Therenal vein was identified as a landmark and the dissection then proceeded

in a cephalad direction, visualizing the adrenal veins which issue directlyfrom the vena cava in a parallel manner. Attention was paid to adrenalveins entering the vena cava from the superior portion of the adrenal glandand arterial branches from the inferior phrenic artery. (4) All vessels weresealed with two clips. (5) Operative exposure of the left adrenal gland byinferior reflection of the splenic flexure of the colon and division of theavascular splenorenal ligament. (6) The spleen was mobilized by means ofa lateral peritoneal incision up to the superior border, allowing mobilizationof the distal pancreas so that it could be elevated in a cephalad manner toexpose Gerota’s fascia. (7) Exposition for clipping and division of the leftadrenal vein entering the superior part of the adrenal vein and the arterialtwigs from the phrenic artery, the renal artery, and the aorta. (8) The rightand left adrenal glands were then gently extracted via the umbilical inci-sion, which was increased to 5 cm in a circular fashion to enable easierremoval of the tumors.

In patients with Cushing’s adenoma, a lateral decubitus approach wasused (Figs. 2 and 3); and in patients undergoing simultaneous bilaterallaparoscopic adrenalectomy it was necessary to change the patient from thelateral-decubitus on one side to the same position on the other side. Fourtrocars were used on each side. The following steps were performed for aright adrenalectomy: (1) The right triangular ligament of the liver wassectioned, the liver was retracted medially, and the vena cava was exposed.(2) Dissection of the superior and anterior aspect of the gland; the vesselswere secured with two clips and transected. (3) Gentle retraction of the

Fig. 1. Supine half-lateral position for simultaneous totalbilateral adrenalectomy. Position of the trocars.Fig. 2. Lateral decubitus approach. Position of the trocars forright adrenalectomy.Fig. 3. Lateral decubitus approach. Position of the trocars forleft adrenalectomy.

104

gland to the right with clipping and sectioning of the vessels. For a leftadrenalectomy: (1) Division of the splenocolic ligament with retraction ofthe spleen and tail of the pancreas to the midline. (2) Identification of therenal vein and clipping and sectioning of the adrenal veins. (3) The superiorpole of the gland was dissected and the left adrenal arteries were securedwith clips and transected.

After right or left adrenalectomy, a sterile plastic bag was used foradrenal removal through the anterior trocar site.

Systemic postoperative analgesia was provided by ketorolac (SyntexLatino, S.A., Barcelona, Spain).

The following parameters were evaluated in each patient with Cush-ing’s syndrome and pheochromocytoma: operative time, estimated bloodloss, length of hospital stay, doses of postoperative pain medication, andthe time needed to achieve normal activity.

For the analysis of all results obtained, standard methods of statisticalanalysis were used. Group values are expressed as mean ± SD (range).Paired tests and nonpaired Mann-Whitney U tests were used to evaluatedifferences between groups. Ap value of less than 0.05 was consideredsignificant.

Results

In the patient with bilateral pheochromocytoma, the dura-tion of total adrenalectomy was 330 min. Plasma epineph-

rine (EPI) and norepinephrine (NE) increased sharply afterinduction of anesthesia and reached very high levels at thetime of tumor isolation on both sides. Catecholamine levelsdecreased substantially at the end of the procedure. A dropin cardiac output, oxygen delivery, and oxygen consump-tion was observed after induction of anesthesia, but thesewere normal thereafter. The systemic vascular resistanceindex was markedly elevated after induction of anesthesiabut remained within normal limits throughout the operation.PaCO2 monitoring was constant in showing normal values(Table 1). Blood loss was noted amounting to 440 ml andtransfusion was not necessary. The total amount of analge-sic requirement was 90 mg. The patient was discharged 4days after the operation with substitutive therapy.

In the comparative study between patients with Cush-ing’s disease and Cushing’s adenoma, the operative timewas significantly higher in patients undergoing simulta-neous bilateral laparoscopic adrenalectomy, 270 ± 30.24min, than in patients undergoing unilateral LpA, 120 ±43.47 min. The blood loss was significantly greater in pa-tients with simultaneous bilateral laparoscopic adrenalecto-

Table 1. Hormone and hemodynamic parameters during bilateral total adrenalectomy using helium pneumoperitoneuma

Epineph-rine

Norepi-nephrine Dopamine pH PaCO2 B.E. H.R. MAP CVP PAMP PAWP CI SVRI DO2 VO2 O2(a-v)C

Basal 177 175 24Anesthesia

induction 2,000 2,661 100 7.46 30.2 2.4 70 79 9 17 13 2.51 2,331 411 104 5.130 min post-

induction 537 888 218 7.45 32.1 0.4 83 80 10 16 12 2.35 1,939 397 90 3.9Left tumor

isolation 3,464 3,540 344 7.39 36.0 −1.5 93 87 9 20 11 4.32 1,446 723 107 2.6Right tumor

isolation 5,501 6,059 390 7.39 34.3 −2.9 98 102 11 21 16 5.5 1,280 668 106 2.5Trocar

removal 109 632 181 7.34 37.8 −3.5 85 85 11 21 13 5.8 1,199 645 125 2.6

a Epinephrine: NV < 150 pg/ml, norepinephrine: NV < 400 pg/ml, dopamine: NV < 10 pg/ml, PaCO2: mmHg, B.E.: base excess, mEq/l, H.R.: heart rate,bpm, MAP: mean arterial pressure, mmHg, CVP: central venous pressure, mmHg, PAMP: pulmonary artery mean pressure, mmHg, PAWP: pulmo-nary artery wedge pressure, mmHg, CI: cardiac index, l/(min/m2), SVRI: systemic vascular resistance index, dyne-s/cm−5-m2, DO2: O2 delivery, ml/min/m2,VO2: O2 consumption, ml/min/m2, O2 (a-v) C: arteriovenous O2 content, ml/dl

Table 2. Laparoscopic adrenalectomy: characteristics, diagnosis, and intraoperative and postoperativeparameters (mean ± SD [range])

Cushing’s disease Cushing’s adenoma

No. patients 4 6Age (years) 44.83 ± 20.32 (34–72) 47.42 ± 24.75 (31–74)Sex (male/female) —/4 2/4Preoperative cortisola

(mg/dl) 34.2 ± 7.06 (27.3–45.2) 30.88 ± 2.92 (27.8–36.2)Unilateral (right/left) — 4/2Bilateral 4 —Size (cm) 4.45 ± 1.94 (4.2–6) 4.4 ± 1.78 (2.5–6.5)Operative time (min) 270 ± 30.24 (215–330)* 120 ± 43.47 (60–165)Blood loss (ml) 365 ± 159.26 (210–580)* 210 ± 106.14 (110–410)Analgesic requirements

(number of dosesb) 7.66 ± 1.52 (6–9) 4.66 ± 1.21 (3–6)Hospital stay (day) 6 ± 2 (4–8)* 3 ± 0.63 (2–4)Normal activity (day) 19.33 ± 2.08 (17–21)* 10.42 ± 4.46 (7–16)Postoperative cortisola

(mg/dl) 3.33 ± 1.04 (2.24–5.02) 3.82 ± 1.55 (1.98–6.79)

a Plasma normal values at 8.00A.M.: 10–25mg/dlb One dose4 30 mg of ketorolac* p < 0.05, Mann-Whitney U test

105

my, 365 ± 159 ml, than in patients undergoing unilateralLpA, 210 ± 106 ml. There was no significant difference inthe analgesic requirements between the groups. However,the 3 days of hospital stay in the unilateral LpA group wassignificantly less than the 6 ± 2 days in patients after si-multaneous bilateral laparoscopic adrenalectomy. Thenumber of days required to return to a normal activity was19.33 ± 2 in patients after simultaneous bilateral laparo-scopic adrenalectomy. This was significantly longer thanthe time required in patients after unilateral LpA 10.42 ±4.46 (Table 2).

One patient with a Cushing’s adenoma 5 cm in diameter,needed conversion to open surgery for uncontrolled bleed-ing. Two patients undergoing simultaneous bilateral lapa-roscopic adrenalectomy had urinary infection. Plasma cor-tisol levels dropped to normal values in patients with Cush-ing’s adenoma and also in patients with Cushing’s diseasein which substitutive therapy was provided with hydrocor-tisone (30 mg/day) and fluodrocortisone acetate (0.1 mg/day). In both groups of patients a remission of the clinicalfeatures was achieved at a mean follow-up of 9.2 months,ranging from 2 to 16 months. Only the patients who had hadsimultaneous bilateral laparoscopic adrenalectomy weredischarged from the hospital with substitutive therapy.

Discussion

The anesthetic management and the surgical treatment ofpheochromocytoma have greatly improved over the last 15years. Most published series have emphasized that approxi-mately 10% of patients have bilateral metastatic or extra-adrenal tumors [12]. Thus, the multicentricity and anatomicvariability of these tumors have in all circumstances in thepast dictated an intraperitoneal operative approach. Thisstrategy allowed the surgeon to explore the contralateraladrenal gland as well as possible ectopic sites for neoplasticchromaffin tissue not identified in the preoperative evalua-tion. However, in recent years, the ability to localize a uni-lateral adrenal pheochromocytoma using 131I-MIBG scin-tigraphy to determine the size and shape of the contralateralgland by computed tomography (CT), while avoiding themorbidity of a transabdominal exploration of both adrenalbeds, has made the retroperitoneal approach the method ofchoice [7]. Patients with MEN IIa or MEN IIb who havebilateral macroscopic phenchromocytomas should, how-ever, have both adrenal glands removed.

For bilateral adrenalectomy, in 1936 H. Young [15] de-scribed a technique for simultaneous exposure and opera-tion on the adrenals; bilateral paraspinal incisions are placedand subperiosteal resection of the 12th rib is carried out oneach side with extraperitoneal and extrapleural dissection todivide Gerota’s fascia and expose the right and left adrenalglands. In 1955 Aird and Helman [1] described the anteriortransperitoneal approach with the advantages of simulta-neous bilateral adrenal exploration with a wider, safer ex-posure through a single incision. In 1990 Scott et al. [12]described the posterior approach through the bed of theresected 11th rib with a two-team approach with bilateralsimultaneous paraspinal incisions. In recent years, the ma-jority of surgeons have also favored the flank approach inunilateral and bilateral tumors, as a variant on the posteriorapproach in which the retroperitoneal space is entered afterthe 11th rib is resected [11].

All our patients with Cushing’s adenoma at the meanfollow-up of 9.2 months met the criteria for the ideal treat-ment: remission of the clinical and biochemical features,removal of all tumor tissue, and no permanent hormonedeficiency [9]. None of our patients with Cushing’s diseasedeveloped Nelson’s syndrome postoperatively. The overallprevalence of this complication after adrenalectomy is be-tween 15 and 29%, and the use of prophylactic radiotherapyfor its prevention remains controversial [8]. With the lapa-roscopic approach [2–4], both adrenal glands can be simul-taneously removed. This approach was performed transperi-toneally and in some circumstances—for example, in thepatient placed in supine half-lateral position with bilateralpheochromocytoma—it allows a careful exploration of theentire abdominal cavity. It may be useful in some patients inexcluding the presence of other pheochromocytoma. Thisposition also allows total bilateral adrenalectomy withoutthe need to change sides and reposition the patient. How-ever, the ability to preoperatively localize an adrenal pheo-chromocytoma and to rule out multifocal lesions with ra-diological and isotopic studies is so good that a direct uni-lateral approach for pheochromocytoma is warranted.Therefore, for most indications of total bilateral adrenalec-tomy, such as adrenal hyperplasia for Cushing’s disease andbilateral pheochromocytoma, we favor the lateral-decubitusapproach, which allows a direct approach to just one glandand results in less dissection than the anterior approach.

In patients with Cushing’s syndrome the laparoscopicapproach shows more favorable results in unilateral LpAthan in simultaneous bilateral laparoscopic adrenalectomyfor the clinical parameters evaluated. In the latter patientsthe operation took longer, but no transfusion was necessary,the hospital stay was not longer than a week, and the timeneeded for this group of debilitated patients to return tonormal activity wasø3 weeks.

CO2 pneumoperitoneum was used in all patients in-cluded in this study except in a patient with bilateral pheo-chromocytoma in whom helium was used to prevent intra-operative cardiovascular derangements associated with hy-percarbia. In the later case it was noteworthy that inductionof anesthesia and even minimal laparoscopic explorationand isolation of the pheochromocytoma on each side wereassociated with the release of catecholamines without he-modynamic alterations. Furthermore, PaCO2 remained nor-mal throughout the 5 h and 30 min of the operation.

In conclusion, laparoscopic bilateral adrenalectomy canbe safely performed using CO2 or helium pneumoperito-neum in patients with adrenal hyperplasia for Cushing’sdisease and in bilateral pheochromocytoma. The techniquewe propose offers many advantages; the optimal magnifi-cation and enhanced resolution enable highly selective dis-section with near-perfect hemostasis and minimal tissuetrauma. This approach needs several hours to perform butthe benefit is clear in terms of hospital stay not longer than1 week and the short time needed for the patients to returnto normal activities.

References

1. Aird I, Helman P (1955) Bilateral anterior transabdominal adrenalec-tomy. Br Med J 2: 708–712

2. Fernańdez-Cruz L, Benarroch G, Torres E, Astudillo E, Saénz A,TauraP (1993) Laparoscopic approach to the adrenal tumors. J Lapa-roendosc Surg 3: 541–546

106

3. Fernańdez-Cruz L, Saénz A, Taura´ P, Benarroch G, Nies CH, AstudilloE (1994) Phaeochromocytoma: laparoscopic approach with CO2 andhelium pneumoperitoneum. End Surg 2: 300–304

4. Gagner M, Lacroix A, Bolte E (1992) Laparoscopic adrenalectomy inCushing’s syndrome and pheochromocytoma. New Engl J Med 327:1033

5. Gagner M, Lacroix A, Prinz R, Bolte E, Albala D, Potvin C, Hamet P,Kuchel O, Querin S, Pomp A (1993) Early experience with laparo-scopic approach for adrenalectomy. Surgery 114: 1120–1125

6. Higashihara E, Tanaka-Horie S, Aruga S, Nutahara K, Minowada S,Aso R (1993) Laparoscopic adrenalectomy: the initial 3 cases. J Urol149: 973–976

7. Irvin LG, Fishman ML, Sher AJ, Yeung KL, Ivani H (1989) Pheo-chromocytoma. Lateral versus anterior operative approach. Ann Surg209: 774–778

8. Jenkins PJ, Trainer J, Plowman N, Shand SW, Grossman BA, WassHAJ, Besser MG (1995) The long term outcome after adrenalectomyand prophylactic pituitary radiotherapy in adrenocorticotropin depen-dent Cushing’s syndrome. J Clin Endocrinol Metab 80: 165–171

9. Orth ND (1995) Cushing’s syndrome. N Engl J Med 332: 791–80310. Prinz RA (1995) A comparison of laparoscopic and open adrenalec-

tomies. Arch Surg 130: 489–49411. Proye CAG, Huart JH, Cuvillier XD, Assez ML, Gambardella B,

Carnaille BML (1993) Safety of the posterior approach in adrenalsurgery: experience in 105 cases. Surgery 114: 1126–1131

12. Scott HW Jr, Abumrad NN, McDougal WS (1990) Surgical ap-proaches for bilateral adrenalectomy and resection of benign and ma-lignant tumors. In: Scott HW Jr (ed), Surgery of the adrenal glands. JBLippincott, Philadelphia, pp 297–310

13. Suzuki K, Kageyama S, Ueda D, Ueda D, Ushiyama T, Kawabe K,Tajima A, Aso R (1993) Laparoscopic adrenalectomy: clinical expe-rience with 12 cases. J Urol 150: 1099–1102

14. Takeda M, Go H, Imai T, Nishiyama T, Morishita H (1993) Laparo-scopic adrenalectomy for primary aldosteronism: report of initial tencases. Surgery 115: 621–625

15. Young HH (1936) A technique for simultaneous exposure and opera-tion on the adrenals. Surg Gynecol Obstet 54: 179–188

107

AbstractsSurgical


170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

New technology

This section is designed to bring forward some of the latest innovative technology with explanations in termsthat will clarify their importance to the discipline of surgery. Through the efforts of the Innovative TechnologyCommittee of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES), leading experts invarious areas will be invited to present a summary of new technology, often including their pioneering work.

Image-guided minimally invasive therapy

R. M. M. Seibel

Institute of Diagnostic and Interventional Radiology, University Witten Herdecke, Mu¨lheim, Germany, Department of Radiology, St. Marien Hospital,Mulheim, Germany

Received: 9 September 1996/Accepted: 15 November 1996

Abstract. From its beginning surgical tomography devel-oped rapidly. The possibility to identify every structure in-side the body and to guide instruments with monitoring oftissue changes increase safety. Different therapies have beenperformed under CT and MRI guidance and evaluated instudies, like biopsies, tumor pain therapy and treatments oftumors as well as treatments of the spine. The results showthat surgical tomography is an accurate guiding technique.Especially in combination with micro endoscopy offers anew way of surgery. The complication rate as well as thefrequency of side effects is very low.

Key words: CT guidance — MRI guidance —Surgical to-mography — Tumor pain therapy — Ethanol therapy —Periradicular therapy — Nucleotomy — Micro endoscopy— Spinal scar dissection

Haaga and co-workers [10] performed the first CT guidedbiopsies a year after the introduction of diagnostic Com-puted Tomography. Indications for therapeutic interven-tions spread rapidly because of the safety of CT guidance.

In 1986 we started Minimally Invasive Therapy in Mu¨l-heim by using CT for guidance technique. The combinationof CT and X-ray fluoroscopy made CT intervention feasiblefor new catheter techniques. Combination of endoscopy toboth these methods offers improvement of guidance tech-nique with the possibility for surgical tomography. Threegroups started interventional MRI nearly at the same time.Ferenc Jolecz [12] in Boston and Robert Lufkin [18] in LosAngeles were the first who performed procedures in a tubedesigned magnet. Since 1988 we have used an open MRI forinterventions. Contrary to diagnostic Radiological Depart-ments therapeutic procedures became the main target in theroutine work of the Institute. Different new treatments havebeen developed and clinically evaluated in prospective stud-ies. On the other hand it was necessary to develop tools,instruments, hardware and software components, which

were compatible for use inside the magnet. Besides the newinterventional possibilities this is the major challenge indevelopment of Interventional MRI.

Biopsies

Percutaneous biopsies were the first interventions per-formed in a CT scanner [10]. The accuracy of tip definitionwith modern CT machines is 1mm3. The possibility to docu-mentate the needle tip inside a lesion is important for opti-mal results. Fine needle biopsies are then easy and accurate.Since the seventies CT guidance became a standard methodfor diagnosis of pulmonary lesions [21]. The accuracy of themethod is with more than 90% histological results veryhigh. The introduction of ultrasound for guidance techniqueof biopsies widened the range of this method for lesions inthe abdomen [20]. The definition of the needle tip is moredifficult than in CT and MRI. At this moment CT is the goldstandard for biopsies. In a prospective study with 1000 bi-opsies in the abdomen the sensitivity was 91.8% and thespecificity 98.9% [38]. In a study with 644 patients withbiopsies of pulmonary lesions we compared 338 patientswith CT guided biopsies with 308 patients with fluoroscopicguidance [30]. With a special fine needle (Cook) definitivehistological results could be obtained after a single biopsy in92,1%. There were no significant differences between bothtechniques. (Table 1).

In tumors with contact to the pleura complications werenot seen. The complication rate increased in relation to thepuncture site. Most complications occurred in the middlelobe (23.9%). Posterior access was mostly used in 483 casesand showed a complication rate of 5.6%. Lateral puncturehad 11.5% in 52 cases and anterior punctures had 20.2%complications out of 109 patients.

The combination of CT and fluoroscopy have been usedin our Institute since 1987. This technique allows increasedmonitoring of the puncture in pulmonary lesions especiallyduring breathing. This accelerates the procedure. On the

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 154–162

other hand CT gives an accurate documentation for thepuncture of lesions and its relation to major vessels. Punc-ture of the lung can be avoided if a lesion is located in themediastinum. The pleura can be distended by injection ofsaline solution under CT monitoring. Step by step the needlecan be advanced to the target (Fig. 1A,B). This technique isused to improve feasibility of biopsies and other interven-tions in the thorax like thoracic sympathectomy ten years inour department.

MRI guidance is increasingly used for abdominal andbone marrow biopsies if tissue characterization is not pos-sible in US and CT. Like CT MRI shows no artifacts be-cause of air or bony structures. In contrast to US full tissuetransparency is possible. An advantage of MRI is the nearlyreal time tip tracking and the lack of radiation. Speciallyopen MRI systems allow a wide access to the patient and asafe and comfortable positioning of patient and physician.At this moment recurrence of tumors in the pelvis and tu-mors in scar tissue as well as tumors in the brain and spineseem to be a good indication for MRI guided biopsies.

Local tumor therapy and tumor pain therapy

CT guided local tumor therapy have been developed aftersuccessful results of biopsy and CT guided pain therapy.Both methods increase the quality of tumor therapy by re-ducing the time for definitive histological diagnosis anddecreasing the tumor pain. For local tumor therapy differentmethods can be used. CT guided ablation techniques withethanol or thermal ablation with radiofrequency and laserhave been developed. For these ablation techniques inter-ventional MRI seems to be the future tool because of moni-toring of tissue changes during therapy. On the other handit is possible to instillate local acting cytostatics to decreasesystemic side effects and complications [7].

CT guided treatment of the brachial plexus and theaxillary plexus

Tumor infiltration of the plexus axillaris and brachialis is asevere problem in breast and lung cancer. Those patientshave intractable pain. The anatomy of the plexus brachialisand axillaris is very complex. Especially in case of patho-logical anatomy and tumor infiltration the risk of injuries ofthe arteria carotis and arteria vertebralis or stellate ganglia is

increased. After accidental injection of local anesthetics intothe arteria vertebralis without CT guidance complicationslike stroke with hypoxemia and convulsion are described[2]. CT guidance has the potential to reduce complicationsin this region. The arteria vertebralis can be detected clearly.Breast and lung cancer patients with tumor infiltration of thebrachial plexus can be treated from a ventro lateral access.For neurolysis 50–70% ethanol is used.

A study with 38 patients, 22 female and 16 male, with atotal of 189 treatments have been performed. Indication fortreatment was incurable pain caused by plexus infiltration.Combination of stellate ganglion block with long actinglocal anesthetics and local infiltration with ethanol showedthe best results with more than 85% good and moderate painrelieve. After the treatment weakness of arm and fingers orHorner’s syndrome could be reduced [6].

Table 1. Complications after fine needle biopsies with CT or fluoroscopicguidance

CT guidancen 4 338

Fluoroscopyn 4 306

n % n %

Pneumothorax 28 8.2 20 6.5Therapy needed 0 0 4 1.3Hemoptysis 3 0.9 6 2.0Shoulder pain 5 1.5 1 0.7

Fig. 1. A Thoracic sympathectomy under ‘‘guidance.’’ Saline solution canpush away structures like vessels and the lung. With a direct puncture theneedle would cross the pleura two times.B CT guided thoracical sympa-thectomy. After injection of saline solution the pleura can be distended.The introducer can pass through it without risk to hit the lung. Even thevessels in the mediastinum can be pushed away by this ‘‘safe accesstechnique.’’

155

CT guided neurolysis of celiac plexus

Tumor infiltration of the celiac plexus causes extraordinarypain. All efferent visceral fiber of the organs in the upperabdomen cross the celiac plexus. Because of this reasonpain in the upper abdomen can be treated by neurolysis ofthe celiac plexus. Most patients with pain in this region havea pancreatic carcinoma or gastric cancer, but also pain byrapidly growing liver metastases (capsule pain) can be

treated very well. Permanent pain relief can be achieved bya single treatment in 80% of the patients.

188 patients were treated by CT guided celiac plexusneurolysis [25]. Mean age of patients was 67 years. After asufficient test block with local anesthetics the neurolysiswas performed with 50% ethanol in the same session. 96%ethanol is normally very painful to the patient and carries anincreased risk of complications [3]. Before the ethanol in-jection contrast media is given to approve the distribution.CT guidance now is standard for neurolysis of the celiacplexus because the pathology can be diagnosed exactly andthe safest access to the plexus can be chosen. We prefer theanterior approach (Fig. 2). In relation to the pathology of thetumor the puncture can be done directly or through liver orstomach.

No major complications occurred after the neurolysis inour patients. Pain during the intervention can be avoided byinstillation of 10 to 20 cc local anesthesia prior to the neu-rolysis. Other side effects are mild diarrhea in up to 60% fortwo days.

Recurrence of pain after a period of complete pain reliefindicates a recurrence of tumor disease or regeneration ofplexus fibers. In those cases the neurolysis should be re-peated.

Lumbar and thoracic sympathectomy

Pain because of tumor infiltration of the sympathetic trunkor spinal metastasis can be treated with lumbar and thoracicsympathectomy. Similar to sympathectomy for OAD theaccess to the lumbar and thoracic sympathetic trunk shouldbe latero dorsal. The position of the patient is stabile and therisk to puncture arteries, bowel or lung is low. The sympa-thectomy should be done at the tumor level. CT guidanceshows the pathological anatomy and allows the monitoringof the ethanol distribution (Fig. 3).

Presacral and precoccygeal sympathectomy

Colorectal and uterus cancer cause very often infiltrations ofthe presacral plexus. This leads to typical pain in the pelvisand leg. In those patients it is possible to treat the pain withlocal injections of ethanol into the plexus under CT guid-ance. Even patients with extraordinary pain could be treatedand had a complete pain relief. In a study we treated 25patients with an infiltration of the plexus. These patients had112 treatments, in the mean 4,5 per patient. The treatmentwas repeated every week with 5 to 8 cc ethanol. The tumorswere either primaries or recurrences of rectal or uterus can-cer. The success rate of the method is nearly 75% [27]. Thepuncture was performed through a lateral access, if the tu-mor could be reached through the foramen piriforme. It isonly possible to reach the upper part of the plexus by ananterior approach.

Patients with combined pelvic and leg pain because oftumor infiltration in the pelvis could be treated with a com-bination of presacral and lumbar sympathectomy success-

Fig. 2. CT guided neurolysis of the celiac plexus. The anterior approachallow distribution of ethanol to both sides of the aorta into the celiacplexus. Only a single puncture technique is necessary for complete neu-rolysis of the plexus.

Fig. 3. CT guided lumbar sympathectomy for treatment of a chondrosar-coma. 21 year old patient with metastases in the vertebral muscles withinfiltration of the spine. The sympathectomy was done in the contralateralside because of the symptoms in the lower limb. After injection of 3 ccethanol the patient had a complete pain relief.

156

fully. Since the introduction of this combination it was notnecessary to perform a chordotomy in any of our patients.

Intratumoral and peritumoral therapy

After local injection of ethanol into tumors for pain therapywe could archive reduction of tumor size. Different groupshad the same experience [15, 34]. Because of this reason westarted a study for local tumor treatment. Aim of the studywas to develop an effective local treatment of tumor pain,treatment of lesions of nerves because of tumor infiltration,stabilization of spinal metastasis and treatment of risk ofparaplegia because of spinal cord compression by tumor.

Prerequisites for the study were tumor stage M1 andunsuccessful radiotherapy or chemotherapy. Treatment wasdone with ethanol or a combination of ethanol and mitox-antron, a low toxic anthracyclin, which is tolerated locally.110 patients were treated. The aim of this study was thepalliation of the patient. So we treated large tumors only inthat part which was responsible for the symptoms.

No major complication were seen. The monitoring ofthe spread of the ethanol during treatment is the reason ofthis result. The use of mitoxantron showed slight side ef-fects with myelotoxicity and minimally reduced thrombo-cytes and granulocytes.

Results of local tumor therapy

110 were treated with local tumor therapy up to ten timeswith gaps of 2 to 4 weeks. The youngest patient was 29, theoldest 81, average 64 years. In 25% of the patients a partialremission with tumor reduction of 50% was found (Table3). 68 patients had no change of tumor size and 13% had aprogression.

The palliation with pain relief, stabilization of bone de-structions and decompression of the spinal cord could bereached in 80% of the patients. 24 cases were treated by anadditional CT guided sympathectomy.

Minimally invasive therapy of the spine

Back pain and disk diseases are very common. In Germanywe spent about 18 billion $ in 1989 for treatment of spinaldiseases. On the other hand loss of Gross National Productis about four times higher. Microtherapy is an alternative toconventional open surgery of disk herniations.

Percutaneous techniques for treatment of spinal diseaseslike disk herniations had been limited to contained disks[23]. 1990 we developed a new technique for micro endo-scopic interventions in the spine. Sequestrectomy for smallfree fragments and dissection of spinal scar tissue became

possible. CT and MRI allow imaging of the structures in thespinal canal.

The accuracy of guidance of instruments in the body is1mm3 for CT. Contrary to X ray fluoroscopy CT and MRIprovide accurate localization of nerves, vessels and hernia-tions in the spine. The possibility of monitoring of instru-ments and effects during the procedure demonstrate the po-tential of these methods.

For diagnosis CT and MRI are important and shouldperformed after conservative treatment of two weeks with-out success or immediately if neurological symptoms arepresent.

Different methods for treatment of chronic diseases ofthe spine are possible and can be performed under localanesthesia:

1. facet joint denervation;2. periradicular therapy (PRT);3. percutaneous nucleotomy (PNT);4. micro endoscopic sequestrectomy;5. micro endoscopic scar dissection.

Facet joint denervation

Low back pain without sciatica is often related to degen-erative diseases of the facet joints. About 80% of facetsyndromes are located in L4/5 and L5/S1. In 1971 Rees [24]introduced facet joint denervation as a surgical procedure. Afew years later electrocoagulation of the facet joints wasintroduced [33]. Facet joint block with local anesthetics orfacet joint denervation with 96% ethanol is performed underCT and MRI guidance in our institute. The neurolysisshould be done in both joints of the level. If there is anadditional compression of the nerve root the treatment canbe combined with PRT. Good results of the treatment can beachieved in 65 to 75% of the patients. CT allows monitoringthe distribution of ethanol, so that accidental injections tothe nerve root or into vessels are nearly impossible [17, 29].

Periradicular therapy

In patients with failed back surgery syndromes we use since1986 CT guided and since 1988 also MRI guided (Fig. 4)periradicular therapy (PRT). Under CT guidance a 22 or 23Gauge coaxial needle system can be advanced step by stepto the nerve root. The puncture should be made in the lowerthird of the foramen and undercross the nerve root. Localanesthetics and local acting cortisone were applied after testinjection of contrast (Fig. 5). The treatment should be re-peated after three to four weeks four to six times. After the

Table 2. Tumor localization in celiac plexus neurolysis

Pancreatic carcinoma 82Gastric carcinoma 54Metastases with direct infiltration of the plexus 43Liver metastases 9

Table 3. Results of CT guided local tumor treatment (N4 110)

Pain relief 80%Tumor size reduction <50% 25%Tumor size no change 62%Tumor progression 13%

157

first good results of this method we treated disk herniationas well as spinal stenoses (Table 4). PRT restitutes the func-tion of the nerve root. If patients denied surgery even weak-ness of the leg could be treated successfully. We performeddifferent studies on PRT [29]. The results are listed in Ta-ble 5.

A prospective randomized study with 40 patients (21male and 19 female) was performed in 1990. The studycompared PRT with 10mg triamcinolonacetonide versus40mg per treatment. Two independent examiners analyzedthe CT images before and after the study and evaluated theresults. All patients suffered from chronic pain by disk her-

niation with radicular symptoms in the treated level. Thepatients had to answer a questionnaire including the visualanalogous scale (VAS) before and after treatment. 0 valuemeans no pain, 100 is maximum pain. Additional neuroorthopedic examinations were made by independent physi-cians. Every 3 to 4 weeks 4 to 8 treatments were performed.Statistical evaluation was carried out with Mann Whitneytest for analysis of comparison of both groups before andafter treatment as double sided plausibility test [9]. Resultsare shown in Table 6.

Statistically the range of pain according the VAS wasnot different in both groups. After therapy in V40 group ahigh significant improvement of symptoms could be ar-chived. In the V10 group we found only slight improve-ment. The neuro orthopedic examination showed a com-plete improvement of the stretched leg test from initial 30 to60 degrees in most cases.

After PRT complication and side effects are seldom.One patient with a diastematomyelia had immediately afterPRT a plegia of the leg. The weakness disappeared afterconservative treatment of a few weeks. Therefore we rec-ommend to exclude patients with spinal cord abnormalitiesfor this treatment. Two patients had a spondylodiscitis anda spondylitis after PRT. No bacterial infection was found atbiopsy. These patients were treated conservative in the hos-pital. A list of side effects is given in Table 7.

Percutaneous nucleotomy

Percutaneous techniques for nucleotomy were first de-scribed by Hijikata in 1975 [11]. Friedman [5], Kambin [14]and Suezawa [36] have further developed this technique.Suezawa added diskoscopy to fluoroscopy for monitoring ofthe treatment. For these procedures general anesthesia wasneeded. Two trocar systems were advanced to both foramenintervertebrale with a diameter of 11 mm. Onik introduceda new automated nucleotomy device with a probe of 2 mmin diameter [23]. This procedure has been performed underlocal anesthesia with fluoroscopy since 1985.

Since 1988 we have performed percutaneous nucle-otomy with combined CT and fluoroscopy guidance. Firstwe started with the automated probe of Onik (PNT). This

Table 4. Indications of periradicular therapy (PRT)

Disk herniationFailed back surgerySpinal stenosisStenosis of recessus lateralisStenosis of foramen vertebralisWeakness and loss of sensibility

Table 5. Results of PRT; first group (N4 370)

Symptoms improved after first therapy 15%Symptoms improved after fourth therapy 72%No symptoms at the end of treatment 78%No symptoms three months after treatment 83%No symptoms two years after treatment 68%Disk surgery after PRT 3%

Fig. 4. MR guided periradicular therapy (PRT) for treatment of lumbardisk herniation. For treatment planning a well known ‘‘digital’’ device canbe used to find the right entry point on the skin. The finger tip is placedunder MR fluoroscopy.

Fig. 5. MR guided PRT. For highlighting we use Gadolinium solution toshow the right distribution of long acting corticoides in the epidural space.

158

cutting device is able to remove material of the nucleuspulposus. Later we switched to Laser nucleotomy (PLNT).We use Holmium and Neodymium YAG Laser. Free frag-ments in the spinal canal or mass herniation can not betreated by percutaneous nucleotomy. Small herniation andcontained disk, this means the herniation is within the disklevel without free fragments, with compression of the nerveroots and correlated neurological symptoms are a good in-dication for percutaneous therapy.

We have treated 110 patients with PNT [31] and 116with PLNT [8]. There were no significant difference inoutcome of both groups, but it is easier to treat with Laser.We use 20 to 18 gauge introducer sets and are testing now23G for Laser. So the diameter is only 1.2 to 0.65 mm. Inthe study we used Laser fibers with 0.4 mm diameter. Therisk of nerve root injury in the foramen is very low. For theautomated suction device we used introducers with diam-eters from 2.8 to 4.0 mm.

For a safer procedure of PNT we modified the set andstarted the puncture with a 22G fine needle with pull offconnector, that could be removed after safe puncture of thedisk under CT monitoring. Then the puncture canal is di-latated step by step until the next size of the introducercould be inserted to the annulus fibrosus under fluoroscopy(PNT).

Patients

PNT

110 patients were treated from 1988 to 1991. The mean agewas 45 years with a range from 25 to 66. Two patientsunderwent PNT at the level L3/4 (2%), fifty eight at levelL4/5 (53%), and fifty at level L5/S1 (45%).

PLNT

In 116 patients a Laser nucleotomy was performed between1989 to 1991. The mean age was 47 years (range 15 to 81years).

The procedures were performed at the level L2/3 inthree patients (2%), at level L3/4 in nine patients (8%), atlevel L4/5 in 71 patients (62%) and at level L5/S1 in 33patients (28%).

Follow up

Time schedule for The follow up examinations was 3weeks, 6 weeks, 3 month, 6 month and 1 year after PNT and

PLNT. The success of both methods was proven by clinicaland neurological examinations at all follow ups and by CT3 weeks and 6 month after therapy. A MRI study of thelumbar spine was routinely done 6 weeks after treatment toexclude discitis. Visual Analogous Scale (VAS) was usedfor evaluation of pain.

Results

In 82% of the patients (n4 110) a complete pain relief afterPNT. Neurological findings like pathological stretched legtest or weak muscle reflexes are improved immediately afterPNT in 92%.

Complications and side effects

One patient after PNT and one patient after PLNT sufferedfrom a spondylodiscitis. 26% of the patients had an irritationof the segmental nerve root for three to five weeks afterPNT and 21% in the PLNT group.

New treatments for uncontained disk and scar tissue

The nucleotomy methods described above were only indi-cated in contained disk. Therefore free fragments in thespinal canal and the foramen could not be treated. We de-veloped a new technique with a combination of CT, fluo-roscopy and micro endoscopy to visualize very small struc-tures inside the foramen and the spinal canal percutane-ously. With new endoscopes this technique now can also beused inside an open magnet. We started the new techniquefor dissection of scar tissue in the spine (Fig. 6) [28] and fortreatment of small free fragments [26] in 1990 in our insti-tute.

A study of 20 patients with chronic low back pain wasinitially published. Four of them had free fragments in thespinal canal. In 8 cases lateral herniation with free frag-ments in the foramen were treated. 7 patients had prior opensurgery with scar tissue in the spinal canal and were alsotreated under endoscopic combined CT fluoroscopic andmicro endoscopic monitoring. In one patient an abscess inthe foramen was removed. Overall mean age was 51.9 ±14.8 years Level L4/L5 was treated in 12 patients and L5/S1in 8 patients. 17 patients were outpatients. Immediately aftertherapy, pain relief and improvement of neurological symp-toms were assessed. Furthermore within the context of a

Table 6.Results of a randomized prospective double blind study (N4 40)on PRT; improvement after treatment 40 mg triamcinolocacetonide versus10 mga

Group N Mean SDA Min Q1 Media Q3 Max

All 40 67.5 34 0 42.5 80 97.5 100V40 30 72.3 34.71 0 57.6 90 100 100V10 10 54 29.14 0 37.5 55 75 100

a V10, triamcinolon 10 mg group; V40, triamcinolon 40 mg group; SDA,standard deviation; Q1, lower quartile; Q3, upper quartile.

Table 7. Side effects of PRT; results of two studies

Study (n4 40) Study (n4 370)

n % n %

Indigestibility 1 2.5 0 —Edema 1 2.5 11 3Acne 1 2.5 15 4Tensions and blushing 1 2.5 7 2Hair loss 2 5 0 —Hormonal disorder 2 5 5 (of 170 female) 3Weight increase 4 10 30 8Stomach disorder 5 12.5 0 —Calf cramps 15 37.5 30 8Petechia 0 2 0.5

159

follow up the patients were asked to evaluate treatmentsuccess, defined as improvement of pain symptoms esti-mated on VAS.

Instrumentation

Special introducer sets with diameter from 1.2 to 4 mm hadto be developed for the procedure. The smallest fiberscopewith a diameter of 0.29 mm fits into a 22G coaxial inter-ventional needle set. A good view is obtained with endo-scopes of a diameter more than 0.6 mm. The best view isgiven by endoscopes with flexible tips and a diameter of 1.4mm. We developed micro forceps, scissors and other instru-ments with 0.9 mm diameter, which can be used in theforamen and spinal canal. Different laser systems were usedfor ablation: a NdYAG and HoYAG laser. An interventionalCT (Somatom plus 4, Siemens) is used for tomography.This CT allows two or three surgeons to perform the pro-cedure. For fluoroscopy we use a digital C arm.

Results

Before treatment 20 patients had a pain score between 70 to100 (VAS). 80% of the patients (n4 16) had neurologic

deficits before micro operation: weakening of the patellarreflex, weakness of the tibialis anterior as well as weaknessof the extensor brevis and decreased ankle jerk. Immedi-ately after the micro surgery improvement of reflexes orweakness could be achieved in 75% of patients. In 94% ofthe cases stretched leg test decreased more than 50° or wasnegative. 50% improvement of pain symptoms was presentin all patients at the end of the procedure. The average painscore decreased from 88 to 8%. Pain relief more than 50(VAS) at follow up after 6 month was obtained in 90% ofthe patients (n4 18).

Complications and side effects

Complications of CT-guided spinal endoscopic procedureswere not seen in a few cases. Five of our patients com-plained slightly of nerve root irritation at the treated level.This lasted several days, sometimes up to two weeks. Con-servative treatment with PRT was necessary in 4 patients.These symptoms were related to mechanical irritation bypressure to the nerve root during PLNT. Three of these 5patients suffered from pseudoradicular pain. This wastreated with blockade of both facet joints (3 times in 8 days)by injecting a local anesthetics.

Discussion

Back pain and disk diseases are very common. In Germanywe spent about 18 billion dollars in 1989 for treatment ofspinal diseases. On the other hand loss of Gross NationalProduct is about four times higher. Open surgery of thespine is limited because of poor results and post nucleotomysyndrome with severe disability. These reasons are chal-lenges to develop new less invasive techniques for treatmentof spinal diseases.

In therapy of the vertebral column and intervertebralstructures by percutaneous procedures offer safe alterna-tives to conventional open surgery of the disk. PRT, PLNTand PNT and micro endoscopic procedures are successfulbecause of decompression and restoration of the function ofthe nerve root. Even if a complete replacement of the opensurgery is not possible at this moment, these micro invasivetechniques have a wide field for application. Especially inpost nucleotomy syndromes with chronicle pain micro-therapy can offer a new way of treatment [32].

The limitation of percutaneous nucleotomy techniqueswere free fragments in the spinal canal and in the foramen[22]. CT or MRI guided micro endoscopy of the spinal canalenables the possibility for treatment of dislocated hernia-tions. First preoperative planning of the best access and theright angle can be done in 3D technique. Then CT and MRIgive nearly real time imaging for monitoring and instrumentguidance. In addition micro endoscopy allows the visual-ization of the surface of the operative field. The combina-tion of these techniques facilitates percantaneus micro sur-gical procedures in the spine.

In Oncology CT or MRI guided Interventions have twomain fields: pain therapy and local tumor therapy. Even inpathologic anatomy CT and MRI allows precise access to

Fig. 6. CT, fluoroscopic and micro endoscopic guided scar dissection inthe spinal canal. The procedure opens a new field in treatment of failedback surgery. Instruments can be monitored during the percutaneous pro-cedure by CT and micro endoscopy.

160

nerves and tumors. The control of the puncture and moni-toring of treatment modalities, like local cytostatic drugs,ethanol injection, heat application with laser, and RF abla-tion, or cryosurgery demonstrate the advantages of thesetechniques. Accidental injection of local anesthetics into thevertebral artery can cause a convulsion immediately [2].Even negative aspiration test gives not enough safety toavoid injuries of arterial wall without CT guidance [19].Because of these reasons we recommend strongly the use ofCT and MRI for local application of toxic materials in tu-mors.

Primary tumors and mestastasis in the liver are success-fully treated [16, 37]. Application of Laser or RF ablationfor metastases of liver and brain seem to become alterna-tives for open surgery as well [1, 3, 4, 13] for therapy ofhead and neck tumors.

An advantage of combination of CT and fluoroscopy isthat patients can be treated or examined without movementor transportation from one machine to the other. This fa-cilitates procedures for diagnosis and therapy in one exami-nation room. With Siemens company we developed 1990 anew CT system for interventional procedures, the SomatomPlus ICT. Special software and hardware components madethis machine suitable for new and fast Minimally InvasiveTherapies like CT monitored liver perfusion in hepatocel-lular carcinoma where it is important to check the cathetertip position and the area of perfusion during therapy. An-other common development was the new MRI system withan open magnet, Magnetom Open, which is feasible to useinside the OR. This machine is patient friendly with a freeaccess from three sides to the patient and has special fea-tures like MRI ‘‘Fluoroscopy’’ with in room foot switch andin room monitor with LCD flat screen and the possibility touse the computer of the system inside the cabin so that everysequence can be started in the MR room. In addition this

machine is very silent, because of low gradient noises. Atleast today we installed our latest common development forSurgical Tomography, the prototype of a real time CT (So-matom Plus 4 Real Time, Siemens) with the extraordinarycomputer capacity that speeds up the machine to have theimage reconstructed at the end of a 0.75 sec scan. An ad-ditional fluoro mode enables this machine to reconstructimmediately 6 images per second. First tests show a verygood ability for Image Guided procedures (Fig. 7). Addi-tional use of endoscopy is the key to transmit every kind oftreatment modality safely into the body.

The use of micro scalpels, micro instruments, radiofre-quency, laser, argon plasma coagulation, cryosurgery, hy-perenergetic ultrasound or micro motor driven remote con-trol micro robots inside the body will be monitored by thiscombination. Future operative room will include modularsystems of imaging machines for MIS.

Finally, it is important to acknowledge the need of bettereducation and training in these new methods. Minimallyinvasive therapy has raised the need for training and edu-cation. Major European Centers have established trainingcourses for endoscopic surgery (Buess, Cushieri, MacMa-hon, Marescaux) and for interventional technology (Mer-land). Currently we are working on training courses on in-terventional MRI and CT and the first courses will takeplace in May 1997.

References

1. Anzai Y, Lufkin R, Castro D, et al. (1991) MR imaging guided inter-stitial NdYAG laser phototherapy: dosimetry study of acute damage inan in vivo model. J Magn Reson Imaging 1: 553–559

2. Ariani J, Parmley J, Ochsner A (1952) Facilities and complicationsafter attempts at stellate ganglion. Surgery 32: 615

3. Cherry D, Lamberty J (1984) Paraplegia following celiac plexus block.Anaesth Intensive Care 12: 59

4. Fan M, Ascher PW, Schrottner O, Ebner F, Germann RH, Kleinert R(1992) Interstitial 1.06 Nd:YAG laser thermotherapy in malignantgliomas under real time monitoring of MRI: experimental studies andphase I clinical trial. J Clin Laser Med Surg 10: 355–361

5. Friedman WA (1983) Percutaneous discectomy: an alternative to che-monucleolysis? Neurosurgery 13: 542–547

6. Gronemeyer DHW, Seibel RMM (1990) CT guided cervical blockadesand methods of cancer therapy for lesions of the cervicobrachialplexus. In: Seibel RMM, Gro¨nemeyer DHW (eds) Interventional com-puted tomography. Blackwell, Oxford, pp 161–179

7. Gronemeyer DHW, Seibel RMM (1993) Mikroinvasive CT gesteuerteTumortherapie von Weichteil- und Skelettmetastasen. Wien MedWochenschr 143: 312–321

8. Gronemeyer DHW, Seibel RMM, Schmidt AM, Kremer G, van Leeu-wen P (1993) A traumatic CT-controlled percutaneous laser nucle-otomy. Min Inv Ther 2: 247–255

9. Gronemeyer DHW, Seibel RMM, Schindler O, Schattauer K, Lange S,Schmidt A (1995) Die mikroinvasive CT gesteuerte periradikula¨reTherapie zur Behandlung von chronischen bandscheibenbedingtenFunktionssto¨rungen. Wien Med Wochenschr 145: 129–139

10. Haaga JR, Alfidi RJ, Havrilla TR, et al. (1976) Precise biopsy local-ization by computed tomography. Radiology 118: 603

11. Hijikata S, Yamiagishi M, Nakayama T, Oomori K (1975) Percutane-ous diskectomy: a new treatment method for lumbar disc herniation. JToden Hosp 5: 5–13

12. Jolecz F, Bleier AR, Jakab P, Ruenzel PW, Huttl K, Jako GJ (1988)MR imaging of laser tissue interactions. Radiology 168: 249–253

13. Kahn TH, Bettag M, Ulrich F, et al. (1994) MRI guided laser inducedinterstitial thermotherapy of cerebral neoplasms. J Comput Assist To-mogr 519.

14. Kambin P, Gellman H (1983) Percutaneous lateral discectomy of thelumbar spine. A preliminary report. Clin Orthop 174: 127

Fig. 7. First intervention in the new Interventional Real Time CT. After0.75 sec the first image appears on the screen. A new image can be madeevery 0.75 sec or in a fluoro mode six images per second can be recon-structed.

161

15. Livraghi T, Festi D, Monti F, et al. (1986) US guided percutaneousalcohol injection of small hepatic and abdominal tumors. Radiology161: 309–312

16. Livraghi T, Vettori C, Lazzaroni S (1991) Liver metastases: results ofpercutaneous ethanol injection in 14 patients. Radiology 179: 709–712

17. Lora J, Long MD (1976) So called facet denervation in the manage-ment of intractable back pain. Spine 1: 121–126

18. Lufkin R, Teresi L, Chiu L, Hanafee W (1988) A technique for MRguided needle placement. AJR Am J Roentgenol 151: 193–196

19. Mastroianni A (1986) The effects of stellate ganglion block. Schmer-zdiagnostik und Therapie, vol 2. Bochum, pp 80–93

20. Matalon TAS, Silver B (1990) US guidance of interventional proce-dures. Radiology 174: 43–47

21. Nordenstro¨m B, Sinner WN (1979) Needle biopsy of pulmonary le-sions. Fortschr Ro¨ntgenstr 129: 414

22. Onik G (1988) Percutaneous automated discectomy: technique. In:Onik G, Helms CA (eds) Automated percutaneous lumbar discectomy.Rad Res Educ Found Univ Cal Print Dep pp 77–110

23. Onik G, Helms CA, Ginsberg I, Hoaglund FT, Morris J (1985) Per-cutaneous lumbar discetomy using a new aspiration probe: porcine andcadaver model. Radiology 155: 251

24. Rees WS (1971) Multiple bilateral subcutaneous rhizolysis of segmen-tal nerves in the treatment of the intervertebral disc syndrome. AnnGen Pract 26: 126–127

25. Schild HH, Seibel RMM, Gro¨nemeyer DHW (1990) Percutaneousneurolysis of the celiac plexus. In: Seibel RMM, Gro¨nemeyer DHW(eds) Interventional computed tomography. Blackwell, Oxford, pp181–191

26. Seibel RMM (1996) CT guided endoscopy of sequestered disk her-niation. J Vasc Interv Radiol 1(Suppl): 222

27. Seibel RMM, Gro¨nemeyer DHW (1990) CT guided neurolysis of thepresacral and precoccygeal sympathetic trunk. In: Seibel RMM,

Gronemeyer DHW (eds) Interventional computed tomography. Black-well, Oxford, pp 193–199

28. Seibel RMM, Gro¨nemeyer DHW (1994) Technique for CT guidedmicroendoscopic dissection of the spine. End Surg 2: 226–230

29. Seibel RMM, Gro¨nemeyer DHW, Grumme TH (1990) New treatmentsof the spinal column diseases using interventional radiological tech-niques. In: Seibel RMM, Gro¨nemeyer DHW (eds) Interventional com-puted tomography. Blackwell, Oxford, pp 95–97, 100–115

30. Seibel RMM, Gro¨nemeyer DHW, Weigand H (1990) Thorax biopsy.In: Seibel RMM, Gronemeyer DHW (eds) Interventional computedtomography. Blackwell, Oxford, pp 15–30

31. Seibel RMM, Gro¨nemeyer DHW, So¨rensen RAL (1992) Percutaneousnucleotomy with CT and fluoroscopic guidance. J Vasc Interv Radiol3: 571–576

32. Seibel RMM, Gro¨nemeyer DHW, Schmidt AM, Melzer A (1996)Minimally invasive treatment of spinal diseases. Min Inv Ther AllTech 5: 226–231

33. Shealy CN (1975) Percutaneous radiofrequency denervation of spinalfacets: treatment of chronic back pain and sciatica. J Neurosurg 43:448–451

34. Sheu JC, Huang GT, Chen DS et al. (1987) Small hepatocellularcarcinoma: intratumor ethanol treatment using new needle and guid-ance systems. Radiology 163: 43–48

35. Silverman ST (1996) Percutaneous abdominal biopsy: recent advancesand future directions. Semin Int Radiol 13: 3–15

36. Suezawa Y (1982) Kreuzschmerzen in der Orthopa¨die—HeutigerStand ihrer Therapie. Schweiz Rundsch Med Prax 71: 783

37. Vogl Th, Mack MG, Scholz WR, Mu¨ller P, Weinhold N, Philipp C,Bottcher H, Roggan A, Felix R (1996) MR imaging guided laserinduced thermotherapy. Min Inv Ther All Tech 5: 243–248

38. Welch TJ, Sheedy PF, Johnson CD et al. (1989) CT guided biopsy:prospective analysis of 1000 procedures. Radiology 171: 493–496

162

A comparison of surgeons’ posture during laparoscopic and opensurgical procedures

R. Berguer,1 G. T. Rab,2 H. Abu-Ghaida,2 A. Alarcon,1 J. Chung1

1 Department of Surgery University of California Davis, 4301 X Street, Sacramento, CA 95817, USA2 Department of Orthopedic Surgery, University of California Davis, 4301 X Street, Sacramento, CA 95817, USA


AbstractBackground:There is increasing recognition of surgeons’physical fatigue in the new ergonomic environment of lap-aroscopic surgery. The purpose of this study was to deter-mine what the differences are in the movement of the sur-geon’s axial skeleton between laparoscopic and open opera-tions.Methods:Surgeons’ body positions were recorded on vid-eotape during four laparoscopic (LAP) and six open (OP)operations. The percent of time the head and back were ina normal, bent, or twisted position as well as the number ofchanges in head and back position were tabulated using acomputer program. A separate laboratory study was per-formed on four surgeons ‘‘walking’’ a 0.5-inch polyethyl-ene tubing forward and backward using laparoscopic andopen techniques. The movements of the surgeons’ head,trunk, and pelvis were measured using a three-camera ki-nematic system (Kin). The center of pressure was recordedusing a floor-mounted forceplate (Fp).Results:In the operating room surgeons’ head and backpositions were more often straight in laparoscopic proce-dures and more often bent in open operations. The numberof changes in back position per minute were significantlydecreased when the laparoscopic-only part of surgery wasanalyzed. In the laboratory the subjects’ head position wassignificantly (p 4 0.02) more upright and the anteroposte-rior (AP) and rotational range of motion of the head wassignificantly reduced during laparoscopy. Subjects’ CP wasmore anterior and there was a significant reduction in theAP range of motion of the CP during laparoscopy.Conclusions:Our study suggests that surgeons exhibit de-creased mobility of the head and back and less anteropos-terior weight shifting during laparoscopic manipulations de-spite a more upright posture. This more restricted postureduring laparoscopic surgery may induce fatigue by limitingthe natural changes in body posture that occur during opensurgery.

Key words: Laparoscopy — Ergonomics — Posture —Fatigue

Surgery can be fatiguing work. Kant et al. [3] reported thatsurgeons and instrument nurses exhibited frequent staticbody postures that were ‘‘distinctly harmful’’ and contrib-uted to physical fatigue during surgery. The rapid growth ofendoscopic surgery over the past 10 years has dramaticallychanged the physical environment of the surgical team,placing new demands and restrictions on the range of move-ment of the surgeon [7]. The effects of these factors onsurgeons’ posture have had limited study to date [8] butsuggest there is a further increase in static postural stress onthe surgeon during laparoscopic operations.

To further investigate the effects of laparoscopic surgi-cal techniques on surgeons’ posture and movement we re-port the results of a study of laparoscopic and open surgicaltechnique in both an operating room and a laboratory set-ting.

Methods

Laboratory setting and simulated surgical tasks

Three staff surgeons and one 4th-year surgical resident stood upright andperformed a two-handed simulated surgical task using either a standardhemostat and forceps (OP) (Aesculap, Germany) under direct vision or twocurved laparoscopic graspers (LAP) (Ethicon Endosurgery, Cincinnati,OH) and vision via a 0° laparoscope and CCD camera (Karl Storz Endos-copy, Germany), and 13-inch video monitor (Sony Corp., Japan). Lightwas provided by a laparoscopic light source (Karl Storz, Germany). Thetask consisted of ‘‘walking’’ a 0.5-inch polyethylene tubing forward andbackward (simulating inspection of the small bowel) for 60 s (see descrip-tion below for explanation of task duration) and was performed in a stan-dard rectangular laparoscopic training box. The position of the work sur-face was standardized for each subject with the instrument handles atelbow height and forearm distance when the instrument tips were locatedat the center of the simulated surgical field.Correspondence to:R. Berguer

Surg Endosc (1997) 11: 139–142

SurgicalEndoscopy


Laboratory kinematic measurements

Subjects’ body movements were recorded by three video cameras placed ina 60° arc 2 ft above flow level and 20 ft behind the subjects. The three-dimensional coordinates of the workspace were established prior to testingby recording a predetermined sequence of lights positioned along the work-space boundaries. Twelve reflective markers were placed on the posterioraspect of subjects’ head, shoulders, and pelvis and their visibility wasenhanced by the reflection from three 100-W incandescent bulbs aimed atthe subjects backside. All markers were clearly visible by the each of thethree cameras at rest. The subjects’ movements during the simulated sur-gical tasks were synchronously recorded on videotape by the three videocameras. The positions of the markers within the established workspaceboundaries were digitally sampled at 10 Hz from the videotape from eachcamera using EV3D Motion Analysis software (Motion Analysis, SantaRosa, CA) and their trajectories were calculated. (The software allowed amaximum of 600 data points; thus the duration of the task was limited to60 s at a sampling rate of 10 Hz). The head, trunk, and pelvis were eachdefined by three markers which described a corresponding plane in space.A script was written using the EV3D software that calculated the relativeangles of anteroposterior flexion (y axis), lateral bending (x axis), androtation (z axis) of the head to the trunk, the trunk to the pelvis, and thepelvis to the workspace floor. In addition to the relative angular movementsof the body segments, the absolute displacement of each marker projectedonto anteroposterior (y) and lateral (x) workspace axes was measured andthe average position, standard deviation, and range of motion of the threebody planes were calculated. The kinematic analysis therefore yielded bothrelative and absolute measurements of body part movements in each sub-ject during the simulated surgical tasks.

Forceplate measurements

The subjects’ center of pressure (analogous to the floor position of thesubjects’ center of gravity) was calculated from measurements recorded bya floor-mounted force plate (Advanced Mechanical Technology Inc., Wa-tertown, MA) located directly under their feet. The forceplate output wassampled at 5 Hz using Bedas-2 software (Advanced Mechanical Technol-ogy Inc., Watertown, MA) to obtain Force (F) and Moment of Force (M)measurements. (The software allowed a maximum of 600 data points; thusthe duration of the task was limited to 60 s at a sampling rate of 10 Hz.)From this data the average position, standard deviation, and range of mo-tion of the CP were derived in the x (lateral) and y (anteroposterior) planesfor each subject.

Operating room posture analysis

A series of four laparoscopic and six open operations were videotaped inthe operating room (OR) at the University of California Davis MedicalCenter in Sacramento, California, and the Veterans Administration ambu-latory surgery clinic in Martinez, California. The video provided a view ofthe surgeon from the knees to the head. The videotapes were reviewed bya single observer (R.B.) who recorded the surgeons’ head and back posi-

tions as straight, bent, twisted, or bent and twisted. This method is amodification and extension of the Ovako Work Analysis System (OWAS)developed by the Finnish Institute of Occupational Health [2]. A customcomputer program written in BASIC and run on an IBM-compatible com-puter recorded the surgeon’s head and back position when the observerpressed one of eight preassigned keys on the computer keyboard. Theprogram recorded the time each head and back position changed, and whenfinished, performed a second analysis by sampling the raw data output atintervals of 5 s and counted the frequency of the 16 possible combinationsof head and body positions during the time of operation. The use of real-time computer recording and analysis of postural data from videotape asused in the present study has previously been shown to be effective [4].

Results

Kinematic data (Table 1)

Subjects’ head position was significantly further forwardduring open manipulation. We observed significant reduc-tions in the AP and rotational range of motion of the headonly during laparoscopic tasks.

Forceplate data (Table 2)

Subjects’ CP was more anterior in three out of four subjectsduring open manipulation but this difference did not reachstatistical significance. Analysis of the lateral and AP move-ment of the CP demonstrated a significant reduction in theAP range of motion.

Video data (Table 3, Fig. 1)

Surgeons’ head and back positions were more often straightin laparoscopic procedures and more often bent in openoperations. We estimated how static surgeons’ head andback postures were by counting the number of changes inhead and back position per minute during all operations.Additionally, we focused on the ‘‘pure’’ portion of the lap-aroscopic procedures (i.e., during actual use of the laparo-scope and not including the opening and closing of theabdominal trocar sites) to better isolate the postural effectsof laparoscopic technique. We found that laparoscopic op-erations caused more frequent head movements than opensurgery overall, but not during the laparoscopic-only part ofthe operations. This data reflects the surgeons’ need to look

Table 1. Kinematic dataa

Kinematic data Head Trunk Pelvis

Lateral position (cm) Lap −3.6 ± 2.8 −2.1 ± 3.4 −3.3 ± 2.1Open −3.1 ± 0.5 1.7 ± 9.9 −4.1 ± 0.9

AP position (cm) Lap 5.6 ± 5.4* 5.8 ± 6.6 8.2 ± 6.9Open −9.7 ± 2.9 4.2 ± 6.0 7.4 ± 7.7

Lateral range (cm) Lap 6.1 ± 1.5 4.3 ± 0.9 3.0 ± 0.7Open 10.3 ± 7.8 9.8 ± 8.6 7.6 ± 6.9

AP range (cm) Lap 9.5 ± 9.2* 7.3 ± 3.5 11.8 ± 15.0Open 13.1 ± 9.4 9.3 ± 4.3 7.8 ± 6.9

Rotational range (degrees) Lap 12.1 ± 3.7* 13.8 ± 7.5 8.9 ± 4.8Open 31.6 ± 5.9 18.9 ± 16.3 15.5 ± 12.3

a Data listed as mean ± standard deviation, AP4 anteroposterior, LAP4 laparoscopic* p < 0.05 Lap vs Open

140

back and forth frequently between the abdominal wall andthe video monitor during the entry and closure periods ofthe laparoscopic operations. In contrast, the frequency ofback movements was the same between laparoscopic andopen operations overall but was significantly decreasedwhen the laparoscopic-only part of surgery was analyzed.These results confirm our observations that during laparo-scopic surgery surgeons tend to hold the trunk very stillwhile concentrating on the video monitor, whereas the trunkmoves more frequently during open operations.

Discussion

Poor posture can cause static muscle loading and fatigue [2]as well as impaired psychomotor task performance [1]. It istherefore not surprising that posture analysis has beenwidely used in industry to determine the strain placed onworkers by different work conditions [2]. Surprisingly, pos-ture analyses among medical workers have been limited.Lee and Chiou [5] reported that 17% of observed postures innursing personnel generated forces higher than the recom-mended action limit of the National Institute for Occupa-tional Safety and Health (NIOSH). Kant et al. [3] appliedposture study and the OWAS classification to OR personnelfor the first time and noted that surgeons and instrumentnurses exhibited repeated static postures characterized bythe head bent forward and the back twisted and bent 54%and 27% of the time, respectively, positions described as‘‘distinctly harmful’’ by the subjects. Recently, Radema-cher et al. applied the OWAS posture analysis system in astudy of minimally invasive procedures [8] and reportedthat during laparoscopic procedures 70% of intraoperativework postures were ‘‘substantially’’ static and caused by the‘‘disadvantageous arrangement of the equipment in the op-erating theatre’’ [8].

The goal of our study was to better understand the effectof laparoscopic surgical technique on surgeons’ axial skel-etal posture and movements. Our direct observations in theoperating room demonstrate that open operations are at-tended by more head-bent and back-bent posture—findingsthat are consistent with the previous report by Kant et al.[3]. In contrast, during laparoscopic procedures, we foundthat surgeons more often hold a head-straight and back-straight posture. It is possible that the more upright skeletal

posture during laparoscopic operations may cause muscularstrain to these areas. However, our analysis of the frequencyof changes in head and back position demonstrated fewerchanges in back position during laparoscopic surgery. Thelatter finding suggests increased static postures of the backwhich may in turn increase surgeon’s fatigue over time.

Our laboratory kinematic and forceplate analyses sup-port the results of the video analysis. These methods havebeen widely used by other investigators to provide precisemeasurements of body movement and shifts in the center ofpressure [6, 9, 10, 11] under various normal and diseaseconditions. In the present study, we found a more upright(centered) head position, less anteroposterior head move-ment, and less rotation of the head during laparoscopic tasksin the laboratory. Our forceplate results similarly demon-strated less anteroposterior movement of the center of pres-sure under the same conditions. Our results are similar torecent findings of increased static postures during laparo-scopic surgery reported by Rademacher et al. [8]. Thuswhile both our video and laboratory data demonstrate po-tentially ‘‘better’’ (more erect) head and back positions dur-ing laparoscopic operations, there appears to be an increasein the static nature (i.e., less movement) of head and backpositions during laparoscopic manipulations which couldincrease postural muscular fatigue in these body areas.

In comparing our laboratory and OR data we acknowl-edge that the limited task complexity and the short period ofobservation we used for the laboratory experiments werelikely to underrepresent the unconscious posture surgeonswould adopt in the operating room, and this effect may havebiased our results. In this light, however, the significantdifferences we did obtain in the laboratory experiment dosupport the video data OR and may encourage further studyof surgeons’ posture in both settings.

Table 2. Forceplate data

Forceplate data Center of pressure

Lateral position (cm) Lap 1.0 ± 2.1

Open 1.4 ± 0.8

AP position (cm) Lap −3.8 ± 5.4

Open −5.9 ± 5.9

Lateral range (cm) Lap 0.6 ± 0.1

Open 1.2 ± 1.1

AP range (cm) Lap 0.5 ± 0.1*

Open 0.8 ± 0.2

a Data listed as mean ± standard deviation, AP4 anteroposterior, LAP4laparoscopic.* p < 0.05 Lap vs. Open

Fig. 1. Bar graph depicting the number of movements per minute of thesurgeons’ head and back during laparoscopic(Lap) and open(Open)op-erations.(Lap only refers specifically to the interval during which thelaparoscope is used to view the operation.) Note that during theLap onlyperiod there is less frequent movement of the surgeons’ head and back. Thedata is displayed as the mean ± standard deviation. *p < 0.05 vsopen.

141

In summary, during laparoscopic surgery surgeons ex-hibit more upright—yet more static—head and back posi-tions. The ultimate physical effects of these alterations insurgeons’ posture during laparoscopic surgery warrant fur-ther investigation in the hopes of optimizing the laparoscop-ic surgeon’s working environment.

References

1. Bhatnager V, Drury CG, Schiro SG (1985) Posture, postural discom-fort, and performance. Hum Factors 27: 189–199

2. Corlett EN, Bishop RP (1976) A technique for assessing posturaldiscomfort. Ergonomics 19: 175–182

3. Kant IJ, de Jong LC, van Rijssen-Moll M, Borm PJ (1992) A surveyof static and dynamic work postures of operating room staff. Int ArchOccup Environ Health 63: 423–428

4. Keyserling WM (1986) A computer-aided system to evaluate posturalstress in the workplace. Am Ind Hyg Assoc J 47: 641–649

5. Lee YH, Chiou WK (1995) Ergonomic analysis of working posture in

nursing personnel: example of modified Ovako Working Analysis Sys-tem application. Res Nurs Health 18: 67–75

6. McCann PD, Wootten ME, Kadaba MP, Bigliani LU (1993) A kine-matic and electromyographic study of shoulder rehabilitation exer-cises. Clin Orthop 288: 179–188

7. Patkin M, Isabel L (1995) Ergonomics, engineering and surgery ofendosurgical dissection. J R Coll Surg Edinb 40: 120–132

8. Rademacher K, Pichler KV, Erbse S, Boeckmann W, Rau G, Jakse G,Straudte H (1996) Using human factor analysis and VR simulationtechniques for the optimization of the surgical worksystem. Healthcare in the information age. IOS Press, Amsterdam, pp 533–541

9. Riley PO, Mann RW, Hodge WA (1990) Modelling of the biomechan-ics of posture and balance. J Biomech 23: 503–506

10. Veeger HE, Lute EM, Roeleveld K, van der Woude LH (1992) Dif-ferences in performance between trained and untrained subjects duringa 30-s sprint test in a wheelchair ergometer. Eur J Appl Physiol 64:158–164

11. Whiting WC, Gregor RJ, Finerman GA (1988) Kinematic analysis ofhuman upper extremity movements in boxing. Am J Sports Med 16:130–136

Table 3. Video dataa

Head position Minutes Back position Minutes

Straight Lap 41.5 ± 19.6* Straight Lap 79.6 ± 10.3Open 8.7 ± 2.2 Open 51.6 ± 17.4

Bent Lap 33.5 ± 16.5* Bent Lap 8.3 ± 4.3*Open 82.9 ± 5 Open 31 ± 16.2

Twisted Lap 23 ± 32 Twisted Lap 5.7 ± 4.8Open 3.1 ± 2 Open 3.6 ± 2.2

Bent and twisted Lap 2.1 ± 2.1 Bent and twisted Lap 6.4 ± 3.6Open 5.4 ± 3.9 Open 14.8 ± 11.1

a Data listed as mean ± standard deviation, LAP4 laparoscopic* p < 0.05 Lap vs Open

142

The author replies

We thank Drs. Watson and Mathew for their valuable com-mentary on our paper ‘‘Laparoscopic surgery in the rat:beneficial effect on body weight and tumor take’’ [1].

Our hypothesis is that the reduced surgical trauma oflaparoscopy is associated with less tumor take and growth.Experimental studies in rats have confirmed that tumor takeand growth after laparoscopic surgery are significantly lessthan after laparotomy [1–3]. However, tumor growth andtake in these studies were considered to represent the gen-eral response to surgical trauma. In our opinion, port sitemetastases are caused by local rather than general factors.Direct implantation of tumor cells in the abdominal wallduring extraction of a specimen and displacement of freeviable intraperitoneal tumor cells have been suggested asimportant factors in the development of abdominal wallmetastases. Insufflation of gas in the peritoneal cavity ap-pears to provide an excellent vehiculum to transport freecancer cells to port sites. Leakage of gas through or alongtrocars, a phenomenon we refer to as the ‘‘chimney effect’’[5], probably predisposes to the development of port sitemetastases.

Drs. Watson and Mathew expressed their interest in anexperiment with introduction of tumor cells at the com-mencement. We have recently reported the results of a studyon the impact of CO2 and gasless laparoscopy on abdominalwall metastases in rats [4]. In accordance with the study byWatson and Mathew we found significantly greater port sitemetastases in rats having CO2 laparoscopy compared togasless laparoscopy [6]. Therefore, further study on gaslesslaparoscopy in digestive cancer is required.

We concur with Drs. Watson and Mathew that the roleof laparoscopic surgery in digestive cancer requires furtherevaluation, both in experimental and clinical studies.

References

1. Bouvy ND, Marquet RL, Hamming JF, Jeekel J, Bonjer HJ (1996)Laparoscopic surgery in the rat: beneficial effect on body weight andtumor take. Surg Endosc 10: 490–494

2. Bouvy ND, Marquet RL, Lamberts SWJ, Jeekel J, Bonjer HJ (1996)Laparoscopic bowel resection in the rat: earlier restoration of IGF-1 andless tumor growth. Surg Endosc 10: 567

3. Bouvy ND, Marquet RL, Jeekel J, Bonjer HJ (1996) Laparoscopicsurgery is associated with less tumour growth stimulation than conven-tional surgery: an experimental study. Br J Surg (in press)

4. Bouvy ND, Marquet RL, Jeekel J, Bonjer HJ (1996) Impact of gas(less)laparoscopy and laparotomy on peritoneal tumor growth and abdominalwall metastases. Surg Endosc 10: 551

5. Kazemier G, Bonjer HJ, Berends FJ, Lange JF (1995) Port site metas-tases after laparoscopic colorectal surgery for cure of malignancy. Br JSurg 82: 1141

6. Mathew G, Watson DI, Rofe AM, Baigrie CF, Ellis T, Jamieson GG(1996) Wound metastases following laparoscopic and open surgery forabdominal cancer in a rat model. Br J Surg 83: 1087–1090

N. D. BouvyH. J. Bonjer

Department of SurgeryUniversity Hospital DijkzigtDr. Molewaterplein 403015 GD RotterdamThe Netherlands

Surg Endosc (1997) 11: 166

SurgicalEndoscopy


VideosSurgical


220

Analysis of complications of endoscopic sphincterotomy for biliarystones in a consecutive series of 546 patients

R. Coppola, M. E. Riccioni, S. Ciletti, L. Cosentino, C. Coco, P. Magistrelli, A. Picciocchi

Department of General Surgery, Catholic University School of Medicine, Rome, Italy. Largo A. Gemelli 8. Rome 00168, Italy


AbstractBackground:Endoscopic sphincterotomy (ES) plays an im-portant role in treatment of biliary stones; however, thereremain some controversies concerning complications of ES,which in most cases seem not to be predictable.Methods:The aim of this study was a retrospective analysisof complications in 546 consecutive patients (267 males,279 females, average age 63.7 years) who underwent endo-scopic retrograde cholangiography (ERCP) for biliarystones from 1988 to 1995.Results:ES was performed in 535 patients (98%), and ex-traction of stones was successful in 493 (92%). In all, 29complications (5.4%) were observed, including bleeding 13,cholangitis seven, cholecystitis four, pancreatitis three, ret-roduodenal perforation two; of these, four (14%) requiredan operation. Overall mortality was 0.3%.Conclusion:While a significant decrease of the incidence ofcomplications was observed in the course of the study, dueto constantly improving experience, no correlation betweenrisk factors and complications was identified.

Key words: Endoscopy — ERCP — Complications

In recent years, laparoscopic surgery has reopened old de-bates concerning the safest and most appropriate treatmentfor biliary stones [3].

Ever since the beginning of the laparoscopic era, endo-scopic retrograde cholangiopancreatography (ERCP) hasbeen widely employed due to the possibility of simplifyinglaparoscopic surgery; however, due to the potential severityof related complications, criticism of ERCP has become

frequent among biliary surgeons and some indications havebeen called into question [14].

Three main indications for ERCP are generally ac-cepted: residual or recurrent stones, complications of stones,and preoperative study in cases of suspected stones [18].

Nowadays, primary endoscopic treatment of residual orrecurrent stones is no longer a matter of controversy.

Similarly, numerous studies have demonstrated the use-fulness of ERCP in the treatment of complications of lithi-asis, especially in patients who are at high risk for surgery.An additional element to take into consideration is that highsurgical risk may in itself be an indication for a preoperativeERCP study [18].

Finally, the use of ERCP as a preoperative examinationin cases of suspected biliary stones has some advantageseven though it has not proved to be cost-effective as for theprevious two indications; thus, stricter rules are required forthis indication, considering also the improvements that havebeen made in laparoscopic biliary surgery.

During the last 8 years, in our department of surgery,ERCP was the treatment of choice for bile duct stones; thispaper reports on our experience with 546 patients with bil-iary stones, focusing on the complications of this technique.


During the period from January 1988 to June 1995, 546 consecutive pa-tients with biliary stones were submitted to ERCP. There were 267 malesand 279 females, with an average age of 63.7 years (range 15–93 years).

Three hundred sixty-three patients presented a complication of stonesat the time of the endoscopic examination; 151 of these patients had jaun-dice, 144 cholangitis, and 68 acute pancreatitis. In this paper we refer tothis group of patients as the ‘‘complications of stones’’ group.

Only patients with jaundice, fever, and pain were included in the chol-angitis group. The majority of these patients (88/1444 61%) underwentERCP 24–48 h after admission; 21 patients (15%) underwent emergencyES less than 24 h after admission due to serious general condition; finally,35 patients (24%) were submitted to ERCP more than 48 h after admission.

Patients with both severe abdominal pain, localized in the upper quad-rants, and hyperamylasemia were included in the acute pancreatitis groupand classified using Ranson’s criteria.

Seven of the 68 patients with acute pancreatitis (10%) had more than

Presented at the annual meeting of the Society of American Gastrointes-tinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA,13–17 March 1996Correspondence to:R. Coppola, Largo G. Rovani 7. Rome, 00137, Italy

Surg Endosc (1997) 11: 129–132

SurgicalEndoscopy


four Ranson’s criteria and were submitted to ERCP within 24 h of admis-sion; 39 patients (58%) had three Ranson’s criteria and underwent ERCP24–48 h after admission; finally, 21 patients (32%) with less than threeRanson’s criteria underwent ERCP more than 48 h after admission.

One hundred thirty patients underwent ERCP for residual or recurrentstones not complicated at the time of the endoscopy. Diagnosis of biliarystones relied frequently on an ultrasound examination performed for re-lapsing colicky pain; this group of patients is identified in our paper as the‘‘residual stones’’ group.

Finally, 53 patients received an ERCP—before being submitted tocholecystectomy—for previous history of jaundice, cholangitis, or pancre-atitis, as well as for dilation of the common bile duct at ultrasonography orelevation of liver function tests. This is the ‘‘preoperative diagnosis’’group.

ERCP was performed with the standard technique. Following cholan-giographic confirmation of stones, endoscopic sphincterotomy was at-tempted at once with a long-nose sphincterotome when possible, or usingthe diathermic needle for performing precut papillotomy.

Fogarty catheters and Dormia baskets were employed for extraction ofstones. Huge stones were fragmented using a mechanical lithotriptor, orExtracorporeal Lithotripsy Shock Waves (ESWL) in selected cases.

Results

Endoscopic sphincterotomy was successfully completed in535 of the 546 patients with biliary stones (98%). In 421patients (79%) ES was performed using a long-nose sphinc-terotome over a guide wire; in 23 cases (4%) a direct can-nulation of the papilla with a long-nose sphincterotome waspossible without the need of a guide wire; finally, in 91patients (17%), due to the impossibility of cannulation ofthe papilla, a precut papillotomy was performed using adiathermic needle. Technical failures (11 cases) were due tounsuccessful selective cannulation of the common bile ductin four cases; gastric resection in four cases, and to thepresence of perivaterian diverticula in three cases. Therewas no significant difference in the success rate among thethree groups of patients (complications of stones, residualstones, preoperative diagnosis); thus, failures of endoscopicsphincterotomy were most likely independent of indica-tions.

Complete extraction of stones was obtained in 92% ofcases following sphincterotomy. Failures (42 cases) weredue mainly to the size and number of stones (15 hugestones, 11 multiple stones, four intrahepatic stones, fourcases of Mirizzi’s syndrome, two impacted stones) as wellas to concomitant benign stenosis of the common bile duct(three cases), which would have made extraction of stonesexcessively difficult and time consuming. In two casesbleeding after endoscopic sphincterotomy prevented com-plete extraction of the stones. Finally, a young patient withCaroli’s disease was submitted to hepaticojejunostomy.

Our experience shows a different success rate of stoneextraction according to indications. The best results wereobtained in the preoperative diagnosis group (success rate of98%), while patients presenting a complication of lithiasiswere the most difficult to manage (success rate of 91%).However, in this group of patients ERCP allowed for theurgent resolution of the complication (for example, in casesof cholangitis) even though complete extraction of thestones could only be completed during elective surgery.

In all, 29 complications of ERCP and endoscopicsphincterotomy were recorded, giving an overall complica-tion rate of 5.4%.

Bleeding was the most frequent complication (13 cases

4 2.4%), followed by cholangitis (7 cases4 1.3%), cho-lecystitis (4 cases4 0.7%), pancreatitis (3 cases4 0.5%),and retroduodenal perforation (2 cases4 0.4%).

In two of these patients (2/294 7%) the complicationarose after a precut papillotomy was performed (one bleed-ing, one pancreatitis).

The average age of patients complicated with acutepost-ERCP pancreatitis was lower than the age of the otherpatients; there were no residual stones among these, nor wasthe papilla traumatized in the course of cannulation. In one,a precut papillotomy was employed; however, in the othertwo, sphincterotomy was performed very easily and on axis.

There was no difference in sex distribution of patientswith complications.

Comparing the three groups of patients, more compli-cations are found in the group with preexisting complica-tions of stones (22/3634 6%) than in the residual stonesgroup (5/1304 3.8%) or the preoperative diagnosis group(2/534 3.8%); however, there is no statistically significantdifference among the three groups.

Four of the 29 complications required a surgical opera-tion (14%): One patient with severe bleeding from sphinc-terotomy was submitted to urgent operation; the remainingthree patients underwent delayed operations for additionalcomplications resulting from retroduodenal perforation (twocases) and acute pancreatitis (one case).

One of the patients with perforation died in the postop-erative period due to sepsis. A second cirrhotic patient, withsevere portal hypertension, died due to liver failure follow-ing bleeding. Thus, the overall mortality rate in our expe-rience was 0.3% (2/535).

Discussion

Endoscopic sphincterotomy plays an important role in thetreatment of biliary stones due to its safety and efficacy.Unfortunately, complications of ES, in most cases, are notpredictable and may occur in young patients as well as inlow-surgical-risk patients [2, 7, 18].

Recently, laparoscopic biliary surgery has gained muchinterest among surgeons, becoming a potential substitute forERCP [3, 14].

At the beginning of the laparoscopic era ERCP wasperhaps too widely employed, due to the new possibilities ofsimplifying laparoscopic surgery.

The first historical application of ES was in patientswith residual or recurrent stones following a previous bili-ary operation [6, 15]; this still represents the main indicationfor ERCP. In these cases the success rate of endoscopy isvery high, thanks to continually increasing expertise as wellas improving instrumentation; in fact, the availability of themechanical lithotriptor and more recently of ESWL in con-junction with ERCP has significantly reduced the number ofrepeated surgical operations for biliary stones.

The second important indication for ERCP is as a treat-ment for patients affected by complications of lithiasis:cholangitis, pancreatitis and jaundice.

Before the introduction of endoscopic drainage acutecholangitis was a life-threatening disease with a very highmortality rate. Endoscopy represented real therapeutic prog-ress; it was now possible to prevent the serious septic com-

130

plications of cholangitis simply by performing ES with ap-plication of a nasobiliary drain. Subsequently, several stud-ies comparing the morbidity and mortality rates of surgeryfor acute cholangitis with those of endoscopy [16, 19] haveconfirmed that there is a clear superiority of endoscopy,especially in high-risk patients.

The same considerations hold for patients with acutepancreatitis: A number of prospective studies [13] havedemonstrated a significant reduction of morbidity and mor-tality in patients submitted to urgent ERCP for acute biliarypancreatitis.

Finally, in patients with jaundice, ERCP is not onlyuseful for achieving diagnosis—it also allows for simulta-neous extraction of stones, although there is no prospectivestudy proving its efficacy.

The third indication of ERCP analyzed in this paper isthe preoperative study of patients with suspected stones. Inour previous laparoscopic experience [5], in which therewas a wide application of ERCP as a preoperative study,about one patient in four was selected for this exam on thebasis of criteria for evaluating the risk of stones, such asdilation of CBD at ultrasound, elevation of LFTs, or com-plications of lithiasis. Since stones were found in less thanhalf of these patients it might be objected that the costs ofthis diagnostic protocol are not justified, even though fur-ther useful information can be obtained from such a preop-erative study.

If we consider our overall endoscopic experience, how-ever, we find that the most important indication for ERCP iscomplications of lithiasis (66% of cases), while the exami-nation is used more rarely as a preoperative study (10% ofcases).

In this study we have attempted to identify the riskfactors associated with each type of complication encoun-tered.

Bleeding was the most frequent complication. It is wellknown that the presence of coagulation disorders representsan elevation of risk of bleeding when performing endo-scopic sphincterotomy; likewise, the use of precut papil-lotomy or reiterative papillotomy after previous surgical orendoscopic papillotomy may increase the risk of bleeding[1, 6, 8]. Some authors attempt to explain bleeding by thepresence of an inflammatory stenosis of the papilla [17]. Inour 13 patients suffering from this complication we wereable to identify these risk factors in only four patients—thatis to say, in about 30% of cases.

More significant was the association between incom-plete extraction of stones and cholangitis: In fact, of theseven cases of acute cholangitis, five were due to an incom-plete extraction of the stones. Similarly, the main cause ofcholecystitis was obstruction of the cystic duct by stones.

With regard to acute post-ERCP pancreatitis, two of thethree patients were less than 50 years of age; they also hadmicrolithiasis and a normal pancreas documented by ultra-sound and pancreatography. For the third patient, the precutpapillotomy technique was employed. None of these pa-tients had a long common channel, nor was any difficultyencountered during cannulation of the papilla. In none ofthese patients was there an overinjection of the pancreaticduct system during the diagnostic phase of the examination[2, 4].

Retroduodenal perforation seems to occur more fre-

quently in the presence of a nondilated distal common bileduct or of a small papilla. In fact, in such cases, whileperforming sphincterotomy it is all too easy to accidentallycut through the duodenal wall into the retroduodenal space,which is what happened in both of our patients. We neverencountered this complication with the precut technique orin the presence of perivaterian diverticula [2].

In our experience both retroduodenal perforations re-quired surgical operation, and this contrasts with severalother reports [7]. In fact, retroduodenal perforation may besuccessfully treated by medical antibiotic therapy associatedwith a nasobiliary drain and a nasogastric tube to avoidspillage of bile and intestinal contents into the retroduodenalspace. Although this treatment was immediately carried outin our patients, unfortunately they both developed a retro-peritoneal abscess, which required surgical drainage. Inthese complicated cases, the relative risk of subsequent sur-gery is very high [7]; in fact, one of our patients died due touncontrolled sepsis.

If we analyze the incidence of the different complica-tions during the period of this work, dividing the entireperiod into four 2-year segments (Fig. 1), it is evident thattotal complications show a tendency to decrease, from10.3% to 2.1%, and this difference is statistically significant(p < 0.05). This is mainly due to improved endoscopic ex-pertise and improved instrumentation.

Table 1. Complications and mortality rates following endoscopic sphinc-terotomy: collective review of 24,091 cases

Author No. patients Complications Mortality

Safrany, 1978 [15] 3,618 7.0% 1.4%Greenen, 1981 [10] 1,250 8.7% 1.2%Cotton, 1982 [6] 865 11.0% 1.0%Vaira, 1989 [20] 1,000 6.9% 1.2%Lambert, 1991 [11] 602 10.0% 2.2%Vennes, 1991 [7] 1,829 7.6% 0.2%Liguory, 1991 [7] 2,881 8.7% 0.9%Armengol-Miro, 1991 [7] 2,981 3.8% 0.4%Landoni, 1993 [12] 547 5.6% 1.8%Siegel, 1994 [19] 839 6.0% 0.4%Vavrecka, 1994 [21] 2,000 3.4% 0.5%Freeman, 1994 [9] 1,494 11.7% 0.5%Bernardi, 1995 [1] 887 9.2% 0.2%Ell, 1995 [8] 2,752 8.3% 0.6%Our experience, 1995 546 5.4% 0.3%

Fig. 1. Incidence of the different complications during the period of thiswork. The entire period is divided into four 2-year segments. It is evidentthat total complications show a tendency to decrease, from 10.3% to 2.1%,and this difference is statistically significant (chi-square4 8.016, p <0.05).

131

Table 1 provides an overview of the literature concern-ing incidence of complications and mortality following en-doscopic sphincterotomy for biliary stones: This review col-lects 24,091 patients, all in studies of more than 500 cases,from 1978 to 1995. The complication rate ranges from 3.4%to 11.7% and the mortality rate from 0.2% to 2.2%. Table 2gives a breakdown of the complications in 20,660 cases,according to type.

Our experience corresponds very well with these studies,showing that while ES is very effective, it must be em-ployed cautiously, like any other surgical procedure.

To sum up the results of our experience, endoscopicsphincterotomy for biliary stones was technically feasible in98% of cases. Complete extraction of stones was successfulin 92% of patients, with a 5.4% incidence of complications.Incidence of complications was very low in the preoperativediagnosis group, which includes younger patients as well aslow-risk patients. Surgery for complications of endoscopywas required in 0.9% of the patients. Overall mortality ac-counted for 0.3%. Our study suggests that there is an inverserelationship between the incidence of complications and theexperience of the endoscopic team: the greater the experi-ence, the fewer the complications.

Finally, although we were able to identify some riskfactors for the complications we observed, there appears tobe no precise correlation between risk factors and compli-cations; this is especially true for pancreatitis, which oc-curred independently of endoscopic experience.

References

1. Bernardi M, Havelka J, Zaugg PY, Buhler H (1995) Endoscopic pap-illotomy in benign biliary tract diseases: nature, frequency and severityof complications. Schweiz Med Wochenschr 125(15): 727–730

2. Boender J, Nix GA, de Ridder MA, van Blankenstein M, Schutte HE,Dees J, Wilson JH (1994) Endoscopic papillotomy for common bileduct stones: factors influencing the complication rate.Endoscopy26(2): 209–216

3. Brodish RJ, Fink AS (1993) ERCP, cholangiography and laparoscopiccholecystectomy. The Society of American Gastrointestinal Endo-scopic Surgeons (SAGES) opinion survey. Surg Endosc 7: 3–8

8. Chen YK, Foliente RL, Santoro MJ, Walter MH, Collen MJ (1994)Endoscopic sphincterotomy-induced pancreatitis: increased risk asso-ciated with nondilated bile duct and sphincter of Oddi dysfunction. AmJ Gastroenterol 89(3): 327–333

5. Coppola R, D8Ugo D, Ciletti S, Riccioni ME, Cosentino L, Magistrelli

P, Picciocchi A (1996) ERCP in the era of laparoscopic biliary surgery.Experience with 407 patients. Surg Endosc 10: 403–406

6. Cotton PB, Frost RG (1982) Computer analysis of a decade of ERCP.Gut 23: A432

7. Cotton PB, Lehman G, Vennes J, Geenen JE, Russel RCG, MeyersWC, Liguory C, Nickkl N (1991) Endoscopic sphincterotomy, com-plications and their management. An attempt at consensus. Gastro-intest Endosc 37: 383–393

8. Ell C, Rabenstein T, Ruppert T, Forster P, Hahn EG, Demling L(1995) 20 years of endoscopic papillotomy. Analysis of 2752 patientsat Erlangen Hospital. Dtsch Med Wochenschr 120(6):163–167

9. Freeman M, Nelson D, Sherman S, Herman M, Haber G, Moore J,Dorshew P, Mackie R, Ryan M, Janidar P, Shaw M, Silverman W,Cumingham J, Man K, Silvis S, Yakshe P, Fennerty M, DiSario J, LoS, Logan G, Pheley A (1994) Complications of endoscopic sphincter-otomy. A prospective multicenter, 30 day out-come study. Gastroen-terology 106: 338 (Abstract)

10. Geenen JE, Vennes JA, Silvis SE (1981) Resume´ of a seminar onendoscopic retrograde sphincterotomy (ERS). Gastrointest Endosc 27:31–38

11. Lambert ME, Betts CD, Hill J, Faragher EB, Martin DF, Tweedle DEF(1991) Endoscopic sphincterotomy: the whole truth. Br J Surg 78:473–476

12. Landoni N, Chopita N, Jmelnitzky A (1993) Endoscopic sphincterot-omy. Complications and follow-up. Acta Gastroenterol Latinoam23(1): 27–31

13. Neoptolemos JP, Carr-Locke DL, London NJ, Bailey IA, James D,Fossard DP (1988) Results of a controlled trial of urgent ERCP andendoscopic sphincterotomy in patients with acute pancreatitis due togallstones. Lancet ii: 979–983

14. Phillips EH, Liberman M, Carroll BJ, Fallas MJ, Rosenthal RJ, HiattJR (1995) Bile duct stones in the laparoscopic era. Arch Surg 130:880–886

15. Safrany L (1978) Endoscopic treatment of biliary-tract disease. Lancetii: 983–985

16. Sheridan WG, Williams HOL, Lewis MH (1987) Morbidity and mor-tality of common bile duct exploration. Br J Surg 74: 1095–1099

17. Sherman S, Ruffolo TA, Hawes RH, Lehman GA (1991) Complica-tions of endoscopic sphincterotomy. A prospective series with empha-sis on the increased risk associated with sphincter of Oddi dysfunctionand nondilated bile ducts. Gastroenterology 101: 1068–1075

18. Siegel JH (1992) Endoscopic retrograde cholangiopancreatography:technique, diagnosis and therapy. Raven Press, New York

19. Siegel JH, Rodriquez R, Cohen SA, Kasmin FE, Cooperman AM(1994) Endoscopic management of cholangitis: critical review of analternative technique and report of a large series. Am J Gastroenterol89(8): 1142–1146

20. Vaira D, Ainley C, Williams S, Cairns S, Salmon P, Russel C, D8AnnaL, Dowsett J, Baillie J, Croker J, Cotton P, Hatfield A (1989) Endo-scopic sphincterotomy in 1000 consecutive patients. Lancet ii: 431–433

21. Vavrecka A, Jancula L, Novotny I, Minarik P (1994) Endoscopictherapy of biliary tract diseases. Bratisl Lek Listy 95(9): 421–424

Table 2. Type and incidence of complications following endoscopic sphincterotomy: collective reviewof 20,660 cases

Author Bleeding Pancreatitis Perforation Cholangitis Others Surgery

Safrany, 1978 [15] 2.5% 1.3% 1.1% 1.3% 0.7% —Geenen, 1981 [10] 2.3% 3.2% 1.1% 2.0% 0% —Cotton, 1982 [6] 3.2% 3.0% 1.1% 2.1% 1.3% —Vaira, 1989 [20] 3.9% 0.9% 0.5% 1.6% 0% —Lambert, 1991 [11] 4.8% 2.7% 0.7% 2.1% 1.0% —Vennes, 1991 [7] 3.0% 3.1% 0.5% 0.9% 0.1% 0.8%Liguory, 1991 [7] 3.1% 1.4% 2.1% 1.5% 0.6% 1.1%Armengol-Miro, 1991 [7] 1.2% 1.3% 0.4% 1.4% 0 0.7%Siegel, 1994 [19] 3.0% 2.8% 0.2% 0 0 0.2%Freeman, 1994 [9] 1.9% 6.7% 0.5% 0.9% 2.9% —Ell, 1995 [8] 2.7% 1.5% 0.5% 1.7% 1.4% —Our experience, 1995 2.4% 0.5% 0.4% 1.3% 0.7% 0.9%

132

Editorial

The impact of technologies on minimally invasive therapy

This issue ofSurgical Endoscopycontains a new TechnologySection set up by the two Editors in Chief with Dr. A. Melzeras the Technical Editor in charge of this Section. The primaryobjective is to review existing and emerging technology relat-ing the whole spectrum of minimally invasive therapy or mini-mal access therapy. However, the preparation of the first Tech-nology Section raised various issues relating to definitions,skills versus technology, co-operation between research clini-cians and scientists/technologists and the anticipated futuremultidisciplinary therapeutic approach by disease-related treat-ment groups forming a coherent team of specialists from vari-ous disciplines working in close co-operation rather than asseparate disciplines, with obvious gain in terms of the appro-priate intervention in the individual case and improved patientoutcome. The emergence of tomographic endosurgery coveredin this issue exemplifies this approach which we must adopt aswe approach the next millennium.

Definitions

Some ten years ago Wickham and Fitzpatrick introduced thephrase‘Minimally Invasive Therapy’(MIT) [1]. Although itlacks semantic accuracy and carries connotations of increasedsafety [2], this terminology has gained widespread acceptanceand, indeed, it describes the essential goal of these endoscopicsurgical and radiological interventions, i.e., the reduction of theoperative traumatic insult without compromise of the thera-peutic benefit. Since then, there has been considerable debateon alternative terminology such as minimal access therapy(MAT). Contention is avoided if the two terms MIT and MATare considered to be synonymous with each comprising:

● minimally invasive surgery (MIS) or minimal access sur-gery (MAS) or endoscopic surgery (ES)

● interventional flexible endoscopy (EFE)● interventional radiology (IR)

The importance of identification of these three compo-nents of MIT/MAT lies in the increasing combined use oftwo or all of these components in specific MIT/MAT treat-ment strategies for complex disorders, e.g., ES with EFE inthe laparoscopic extraction of ductal calculi and ultrasoundguided laparoscopic cryotherapy for hepatic metastases [3].

The combination of real time tomographic imaging, inter-ventional radiology with frameless stereotactic navigation andendoscopic surgery has lead to the increasing use of imageguided therapy or surgery (IGT/IGS). Although descriptive,this terminology does not distinguish between surface viewingby the endoscope and tomographic imaging by fluoroscopy,ultrasound, CT and MRI. The essence of this exciting newMIT approach is the combination of surface viewing of the

operative field with near real time tomographic imaging of theoperative tissue volume. The benefit of this dual imaging ap-proach is increased precision of the therapeutic effect espe-cially when frameless stereotaxis is used, thereby allowingfurther reduction in the size of the instruments and endoscopes,and the operative insult during the intervention. The currentpopular terminology for this evolving dual approach is imageguided endoscopic microsurgery (IEM) [4] although the cur-rent imaging systems do not, as yet, provide a resolution higherthan 1mm3. Alternative terms include tomographic endo-scopic microsurgery or, simply tomographic endosurgery(TES). IEM (or TES) represents the next phase in the reduc-tion of the operative traumatic insult as it will lead to the useof microscopic instruments and sub-millimeter ports within theforeseeable future. The ultimate goal of MIT or MAT is non-invasive treatment with no instrument or ports being neces-sary. Although examples of this can be cited such as focusedhigh intensity ultrasound for the ablation of hepatic tumours(tractless surgery), it is unlikely that this approach will havemore than a limited application in the wide spectrum of thesurgical disorders.

Technology vs surgical skill

The significant advances in MAT and MAS have been pos-sible as a result of the combined influence of technologicaladvances and skills of the operators. In this respect, the endo-scopic operations and interventions that are performed today,and new procedures that are likely to come in the future, do notdepend on technological advances alone; the latter facilitatethe execution or make an operation possible. In the final analy-sis, safe and efficient execution is dependent on the uniqueskills of the surgeon or interventionalist. For example, thecurrent MAT of localized coronary artery stenosis involves aradiological transvascular approach with dilatation of the ste-nosis and insertion of a stent to maintain luminal patency.Bypass surgery is required for more extensive coronary dis-ease involving several vessels. Over the past few years avideothoracoscopic approach to coronary artery bypass sur-gery, avoiding median sternotomy has been introduced in afew pioneering centers. This significant development is clearlythe outcome of unique surgical endoscopic skills based on theextensive experience of pioneering cardiac endo-surgeons.This example illustrates the fact that technology makes a pro-cedure possible, but in the end it is the surgeon or interven-tional radiologist who must use the technology to execute theoperation. Similar considerations apply to the need for con-version to open surgery during endoscopic surgery of the ab-domen and chest. This is and always will be a fine balancebetween surgical experience, clinical judgment and availabletechnology.

SurgicalEndoscopy


Future objectives of MIS/MAS

The broad objectives which govern the future developmentof MAS interventions can be itemized as follows:

● Accurate and complete pre-surgical diagnosis● On-line tomographic imaging of the operative tissue volume● Presurgical planning of procedure and access trajectories● Further minimization of the access to the operative site● Improved technology for in-situ ablation and tissue ap-

proximation

Aside from the imaging system (Hopkin’s rod lens tele-scope and CCD camera), most of the current instrumentationused in MAS has been developed by pioneers with techno-logical knowledge and expertise in endoscopic surgery such asWittmoser, Semm, Buess etc., working in close co-operationwith industry. This association from the early design stage tothe final prototype and its evaluation is crucial and, if anything,will assume greater importance in the R&D of sophisticatedtechnologies and imaging systems in the future. There is an-other factor. It seems likely with the introduction TES thatcomplex technologies that are currently used by disciplinesother than surgery, e.g., tomographic imaging by open MRIand hybrid fluoroscopy-spiral CT system, will be increasinglyused in the future. Obviously expectations that surgeons in thenext century would be required to be totally familiar withsurgical tomography and advanced techniques for in-situ tissueablation is unrealistic. As procedures become more complexrequiring sophisticated imaging and technology, expert thera-peutic intervention will necessitate a multidisciplinary teamconsisting of surgeons, radiologists, flexible endoscopists andphysicists working together on specific disorders in accor-dance with their clinical interests—disease-related treatmentgroups. A patient with a specific disorder will be assessed bythis team and the best MAT option depending on nature, stageof the disease and general condition will be determined by thegroup. In this fashion the optimal interventional therapeuticapproach would be individualized to the patient’s needs.

The co-operation with industry in technological and imag-ing Research and Development (R&D) works both ways. Fora productive association, the scientists working in industry andhigher research institutions must understand the concepts be-hind the intended therapeutic advances. This requires somebasic knowledge of the medical and surgical principles in-volved. In practice, this translates to whether the two partiescan communicate and understand each other. The only validpractical approach to this problem is education of cliniciansinvolved in R&D in the basic principles of high technologyrelated to MAT development concurrent with medical instruc-tion of scientists, engineers and technologists involved in thisco-operative research. Joint regular workshops on ‘Surgerymeets Technology’ and information in journals such as theTechnology Section introduced in this and other journals areobvious ways of achieving this mutual understanding.

Tomographic Endosurgery (TEM)

The first session of the new Technology Section ofSurgicalEndoscopydeals with the clinical results and technical data oftomographic interventions and future aspects of tomographicendosurgery [6]. Tomographic guidance and endoscopic sur-gery have to be regarded as complementary technologies. The

endoscope provides a clear sharp colored image of the surfacefeatures of the operative field, whereas the on-line tomo-graphic imaging provides data on the sub-surface nature of theentire operative tissue volume of the same operative field. Thisdual imaging yields the most comprehensive information ofthe anatomy, parenchymal details and intra-parenchymal pa-thology of all the organs and structures contained within theoperating zone. In addition, the on-line tomographic informa-tion provides accurate information of the access trajectories ofthe instruments used for the intervention in real time and, ifframeless stereotaxis is deployed, the exact position of thepoint of the instrument in space (x,y,z location) can be mappedon to the tomographic image of the operative tissue volume inreal time. The precision and safety of the manipulations par-ticularly in anatomically crowded and dangerous areas arethereby assured.

A further application of fast tomographic image guidanceis in the field of surgical robotics and master slave manipula-tors that are under development by several groups [7–9]. Allthese systems are motorized and their effective control requiresfail-safe systems in the as low as is reasonably possible(ALARP) region based on software guided by an externalposition control system, i.e., near real time tomographic im-aging. The problem with such an image guided operating sys-tem (IGOS) is cost, and careful studies will be needed toestablish the cost benefit to patient care with the introductionof these systems. If they convert a major MAS to a minimalMAS such that the procedure can be performed under localanaesthesia with immediate discharge, then specialized TESunits equipped with this technology will flourish. As shown byDr. Seibel’s review, spinal TES has already reached this stage.Indeed, it has come of age.

References1. Wickham J. Editorial. Min Invas Ther & Allied Tech 1996: 12. Cuschieri A, Buess G. Nature and scope of endoscopic surgery. In:

Operative Manual of Endoscopic Surgery, Cuschieri A, Buess G, Peris-sat J (eds) 1992; Springer: Berlin, pp 9–13

3. Cuschieri A, Crosthwaite G, Shimi S, Pietrabissa, Joypaul V, Tait I,Naziri W. Hepatic cryotherapy for liver tumours: development andclinical evaluation of a high-efficiency multineedle probe system foropen and laparoscopic use. Surg Endosc 1995: 483–489

4. Buess G. IEM or what shall we call it? Min Invas Ther & Allied Tech1996: 1

5. Buess G. Why this journal? End Surg 1993: 1–26. Seibel R. Image guided minimal invasive therapy. Surg Endosc 1997:

154–1627. Melzer A, Schurr MO, Kunert W, Buess G. Intelligent surgical instru-

ment system. Concept and preliminary experimental application ofcomponents and prototypes. End Surg 1993: 165–170

8. Satava R, Simon LB. Teleoperation, telerobotics, and telepresence sur-gery. End Surg 1993: 151–153

9. Schurr MO, Melzer A, Kunert W, Dautzenberg P, Neisius B, BretwieserW, Buess G. Experimental evaluation of components of an endoscopicmanipulator system Artemis. October 1989, SMIT 6th InternationalMeeting, Berlin

A. Cuschieri

Department of SurgeryNinewalls Hospital and Medical SchoolDundee, United Kingdom

A. Melzer

University Witten/HeidedreSchulstrasse 10Mulheim, Germany

92

News and notices

New Address for the European Association forEndoscopic Surgery (E.A.E.S.)

Effective January 1, 1997, the new correspondence, telephone, and faxnumbers of the E.A.E.S. office are:

E.A.E.S. Office, c/o Mrs. Ria Palmen

Luchthavenweg 81Unit 1.425657 EA EindhovenThe Netherlandsor: P.O. Box 3355500 AH VeldhovenThe NetherlandsTel: +31 40 2525288Fax: +31 40 2523102

Fellowship in Minimally Invasive SurgeryGeorge Washington Medical CenterWashington, DC USA

A one-year fellowship is being offered at the George Washington Univer-sity Medical Center. Interested candidates will be exposed to a broad rangeof endosurgical Education and Research Center. Active participation inclinical and basic science research projects is also encouraged.

For further information, please contact:

Carole Smith202-994-8425

or, send curriculum vitae to:

Dr. Jonathan M. SackierDirector of Endosurgical Education and ResearchGeorge Washington University Medical CenterDepartment of Surgery2150 Pennsylvania Avenue, N.W.6B-417Washington, DC 20037, USA

Essentials of Laparoscopic SurgerySurgical Skills UnitUniversity of DundeeScotland, UK

Under the direction of Professor A. Cuschieri the Surgical Skills Unit isoffering a three-day practical course designed for surgeons who wish toundertake the procedures such as laparoscopic cholecystectomy. This in-tensely practical program develops the necessary operating skills, empha-sizes safe practice, and highlights the common pitfalls and difficultiesencountered when starting out. Each workshop has a maximum of 18participants who will learn both camera and instrument-manipulation skillsin a purpose-built skills laboratory. During the course there is a live dem-onstration of a laparoscopic cholecystectomy. The unit has a large libraryof operative videos edited by Professor Cuschieri, and the latest books onendoscopic surgery are on display in our Resource area. Course fee in-cluding lunch and course materials is $860.

For further details and a brochure please contact:

Julie Struthers, Unit Co-ordinatorSurgical Skills UnitNinewells Hospital and Medical SchoolDundee DD1 9SYTel: +44 382 645857Fax: +44 382 646042

Advanced Endoscopic SkillsSurgical Skills UnitUniversity of DundeeScotland, UK

Each month Professor Cuschieri Surgical Skills Unit offers a 41⁄2 daycourse in Advanced Endoscopic Skills. The course is intensely practicalwith ‘‘hands on’’ experience on a range of simulated models. The programis designed for experienced endoscopic surgeons and covers advanceddissection techniques, extracorporeal knotting techniques, needle control,suturing, internal tying technique, stapling, and anastomotic technique.Individual workstations and a maximum course number of 10 participantsallows for personal tuition. The unit offers an extensive collection of sur-gical videos and the latest books and publications on endoscopic surgery.In addition, participating surgeons will have the opportunity to see liveadvanced laparoscopic and/or thoracoscopic procedures conducted by Pro-fessor Cuschieri and his team. The course is endorsed by SAGES. Coursefee including lunch and course materials is $1850.



The Practical Aspects of Laparoscopic FundoplicationSurgical Skills UnitUniversity of DundeeScotland, UK

A three-day course, led by Professor Cuschieri, designed for experiencedlaparoscopists wishing to include fundoplication in their practice. Thecourse covers the technical details of total and partial fundoplication usingsmall group format and personal tuition on detailed simulated models.There will be an opportunity to observe one of these procedures live duringthe course. Maximum course number is six. Course fee including lunch is$1850.



SurgicalEndoscopy


Courses at George Washington University EndosurgicalEducational and Research Center

George Washington University Endosurgical Educational and ResearchCenter is proud to offer a wide range of surgical endoscopy courses. Thesecourses include advanced laparoscopic skills such as Nissen fundoplica-tion, colon resection, common bile duct exploration, suturing, as well assubspecialty courses. Individual surgeons needs can be met with privatetuition. The Washington D.C. area is a marvelous destination to visit forrecreational pursuits which can be arranged by the facility to suit yourpersonal agenda.

For further details please contact:

Carole Smith:Department of Surgery2150 Pennsylvania Avenue NW6BWashington, DC 20037, USATel: (202) 994-8425

Courses at the Royal Adelaide Centre forEndoscopic Surgery

Basic and Advanced Laparoscopic Skills Courses are conducted by theRoyal Adelaide Centre for Endoscopic Surgery on a regular basis. Thecourses are limited to six places to maximize skill development and tuition.Basic courses are conducted over two days for trainees and surgeons seek-ing an introduction to laparoscopic cholecystectomy. Animal viscera insimulators is used to develop practical skills. Advanced courses are con-ducted over four days for surgeons already experienced in laparoscopiccholecystectomy who wish to undertake more advanced procedures. Awide range of procedures are included, although practical sessions can betailored to one or two procedures at the participants request. Practical skillsare developed using training simulators and anaesthetised pigs.

Course fees: $A300 ($US225) for the basic course and $A1,600($US1,200) for the advanced course.

For further details and brochure, please contact:

Dr. D. I. Watson or Professor G. G. JamiesonThe Royal Adelaide Centre for Endoscopic SurgeryDepartment of SurgeryRoyal Adelaide HospitalAdelaide SA 5000 AustraliaTel: +61 8 224 5516Fax: +61 8 232 3471

Advanced Laparoscopic Suturing and SurgicalSkills Courses

MOET InstituteSan Francisco, CA, USA

This intensive hands-on training program is intended to help the surgeondevelop proficiency in the essential laparoscopic surgical techniques.A sequence of progressively challenging exercises has been designedto enable the surgeon to improve his or her laparoscopic dexterity,efficiency, and creativity. Exact and meticulous technique is emphasized sothat the surgeon can apply these skills with confidence. Personal instructionis provided by Zoltan Szabo, Ph.D., F.I.C.S., Director of the MOETInstitute, and surgeons are allowed to progress their own pace. Each par-ticipant has sole use of a laparoscopic training station equipped withhigh-quality clinical laparoscopic equipment and instrumentation. In-animate, animal tissue, and optional live animal models are utilized. Fea-tures of these program include: fluently choreographed instrumentmovements; economy of movement and ‘‘flawless’’ technique; needle andsuture handling skills (2-0 to 7-0); precision suturing, knotting, ligature,and anastomosis techniques; atraumatic, hemostatic tissue handlingand dissection; optimal angles of approach (coaxial alignment of setupand geometry of port positioning); laparoscopic surgical strategy, tech-

nical nuances, and troubleshooting; visual perception problems and solu-tions; magnified eye-hand coordination; and two-handed (ambidextrous)technique.

Courses are offered year-round by individual arrangement. The MOETInstitute is accredited by the Accreditation Council for Continuing MedicalEducation (ACCME) to provide continuing medical education for physi-cians and designates these CME activities for 20–40 credit hours in Cat-egory 1 of the Physician’s Recognition Award of the American MedicalAssociation. These programs are also endorsed by the Society of Gastro-intestinal Endoscopic Surgeons (SAGES).


Wanda Toy, Program AdministratorMicrosurgery & Operative Endoscopy Training (MOET) Institute153 States StreetSan Francisco, CA 94114, USATel: (415) 626-3400Fax: (415) 626-3444

IIIrd European Workshop on Digestive SurgeryMarch 13–14, 1997Brussels, Belgium

The IIIrd European Workshop on Digestive Surgery is focused on liveoperative demonstrations performed and narrated by European experts.Panel discussions, quizzes, and selected communications will take placeduring the sessions. The aims are to confront alternative procedures and toprovide young surgeons with an overview of selected fields. Topics will be:functional anorectal surgery, proctology, colorectal surgery, and updates inlaparoscopic surgery. Course direction: J. J. Houben, MD.


Administrative SecretariatConference Services s.a.Avenue de l’Observatoire 3, bte 17B-1180 Brussels, BelgiumTel: +32 2 375 16 48Fax: +32 2 375 32 99

Medicine and the Law for Junior Hospital DoctorsApril 11, 1997Middlesbrough, UK


Miss WelshENTERNorth Riding InfirmaryNewport RoadMiddlesbrough TS1 5JEUK

First International Baltic Conference of Videosurgery ofthe Viscerosynthesis Section of the Association ofPolish Surgeons

April 24–27, 1997Gdansk, Poland


Organizing SecretariatSecond Department of SurgeryMedical University of Gdansk1 Prof. Z Kieturakisa Street80-742 Gdansk, PolandTel/Fax: (0 048 58) 31 87 75

222

4th International Meeting on Laparoscopic SurgeryMay 1–7, 1997Berne, Switzerland

Main topic: Acute appendicitis: Standard treatment and the role of lapa-roscopic surgery


Mrs. Caroline ZürcherKlinik für Viszerale und TransplantationschirurgieUniversität BernInselspitalCH-3010 Bern, SwitzerlandTel: +41 31 632 97 22Fax: +41 31 632 97 23

European Course on Laparoscopic Surgery(French language) May 13–16, 1997(English language) November 18–21, 1997Brussels, Belgium

Course director: G.B. Cadiere


Administrative SecretariatConference Services s.a.Avenue de l’Observatoire, 3 bte 17B-1180 Bruxelles

Tel: (32 2) 375 16 48Fax: (32 2) 375 32 99

Joint Euro Asian Congress of Endoscopic Surgery5th Annual Congress of the European Association forEndoscopic surgery (EAES)3rd Asian-Pacific Congress of Endoscopic SurgeryJune 17–21, 1997Instanbul, Turkey

The Congress will include a joint postgraduate course EAES/SAGES/ELSA on June 17th. For information and registration:

SETUR Congress DepartmentCumhuriyet Cad. No. 10780230 ElmadagIstanbul, TurkeyTel: (90.212) 23003 36Fax: (90.212) 240 82 37

9th International MeetingSociety for Minimally Invasive Therapy

July 14–16, 1997Kyoto, Japan

Scientific program to include: Plenary, Parallel, Poster, and Video sessions.Host Chairman: Professor Osamu Yoshida, Department of Urology, KyotoUniversity, 54 Shogoin Kawahara-sho, Sakyo, Kyoto 606, Japan. Phone: +8175 751-3328, Fax: +81 75 751-3740. This meeting coincides with the GionFestival in Kyoto, one of the greatest festivals in Japan.


Secretariat of SMIT 9th Annual International Meetingc/o Academic Conference Planning383 Murakami-choFushimika, Kyoto 612 JapanTel: +81 75 611-2008Fax: +81 75 603-3816

6th World Congress of Endoscopy Surgery ‘‘Roma 98’’6th International Congress of European Associationfor Endoscopic Surgery

June 3–6, 1998Rome, Italy

The program will include: the latest, original high quality research; sym-posia; plenary lectures; abstract presentations (video, oral, and posters);EAES and SAGES postgraduate courses, OMED postgraduate course ontherapeutic endoscopy; working team reports; educational center and learn-ing corner; meeting of the International Society of Nurses and Associates;original and non original scientific reports; and a world expo of newtechnology in surgery.


Congress Secretariat: Studio EGAViale Tiziano, 1900196 Rome, ItalyTel: +39 6 322-1806Fax: +39 6 324-0143

223

Flexible sigmoidoscopy

A reliable determinant of colonic ischemia following ruptured abdominal aortic aneurysm

C. P. Brandt, J. J. Piotrowski, J. J. Alexander

Department of Surgery, MetroHealth Medical Center, Case Western Reserve University School of Medicine, 2500 MetroHealth Drive,Cleveland, OH 44109-1998, USA


AbstractBackground:The development of colonic ischemia follow-ing repair of ruptured abdominal aortic aneurysm (AAA) isassociated with significant morbidity and timely diagnosisis essential. The purpose of this study was to determine theefficacy of endoscopy in the diagnosis of colonic ischemiaand in prediction of need for resection.Methods:Patients who underwent postoperative lower en-doscopy after ruptured AAA from 1986 to 1995 were re-viewed for endoscopic findings, clinical course, and patientoutcome.Results:A total of 80 patients had ruptured AAA during thestudy period, of which 56 survived for longer than 24 hpostoperatively. Flexible lower endoscopy was done in 18patients (32%) on an average of 4.4 days following AAArepair (range 1–16). Indications for initial endoscopy in-cluded early or bloody stools in 12 (67%), hemodynamicinstability or sepsis in eight (44%), and acidosis in four(22%). The extent of the examination was sigmoid or de-scending colon in 13, cecum in four, and transverse colon inone. Endoscopic findings were normal in four patients. Fiveexaminations showed only areas of hemorrhagic mucosa.Absence of full-thickness ischemia was confirmed by clini-cal course or autopsy in these nine patients. Two examina-tions demonstrated full-thickness necrosis which was con-firmed at subsequent laparotomy. In six examinations, isch-emia was noted but judged to be limited to mucosa only.Absence of full-thickness disease was demonstrated by lap-arotomy in three and subsequent course in three. Eight pa-tients (57%) with initial abnormal examinations underwentrepeat endoscopy showing improved interval appearance inseven cases and progression to full-thickness ischemia inone patient.

Conclusions:Flexible sigmoidoscopy reliably predicts full-thickness colonic ischemia following repair of ruptured aor-tic aneurysms. Patients with non-confluent ischemia limitedto the mucosa can be safely followed by serial endoscopicexaminations.

Key words: Ruptured abdominal aortic aneurysm — Isch-emic colitis — Flexible sigmoidoscopy

The development of colonic ischemia following repair ofruptured abdominal aortic aneurysm (AAA) continues to bea significant cause of morbidity and mortality. While theincidence of clinically significant colonic ischemia follow-ing elective aortic reconstruction is low, the reported occur-rence of this complication after ruptured AAA repair rangesfrom 6.9% to 35% and has been associated with a mortalityrate exceeding 60% [2, 6–9]. Prompt diagnosis of postop-erative intestinal ischemia is critical in order to minimizesubsequent morbidity, but clinical signs are frequently vari-able and nonspecific. Previous authors have recommendedthe use of routine postoperative lower endoscopy in thesepatients [1, 5, 6]. However, data on the reliability of thismethod is scant. The purpose of this study was to determinethe safety and reliability of endoscopy in the diagnosis ofcolonic ischemia and in the prediction of need for intestinalresection.


A retrospective review was performed of all patients treated for rupturedAAA at MetroHealth Medical Center from 1986 to 1995. Patients who didnot survive longer than 24 h after repair were excluded from further analy-sis. Data collected included patient demographics, operative details, andpostoperative complications. All patients who underwent postoperativelower endoscopy were reviewed for timing and indications for the proce-dure, endoscopic findings, length of colon examined, subsequent clinicalcourse, and patient outcome. All initial endoscopies were performed at the

Presented at the annual meeting of the Society of American Gastrointes-tinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA,March 13–17 1996

Correspondence to:C. P. Brandt

Surg Endosc (1997) 11: 113–115

SurgicalEndoscopy


bedside in the intensive care unit. Irrigation was used as needed to clear themucosa for adequate visualization and mucosal biopsy was performed asnecessary to detect presence or absence of submucosal blood flow.

Results

A total of 80 patients were treated for ruptured AAA duringthe study period, of which 56 survived for longer than 24 hpostoperatively. There were 51 males and five females witha mean age of 71 years, ranging from 51 to 91 years. Nine-teen patients (34%) underwent lower endoscopy from 1 to16 days following the initial surgery (mean4 4.4 days).This consisted of one rigid proctoscopy and 18 flexibleendoscopic examinations. Indications for the initial endos-copy were early or bloody bowel movements in 12 (63%),hemodynamic instability or signs of sepsis in nine (47%),systemic acidosis in five (26%), and prolonged ileus in onepatient (5%). The rigid endoscopy included only the rectumand was felt to be technically inadequate because of poormucosal visualization due to fecal material. The extent ofexamination in the flexible studies included the sigmoid ordescending colon in 13, transverse colon in one, and cecumin four.

Table 1 outlines the endoscopic findings and clinicaloutcome of these 18 patients. The endoscopic findings werejudged to be normal in four patients, of which three died andone survived. The absence of full-thickness ischemia wasconfirmed by autopsy in two of the nonsurvivors. The re-maining death occurred at postoperative day 66 withoutmanifest evidence of intestinal ischemia.

Endoscopy showed hemorrhagic mucosa in five pa-tients, of which three survived and two died. One survivorunderwent subsequent laparotomy which did not reveal full-thickness ischemia. Autopsy in one patient confirmed ab-sence of ischemia and the other death was of respiratoryetiology. This patient had undergone a follow-up endoscopywhich had been normal.

Endoscopy in six patients was interpreted as showingischemia which was patchy in distribution and limited to themucosa. Four patients survived and two died. Laparotomywas performed in three patients where absence of full-thickness ischemia was noted, and no resection was per-formed. All of these three patients survived. One patientwho died on the 20th postoperative day had undergone afollow-up endoscopy which showed interval mucosal heal-ing. The other nonsurvivor died 4 months postoperativelywithout evidence of intestinal ischemia.

Two endoscopic examinations were felt to show full-thickness necrosis. In both cases, this was confirmed atsubsequent laparotomy and colonic resection was per-

formed. One initial endoscopy showed diffuse abnormalmucosal exudate which was of unclear significance to theendoscopist. Follow-up examination 48 h later demon-strated confluent ischemia requiring operative resection.Two of the three patients with full-thickness necrosis died.

Eight of 14 patients with initial abnormal examinationsunderwent repeat endoscopy an average of 4.6 days (range4 1–14 days) following initial examination and improvedinterval appearance was noted in seven. There were no com-plications attributable to the endoscopic procedure itself.

Discussion

Colonic ischemia is a well-recognized and often lethal com-plication of ruptured abdominal aortic aneurysm. Patientsdescribed in four recent large series who survived the initialoperation for repair showed an incidence of clinically sig-nificant colonic ischemia ranging from 15 to 35%, withassociated mortality rates of 42 to 68% [2, 6–8]. It is rea-sonable to assume that, as with any intraabdominal sourceof sepsis in a critically ill patient, an early and promptdiagnosis would be required to minimize associated mor-bidity. An accurate assessment of the degree of ischemia isalso important, as some patients with infarction limited tothe mucosa will recover without need for operative resec-tion, while others with full-thickness necrosis will requireprompt exploration and resection.

Early recognition of postoperative intestinal ischemia ishampered in part by difficulty in predicting which patientsare at risk, as well as by variability in clinical signs andsymptoms manifested in patients with this complication.The pathogenesis of intestinal ischemia following rupturedAAA appears to be multifactorial, and attempts at determin-ing pre- and postoperative predictive risk factors have beenmade. Piotrowski et al. [8] found that colon ischemia oc-curred more frequently in patients with preoperative shockand greater intraoperative blood loss, but no correlation wasfound with patient age, comorbid medical conditions, timeto operation, or the patency or treatment of the inferiormesenteric artery. Similarly, Meissner and Johansen [7]found no demographic, clinical, or operative factors thatcorrelated with the development of ischemia other than lowperioperative cardiac output. Unfortunately, no specific fac-tors have been found to be sufficiently reliable to predictthis particular complication in an individual patient.

Further difficulty in establishing an early diagnosis ofintestinal ischemia arises from the variability and nonspec-ificity of the clinical presentation in the presence of theusual hemodynamic and abdominal changes expected after

Table 1. Comparison of endoscopic and pathologic findings following flexible lower endoscopy

Endoscopic impression nSubsequentlaparotomy (n)

Full-thicknessischemia present (n)

Normal 4 0 0Hemorrhagic mucosa only 5 1 0Nonconfluent ischemia limited to mucosa 6 3 0Full-thickness colonic ischemia 2 2 2Confluent mucosal exudatea 1 1 1

a 4 Follow-up sigmoidoscopy displaying full-thickness ischemia

114

emergent aortic surgery. Meissner and Johansen [7] foundthat several clinical factors, including early (less than 24 h)bowel movement, postoperative bloody bowel movement,persistent acidosis, coagulopathy, excessive fluid require-ments, and persistent hemodynamic instability were morecommon among patients with ischemia. However, the sen-sitivity of any single variable was less than 50%. Maupin etal. [6] described 20 patients who had bloody or explosivediarrhea within 4–8 h after surgery, of which 13 (65%) hadischemic colitis. Four of these patients had only mild isch-emia not requiring surgery. Six additional patients who de-veloped ischemic colitis later in the postoperative course didnot have diarrhea but presented with sepsis, acidosis, hy-perglycemia, renal failure, or abdominal tenderness. There-fore, while the presence of these perioperative findingsshould heighten the suspicion for intestinal ischemia, theyare generally not sufficiently specific to determine the needfor reexploration.

Based on these considerations, alternative techniquesfor diagnosis of intestinal ischemia following ruptured AAAhave been proposed. Fiddian-Green et al. [3] studied the useof an intracolonic silicone tonometer to measure intralumi-nal PCO2 as an indirect measure of intramural pH in 25high-risk patients undergoing aortic surgery. They foundthat a low normal pH was predictive of ischemic colitis andthat the duration of pH evidence for ischemia on the day ofoperation was the best predictor for the development ofsigns and symptoms of ischemic colitis and for death afteroperation. However, this technique is somewhat cumber-some, has not been subsequently studied, and was not pre-dictive of which patients would require colonic resection.

Forde et al. [4] in 1979 reported on the use of flexiblecolonoscopy to establish the diagnosis of ischemic colitisand found that the endoscopic appearance correlated wellwith subsequent microscopic pathology. Since then, severalauthors have recommended routine postoperative lower en-doscopy for all patients who initially survive repair of rup-tured AAA, although a positive effect on ultimate survivalwith this approach has not been demonstrated [1, 5, 6].Although postoperative ischemia may involve the right andtransverse colon as well as small intestine in up to 20% ofcases, involvement of the rectum or sigmoid or descendingcolon is almost invariably present [5, 7]. Therefore, flexiblesigmoidoscopy would appear to be the most appropriatetechnique for diagnosis and monitoring of intestinal isch-emia in these situations. Advantages of flexible sigmoidos-copy include ability to perform it at the bedside in the ICU,relative ease and speed and performance, and enhanced mu-cosal visualization and extent of examination when com-pared to rigid proctosigmoidoscopy.

In our experience, flexible sigmoidoscopy was used ona selective basis when there was clinical suspicion of co-lonic ischemia, and abnormal findings were noted in 14 of18 cases. All patients with normal examinations or those

showing only hemorrhagic mucosa were confirmed to haveabsence of significant ischemia by either subsequent clinicalcourse or autopsy. Three patients were felt to have clearfull-thickness ischemia on initial or follow-up sigmoidos-copy which was confirmed at immediate laparotomy whereresection was performed. The most difficult managementdecisions occur in those patients in whom sigmoidoscopyreveals nonconfluent ischemia which appears to be limitedonly to the mucosa. While retrospective analysis revealedthat the endoscopic impression of absence of full-thicknessinfarction was correct in all cases, three of six patients inthis series underwent subsequent laparotomy due to highclinical concern. While we feel that a liberal approach toreexploration cannot be condemned in this situation, thispatient population may often be best managed by close ob-servation, optimization of hemodynamic status and the lib-eral use of follow-up endoscopy in 24 hours to monitor forprogression of ischemic changes.

In conclusion, we have found that flexible sigmoidos-copy in patients who have survived repair of ruptured AAAcan safely and reliably predict the presence of full-thicknesscolonic ischemia requiring resection. Patients with noncon-fluent ischemia limited to the mucosa may be safely fol-lowed by serial endoscopic examinations. Due to the highincidence of ischemic colitis following ruptured AAA andsignificant associated morbidity and mortality, a low thresh-old for or even routine use of postoperative flexible sig-moidoscopy is warranted.

References

1. Brewster DC, Franklin DP, Cambria RP, Darling RC, Moncure AC,Lamuraglia GM, Stone WM, Abbott WM (1991) Intestinal ischemiacomplicating abdominal aortic surgery. Surgery 109: 447–454

2. Farooq MM, Freischlag JA, Seabrook GR, Moon MR, Aprahamian C,Towne JB (1996) Effect of the duration of symptoms, transport time,and length of emergency room stay on morbidity and mortality inpatients with ruptured abdominal aortic aneurysms. Surgery 119: 9–14

3. Fiddian-Green RG, Amelin PM, Herrmann JB, Arous E, Cutler BS,Schiedler M, Wheeler HB, Baker S (1986) Prediction of the develop-ment of sigmoid ischemia on the day of aortic operations. Arch Surg121: 654–660

4. Forde KA, Lebwohl O, Wolff M, Voorhees AB (1979) Reversible isch-emic colitis—correlation of colonoscopic and pathologic changes. AmJ Gastroenterol 72: 182–185

5. Hagihara PF, Ernst CB, Griffen WO (1979) Incidence of ischemiccolitis following abdominal aortic reconstruction. SG&O 149: 571–573

6. Maupin GE, Rimar SD, Villalba M (1989) Ischemic colitis followingabdominal aortic reconstruction for ruptured aneurysm. Am Surg 55:378–380

7. Meissner MH, Johansen KH (1992) Colon infarction after rupturedabdominal aortic aneurysm. Arch Surg 126: 979–985

8. Piotrowski JJ, Ripepi AJ, Alexander JJ, Yuhas J, Brandt CP (1996)Colonic ischemia: the Achilles Heel of ruptured aortic aneurysm repair.Am Surg 62: 557–561

9. Welling RE, Roedersheimer R, Arbaugh JJ, Cranley JJ (1985) Ischemiccolitis following repair of ruptured abdominal aortic aneurysm. ArchSurg 120: 1368–1370

115

Laparoscopic resection of high rectovaginal fistula with intracorporealcolorectal anastomosis and omentoplasty

W. Schwenk, B. Bohm, K. Gru ndel, J. Muller

Department of General Surgery, Humboldt-University, Charite´, Schumannstr. 20/21, 10117 Berlin, Germany

Received: 4 December 1995/Accepted: 23 March 1996

Abstract. A 46-year-old Caucasian female underwentvaginal hysterectomy for myoma in another hospital anddeveloped a high rectovaginal fistula 6 weeks later. A di-verting-loop colostomy of the sigmoid colon was performed2 months later. The patient was admitted to our service withpersistent high rectovaginal fistula 6 months later. We re-sected the sigmoid colon and two-thirds of the rectum in-cluding the fistula tract using laparoscopic techniques. Anintracorporeal anastomosis was accomplished using adouble-stapling technique. An omental flap was mobilizedand placed between the colorectal anastomosis and the va-gina. Except for a subcutaneous wound infection at theformer colostomy site, the postoperative course was un-eventful. The patient was discharged at the 7th postopera-tive day and remained free of symptoms. We conclude thatlaparoscopic resection of high rectovaginal fistula with pri-mary intracorporeal anastomosis is feasible and should beconsidered in selected cases as an alternative ‘‘minimal-invasive’’ approach to this disease.

Key words: Omentoplasty — High rectovaginal fistula —Intracorporeal colorectal anastomosis

Rectovaginal fistulas (RVFs) are rare and account for lessthan 5% of all anorectal fistulas [1]. RVFs can be classifiedaccording to their cause [2], which may influence the choiceof surgical repair: (1) RVF with or without anal sphincterdisruption, (2) inflammatory bowel disease, (3) radiationinjury, and (4) postoperative fistula. While obstetric traumais the most common cause for RVF type I [3], about 10% ofall RVFs are caused by inflammatory bowel disease (type 2)[4]. RVFs following pelvic radiotherapy (type 3) have ac-counted for about 15–20% of all RVFs [5] and RVF type 4following previous surgical procedures constitutes onlyabout 7% of all higher RVFs [3]. RVFs following surgery

usually require temporary diversion followed by resectionwith colorectal anastomosis [2]. The interposition of anomental flap between vagina and colorectal anastomosismay reduce the chances of recurrence. We combined thisapproach with a ‘‘minimally invasive access’’ to the peri-toneal cavity and report the first case of a laparoscopicanterior resection with simultaneous takedown of a loopcolostomy and interposition of an omental flap.

Medical history and preoperative findings

A 46-year-old Caucasian female was admitted to our service in July 1995with a 9-month history of RVF. She had had regular bowel movementswithout blood, pus, or mucous discharge and had undergone vaginal hys-terectomy for myoma of the uterus with an uneventful postoperative coursein another hospital in September 1994. She noticed vaginal discharge andpassed air through the vagina 6 weeks following hysterectomy. An RVFwas diagnosed and conservative treatment with antibiotics was started.Since the symptoms did not improve, a diverting-loop colostomy of thesigmoid colon was performed through a left-sided transrectal laparotomy inthe same hospital in January 1995. The patient complained about persistentvaginal discharge and developed recurrent episodes of colitis which weretreated with antibiotics. A colonoscopy revealed an opening in the anteriorrectosigmoid wall 12 cm above the anal verge in July 1995. The patientwas admitted to our service for definitive treatment of the high RVF.

At admission, the obese patient (165 cm, 77 kg) presented with awell-functioning sigmoid-loop colostomy and otherwise unremarkablefindings at physical examination. Rectal temperature was 36.4°C and bloodtests were within normal range. An edematous region was noticed at theanterior wall about 12 cm from the anal verge at rigid proctoscopy but anopening of the fistula was not visualized. An enema with water-solublecontrast showed a direct communication of the rectosigmoid and the vagi-nal roof (Fig. 1). Although the intravenous pyelogram was unremarkable,a ureteral stent was placed into the left ureter to allow easier intraoperativeidentification of the left ureter. The bowel was prepared with 4 l of poly-ethylenglycol solution and repeated enemas to clean out the colorectumdistal to the colostomy. Perioperative antibiotics consisted of 2 g cefotiamand 0.5 g metronidazole at induction of anesthesia with repeated dosesevery 2 h during surgery.

Operative technique

The patient was placed in a modified lithotomy position. The colostomywas resected using linear staplers and both ends of the bowel were placedCorrespondence to:W. Schwenk

Surg Endosc (1997) 11: 147–149

SurgicalEndoscopy


inside the peritoneal cavity. A 12 mmHg CO2 pneumoperitoneum wasestablished with a high-flow insufflator and four additional cannulas (two5 mm, one 10 mm, one 12 mm) were inserted under visual control (Fig. 2).The laparoscopic dissection was started with lateral mobilization of thesigmoid and descending colon and was continued until the ureter and theovarian vessels were identified. The mesocolon was medially dissectedfrom the retroperitoneum, preserving the hypogastric nerves and identify-ing the left ureter and the ovarian vessels again. The inferior mesentericartery was divided with an endoscopic stapler and the mesocolon wastransected by electrocoagulation. The left colonic flexure was taken downand the omentum majus was dissected free of the transverse colon. Thegreater omentum was dissected along the epiploic vessels and the majorvessels of the omentum so that a long flap on the left side could be created.The plane between the mesorectum and the presacral fascia was entered atthe level of the promontory by sharp dissection and the rectum was mo-bilized in the presacral space down to the tip of the coccyx. The rectum wascarefully mobilized laterally and all adhesions between ovary, tube, andrectum were lysed. The lateral stalks were divided and the vagina wastransvaginally lifted ventrally to facilitate the dissection the rectovaginalseptum 5 cm below the fistula tract. The mesorectum and the rectum weretransected below the fistula with an endoscopic stapler. The specimen wasplaced into an endoscopic bag and removed through the incision at theformer colostomy site. The anvil of a 28-mm circular stapler was insertedinto the descending colon and placed back into the peritoneal cavity. Thewound was irrigated and closed. The pneumoperitoneum was reestablishedand the colorectal anastomosis was accomplished in ‘‘double-stapling’’technique. The donuts were checked for completeness and the anastomosiswas tested by air insufflation into the rectum. The omental flap was placedbetween the colorectal anastomosis and the vagina and fixated with fourstaples to the dorsal side of the vagina. The vagina was left open. Drainswere inserted into the pelvis through the right lower abdominal incisionand into the vagina. The operative time was 500 min with an estimatedblood loss of less than 500 ml.

Postoperative course

The drains were removed on the 1st p.o.d. The patient had her first bowelmovement on day 4. Oral feeding was started with liquids on day 2 and

advanced to normal oral diet within 4 days. Serous secretion from theformer colostomy site was noticed on the 6th p.o.d. The wound was openedsubcutaneously and treated with regular dressings. The patient was dis-charged on the 7th p.o.d. No further passing of gas or stool via the vaginawas noticed during the follow-up of 4 months.

Discussion

Lower RVFs with or without sphincter disruption caused byobstetric injury (type 1) [2], inflammatory bowel disease(type 2), or pelvic radiation (type 3) are treated by transanalor transvaginal procedures. While the transabdominal ap-proach to repair high postoperative RVF (type 4) was rarelyrecommended in the past [6], it has become more popular asthe introduction of circular stapling devices has allowedeasier performance of low colorectal anastomosis in thepelvis [1]. Currently, postoperative high RVFs are repairedby proctosigmoidectomy with colorectal anastomosis [2],which may be combined with the interposition of a vascu-larized pedicle of omentum between the rectum and vagina[1]. While high RVFs due to inflammatory bowel disease orradiation injury are usually accompanied by extensive in-flammation of the surrounding tissue, the patient presentedabove did have an RVF following transvaginal hysterec-tomy. Since there only was inflammation close to the fis-tula, we were able to dissect the rectum from the vagina wellbelow the level of the RVF so that an intraperitonealdouble-stapling anastomosis could be accomplished in themidrectum. The operation required much more time (500min) than a regular laparoscopic anterior resection (200–300 min): This was due in part to the additional steps (co-lostomy take-down and formation of a vascularized omentalflap) and in part to meticulous dissection of the dense ad-hesions in the pelvis. Despite the long operative time, therecovery was very rapid. The superficial wound infection atthe former colostomy side does not relate to the minimallyinvasive access but is a well-known complication of colos-tomy take-down procedures. The presented case demon-

Fig. 2. Position of the cannula during laparoscopic resection of rectovagi-nal fistula: (A) 10-mm cannula, (C) 12-mm cannula, (B and E) 5-mmcannula, (D) former colostomy site, 12-mm cannula and removal of thespecimen.

Fig. 1. Barium enema showing the fistula between the upper rectum andthe vaginal stump following hysterectomy.

148

strates that even complex colorectal procedures to cure highRVF may be performed laparoscopically. However, the se-lection of the patient is as important as the experience of thelaparoscopic surgeon, who has to be very familiar with thelaparoscopic approach to pelvic diseases.

References

1. Bandy LC, Addison A, Parker RT (1983) Surgical management ofrectovaginal fistulas in Crohn’s disease. Am J Obstet Gynecol 147:359–633

2. Given FT (1970) Rectovaginal fistula: a review of 20 years’ experiencein a community hospital. Am J Obstet Gynecol 108: 41

3. Lescher TC, Pratt JH (1967) Vaginal repair of the simple rectovaginalfistula. Surg Gynecol Obstet 124: 1317–1321

4. Lowry AC, Thorson AG, Rothenberger DA, Goldberg SM (1988) Re-pair of simple rectovaginal fistulas: influence of previous repairs. DisColon Rectum 31: 676–678

5. Rothenberger DA, Goldberg SM (1983) The management of rectovagi-nal fistula. In: Faszio VW (ed) The surgical clinics of North America.WB Saunders, Philadelphia, p 61–79

6. Senagore A (1995) Rectovaginal fistula. In: Mazier PW, LuchtefeldMA, Levien DH, Seagore AJ (eds) Surgery of the colon, rectum andanus, 1XX ed. WB Sanders, Philadelphia

149

Comparison of acid reduction in antiulcer operations

F. J. Brody,* K. S. Trad

Department of Surgery, The George Washington University, 2150 Pennsylvania Avenue, NW, Washington, DC 20037, USA


AbstractBackground:Three acid-reducing operations have recentlybeen described for the laparoscopic treatment of peptic ulcerdisease. These consist of a posterior truncal vagotomy com-bined with either (1) an anterior seromyotomy (SERO), (2)an anterior highly selective vagotomy (AHSV), or (3) alinear stapled lesser curvature excision (STAP). The pur-pose of this study was to investigate the physiologic effectsof these procedures in terms of basal and maximal acidoutputs.Methods: Fifty New Zealand rabbits were prospectivelyrandomized into five open laparotomy groups (n 4 10): acontrol group without vagal manipulation (CON), a bilateraltruncal vagotomy with pyloromyotomy group (VP), aSERO group, an AHSV group, and a STAP group. Allanimals underwent placement of a gastrostomy tube for sub-sequent gastric secretory analysis. On postoperative day 6,basal acid outputs (BAO) and maximal acid outputs (MAO)following IV pentagastrin stimulation (30mg/kg/h) weremeasured. Results were compared statistically using theANOVA method.Results:Pentagastrin stimulation was associated with a sig-nificant increase in MAO in the CON group (p < 0.05 vsBAO); however, this response was effectively blunted in allthe experimental groups. There were no differences in BAOor MAO between any of the vagotomized groups (SERO,HSV, STAP, VP).Conclusions:We conclude that the three acid-reducing pro-cedures modified for laparoscopy are equally efficacious inreducing gastric acid secretion and that they compare fa-vorably with VP. To our knowledge, this is the first reportcomparing basal and stimulated gastric acid secretion be-tween these new acid-reducing techniques.

Key words: Acid reduction — Antiulcer operations — Pep-tic ulcer disease

Peptic ulcer disease (PUD) remains a major health careproblem in the United States. Approximately 10% of thegeneral population develop some manifestation of ulcer dis-ease during their lifetime. The introduction of H2-blockersin the mid-1970s and of proton pump inhibitors more re-cently has markedly decreased the volume of acid-reducingprocedures performed annually. Our increased understand-ing of the role of Helicobacter pylori in PUD and thechanges in medical management may effect the ultimaterole of surgical therapy. However, surgery remains an im-portant modality in the treatment of intractable, recurrent,and complicated peptic ulcer disease. Today laparoscopicprocedures are offered as an alternative to traditional opentechniques in selected cases when an operative interventionis warranted.

Several laparoscopic procedures for the management ofPUD are currently being taught and performed across Eu-rope and the United States. In 1991, Francois Dubois pub-lished his data on laparoscopic bilateral truncal vagotomy[2]. This technique requires, however, the addition of agastric drainage procedure (pyloromyotomy, pyloroplasty,or endoscopic balloon dilatation of the pylorus) and is fre-quently associated with postoperative diarrhea, dumping,and other postvagotomy syndromes. Unlike bilateral truncalvagotomy, highly selective vagotomy (HSV) preserves an-tral innervation, does not require a drainage procedure, andcauses little or no postvagotomy side effects [5–7]. Withwell-documented physiologic and clinical outcomes, lapa-roscopic HSV is feasible, but the technical difficulty in-volved in dissecting the posterior leaflet of the gastrohepaticomentum and operative times in excess of 5 h have limitedits widespread use [1].

To date, three major laparoscopic modifications of theclassic HSV are available to alleviate the technical difficul-ties. All three consist of a posterior truncal vagotomy (PTV)combined with some form of anterior selective denervation:

1. PTV with anterior seromyotomy: The seromyotomy isperformed along the lesser curvature of the stomach andinterrupts the small vagal fibers which course obliquelythrough the seromuscular layer before reaching the acid-secreting mucosa. This operation was first conceived and

Presented at the annual meeting of the Society of American Gastrointes-tinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA,13–17 March 1996* SAGES Grant RecipientCorrespondence to:F. J. Brody

Surg Endosc (1997) 11: 123–125

SurgicalEndoscopy


popularized in the United Kingdom by Taylor [9]. Kat-khouda and Mouiel from the University of Nice, France,adapted this operation for laparoscopic surgery and re-ported their early results in a clinical series [8].

2. PTV with anterior highly selective vagotomy: Selectiveanterior denervation is achieved by individual ligation ofthe neurovascular bundles of the anterior leaf on thelesser curvature of the stomach. This technique wasoriginally descirbed by Hill and Barker in 1978 [4].Zucker and his co-workers developed a similar operationusing the laparoscopic approach [1].

3. PTV with stapled lesser curvature: A full-thickness stripof gastric wall is excised along the lesser curvature witha linear stapling device, thereby interrupting the vagalfibers traveling to the gastric mucosa. This original ap-proach was described by Hannon and Snow [9].

The underlying assumption is that all three modifica-tions are physiologically equivalent to classic open acid-reducing techniques. While this may be argued on a theo-retical basis, little factual evidence exists in the literature tosupport this claim. The aim of this research was to test thevalidity of this assumption by investigating the effect ofthese procedures on basal and stimulate acid secretion. Weused a chronic animal model, developed in our laboratory,to compare the reduction in acid secretion observed in eachof these three modifications against the more clinicallyproven and physiologically accepted bilateral truncal vagot-omy with pyloromyotomy. This study was supported in fullby a 1995 SAGES research grant.

Methods

Fifty New Zealand rabbits (mean weight 3.0-kg) were prospectively ran-domized into five groups (n 4 10); a control group without vagal ma-nipulation (CON), a bilateral turncal vagotomy with pyloromyotomy group(VP), a PTV with anterior seromyotomy group (SERO), a PTV with an-terior highly selective vagotomy group (AHSV), and a PTV with linearstapled lesser curvature excision group (STAP). Animals were fed a pre-operative liquid rabbit chow diet (Liquid diet, Bioserv, Inc.; Frenchtown,NJ) for 1 week and fasted for approximately 24 h prior to surgery. Ad-equate anesthesia was achieved with ketamine (100 mg/kg) and xylazine(50 mg/kg) administered intramuscularly followed by 0.5% lidocaine in-jected subcutaneously along the incision site. A midline laparotomy wasthen performed. Rabbits randomized to the VP group underwent a bilateraltruncal vagotomy with a pyloromyotomy extending longitudinally from thefirst portion of the duodenum through the entire seromuscular layer of thepylorus. Animals randomized to one of the selective anterior vagal dener-vation groups (SERO, AHSV, or STAP) underwent a posterior truncalvagotomy prior to selective denervation. The seromyotomy was completedutilizing the electrocautery as described by Taylor [9]. Our AHSV wasperformed with 4-0 chromic ties and the STAP group underwent a lessercurvature resection with an endoscopic linear stapler (Ethicon, Inc., Somer-ville, NJ) as described by Hannon [3]. All procedures were completed in an

open fashion utilizing laparoscopic instruments as described by the se-lected authors. All animals subsequently received a modified gastrostomytube (BARD Interventional Products; Tewksbury, MA) placed in the mostdependent and inferior aspect of the gastric body as access for gastricsecretory analysis. The gastrostomy tube was brought through a stabwound in the abdominal musculature and placed in a subcutaneous pocket.Animals from the CON group underwent gastrostomy tube placement onlywithout any vagal manipulation. The abdomen was closed and the animalsrecovered. Postoperatively, all animals were allowed a liquid diet. Onpostoperative day 5, the subcutaneous portion of the gastrostomy wasexteriorized and the stomach was lavaged until clear via the gastric cannulaand the rabbits were fasted overnight. On postoperative day 6, the animalswere placed in a restraining device and gastric lavage was performed againuntil clear and gastric analysis was performed using the standard acidtitration method. Basal acid outputs (BAO) were collected via the cannulaat 15-min intervals for 1 h. After each 15-min interval, the stomach waslavaged with three 30-ml boluses of distilled water. The volume collectedwas measured to the nearest milliliter and a 5-m aliquot was collected fromeach 30-ml lavage. This 15-ml aliquot was titrated to pH 7.0 with 0.1N

NaOH. The animals were then rested for 1 h. Pentagastrin was then ad-ministered intravenously 30mg/kg/h). Fifteen minutes after pentagastrininjection, maximal acid outputs (MAO) were collected at 15-min intervalsand titrated to pH 7.0 using the same method. BAO and MAO are ex-pressed inmmol/min of hydrochloric acid secretion. Data are expressed asmean ± SD and compared using the ANOVA method. The conduct ofexperiments and care of animals conforms to theGuide for the Care andUse of Laboratory Animals,NIH publication No. 86-23.

Results

Validation of the animal model

In our initial pilot study, we validated the chronic consciousrabbit model developed in our laboratory by comparing theMAO and BAO in rabbits from the VP group and sham-operated rabbits. This revealed a remarkable (2–3 fold) andreproducible decrease in pentagastrin-stimulated MAO inthe VP group as compared to the sham-operated group.There were no marked difference in the BAO between thetwo groups.

Comparison of acid reduction

Our results are summarized in Table 1. Pentagastrin stimu-lation was associated with a significant increase (2.5-fold)in MAO as compared to BAO in the control (CON) group(p < 0.05 vs BAO). This response was effectively blunted inall the vagotomized groups (VP, SERO, AHSV, STAP):The mean MAO in these groups never exceeded a 50%increase as compared to the basal values, and the differencedid not reach statistical significance. There were no differ-ences in BAO or MAO between any of the vagotomized

Table 1.Basal acid outputs (BAO) and pentagastrin-stimulated maximal acid outputs (MAO) expressedin mmol/min ± SD

VP SERO AHSV STAP CON

n 10 10 10 10 10BAO 34.2 ± 8.6 32.2 ± 9.0 27.7 ± 12.1 23.9 ± 9.1 41.4 ± 16.0MAO 48.7 ± 6.9* 47.9 ± 7.4* 47.0 ± 12.2* 35.9 ± 14.8* 107.2 ± 20.1**

* p < 0.05 vs CON** p < 0.05 vs BAO

124

groups (VP, SERO, AHSV, STAP). However, MAO wassignificantly lower in each of these groups as compared tothe control group (p < 0.05 vs CON).

Discussion

Since the successful introduction of laparoscopic cholecys-tectomy in 1987, there has been a widespread thrust bysurgeons across the world to extend the applications ofminimally invasive surgery to the treatment of other intra-abdominal disease entities. The laparoscopic managementof peptic ulcer disease is now routinely practiced in severalcenters to deal with complicated and intractable ulcer dis-ease where medical therapy has failed. In some selectedcases, it is even regarded as a reasonable alternative tolife-long medical therapy. As a result of this enthusiasm forminimally invasive approaches, new procedures with absentor limited track records as open procedures have been in-troduced with the specific goal of being made technicallyfeasible through the laparoscope. Long-term ulcer recur-rence rates will not be available for a number of years, andthere are no ongoing randomized trials to compare thephysiologic effects and long-term clinical effects of thenewly available laparoscopic acid-reducing procedures.

To our knowledge, this is the first report comparingbasal and stimulated gastric acid secretion between a well-established, clinically proven antiulcer operation (VP) andthe newer acid-reducing techniques. In our randomized pro-spective animal study, we showed that the three modifiedapproaches were equally efficacious in reducing gastric acidsecretion, suggesting that they were equivalent from aphysiological standpoint. The choice of the preferred lapa-roscopic antiulcer procedure should therefore depend on thesurgeon’s familiarity with a specific technique. Further-more, they were each found to compare favorably with VPin terms of acid reduction. This suggests that the three an-

tiulcer procedures are physiologically sound and may prove,in the long-term, to be associated with acceptable ulcerrecurrence rates. Final determination regarding this matterwill have to await the 5- and 10-year follow-up studies fromthe ongoing series.

Although our model did not incorporate a laparoscopicapproach, we were investigating only the physiologic end-points for each procedure regardless of the surgical ap-proach. Future protocols utilizing a laparoscopic approachcomparing these acid-reducing procedures are currentlypending at our institution in order to account for laparo-scopic variables such as the pneumoperitoneum and dissec-tion difficulties utilizing laparoscopic instruments.

References

1. Bailey RW, Flowers JL, Graham SM, Zucker KA (1991) Combinedlaparoscopic cholecystectomy and selective vagotomy. Surg LaparoscEndosc 1: 45–49

2. Dubois F (1991) Surgical endoscopic management of peptic ulcer dis-ease. Programs and abstracts of the Sages 1991 postgraduate course,Monterey

3. Hannon JK, Snow LL, Weinstein LS (1992) Linear gastrectomy: anendoscopic staple assisted anterior highly selective vagotomy combinedwith posterior truncal vagotomy. Surg Laparosc Endosc 2(3): 254–257

4. Hill GL, Barker CJ (1978) Anterior highly selective vagotomy withposterior truncal vagotomy: a simple technique for denervating theparietal cell mass. Br J Surg 65: 702–705

5. Hoffman J, Olesen A, Jensenn HE (1987) Prospective 14–18 year fol-low-up after parietal cell vagotomy. Br J Surg 74:1056–1059

6. Johnston D (1975) Operative mortality and postoperative morbidity ofhighly selective vagotomy. Br J Surg 62: 160–164

7. Jordan PH (1991) Surgery for peptic ulcer disease. Curr Probl Surg 28:265–330

8. Katkhouda N, Mouiel J (1991) A new technique of surgical treatment ofchronic duodenal ulcer without laparotomy by videiocoelioscopy. AmJ Surg 161:361–364

9. Taylor TV, Gunn AA, MacLeod DAD (1982) Anterior lesser curveseromyotomy and posterior truncal vagotomy in the treatment ofchronic duodenal ulcer. Lancet 2: 846–848

125

Original articles

Thoracoscopic decortication in infants and children

S. S. Rothenberg, J. H. T. Chang

Columbia-HealthONE Presbyterian, St. Luke’s Hospital, 1601 East Nineteenth Avenue, Suite 5200, Denver, CO 80218, USA

Received: 25 March 1996/Accepted: 28 May 1996

AbstractBackground:The treatment of pediatric empyemas remainscontroversial. While thoracentesis and tube thoracostomyappear adequate for relatively benign organisms, virulentbacteria cause thick fibrinous pleural peels entrapping thelung. Open thoracotomies have been effectively used fordecortication but are painful.Methods:We report the use of minimally invasive thoraco-scopic decortication in 12 patients (mean age 5 years). Allfailed conventional management with persistent fever, in-creasing oxygen requirement, recurrent effusion, and pleu-ral consolidation; 5- and 10-mm trocars were used and com-plete decortication was accomplished.Results:Ten of 12 patients were afebrile by 72 h and dis-charged 4–12 days after surgery. Eleven of 12 patients hadclear chest x-rays by 1 month.Conclusion:Thoracoscopic decortication is a safe and ef-fective means of treating pediatric empyemas.

Key words: Thoracoscopic decortication — Pediatric em-pyema — Thoracentesis

Although the surgical drainage of empyemas has been ad-vocated from the time of Hippocrates, the treatment contin-ues to be debated [1, 6]. As in most of medicine, the besttherapy is prevention of the empyema by proper antimicro-bial therapy, timely follow-up, and early tube drainage ofpleural effusions [4]. Unfortunately, a significant number ofinfants and children continue to develop multiloculated em-pyemas and suffer prolong hospitalization and convales-cence. It has been shown since the early 1980s that decor-tication shortens hospitalization and convalescence [5, 7, 8].However, a formal thoracotomy is painful and hinders post-operative respiratory excursion. Even though Rogers sug-

gested the use of thoracoscopy as early as 1981, its use forthe definitive treatment of pediatric empyemas is relativelyrecent [5].


From February 1993 to December 1995, 12 patients between the ages of 22months and 14 years (mean: 5 years) and from 13 kg to 80 kg (mean: 28kg) had thoracoscopic decortication of their empyemas. The patients hadbeen hospitalized from 3 to 7 days prior to surgery. In addition to antibi-otics, six received thoracentesis and three had tube thoracostomies. Twopatients showed evidence of necrotizing pneumonia. Eight empyemas werein the left and three in the right chest. Indications for surgery includedpersistent fever, increased oxygen requirement, recurrent or increased ef-fusions, and pleural consolidation.

The patients were placed in a lateral decubitus position. Ten of 12patients had contralateral mainstem intubations. When necessary, carbondioxide at low flow (1 l/min) and low pressure (4 mmHg) was used tomaintain lung collapse. Three 5-mm trocars or two 5-mm and one 10-mmtrocars were inserted and 0° and or 30° scopes were used. Bowel forcepswere most effective for complete decortication and abscess debridement.Two chest tubes were left in all patients. Operative time ranged from 50 to130 min (mean: 86 min).

Results

Cultures revealed six cases ofStreptococcus pneumoniae,three of group A beta-hemolyticStreptococcus, two alphaStreptococcus,and two with no growth.

Ten of 12 patients were afebrile by 72 h. One patientneeded postoperative ventilation for 48 h. The chest tubeswere removed between days 3 and 7. All patients weredischarged within 4 to 12 days (mean: 6.3 days). Six pa-tients were discharged off antibiotics and 6 were given a7–14-day course of oral antibiotics.

One month follow-up showed a clear chest roentgeno-gram in 11 of 12 patients. One developed a parenchymalabscess which was drained percutaneously. One patient de-veloped a deep-vein thrombophlebitis 2 weeks after surgeryand was treated conservatively.

Discussion

Thoracoscopy was used as early as 1910 by Jacobaeus tolyse adhesions and collapse the lung of a patient with tu-


Correspondence to:S. S. Rothenberg

SurgicalEndoscopy


berculosis. Limited visibility and inadequate instrumenta-tion prevented its wide usage. In the late 1980s, the devel-opment of the microchip camera and proper instrumentsresulted in its widespread use in the abdominal cavity. Littleadaptation was needed for its application in the chest cavity.Further miniaturization of instruments and improvement ofoptics have allowed for its use in the pediatric age groupincluding mediastinal tumor and lung resections [11].

Pneumonia continues to be a significant pediatric afflic-tion with an annual incidence of 1–4% [1]. The increasingnumber of penicillin-resistantStreptococcus pneumoniaereports will undoubtedly result in more empyemas due todelay in proper antimicrobial administration. While earlythoracenteses and tube thoracostomies will be curative forexudative and early fibrinopurulent pleural effusions, themultiloculated empyemas require more aggressive therapy.Conservative management of the residual fibrinous empy-ema peel with prolonged antibiotic therapy results in alengthy convalescence with general malaise, a marked re-striction of activities, and occasionally, recurrence with aresistant organism. Thoracic wall deformities have also re-sulted from inadequately treated empyemas [11].

The therapy of empyemas in infants and children con-tinues to be debated due to reports of successful conserva-tive management of relatively antibiotic-susceptible bacte-rial strains such as pneumococci, streptococci, orStrepto-coccus aureus.The emergence ofH. influenza,B-hemolyticstreptococci,S. pneumoniae,and anaerobes have dramati-cally altered the intrapleural pathology. These bacteria com-monly cause multiple fluid loculations and a thick fibrinouspeel which entraps the lung, and they do not respond toconventional antibiotic therapy. Recently, Rosen et al. re-ported the use of streptokinase for the dissolution of suchempyemas. Unfortunately, only one patient had bacterialculture results and the hospitalization ranged from 10 to 26days (average 19.2 days). In addition three of their fivepatients required 2–3 months before having a normal physi-cal examination and activity [10]. Further investigationsneed to be performed.

The use of open thoracotomy for the treatment of pedi-atric empyemas has been advocated for over a decade withproven efficacy [2, 3, 9]. With the advent of minimallyinvasive techniques of thoracoscopy, even the dread of thepain of a thoracotomy has been eliminated. Usually, 5-mmtrocars are all that is necessary and the discomfort is mini-mal; 10-mm trocars may be used in the larger-sized patient

as it obviously facilitates removal of the peel to shortenoperative time. In Kern and Rogers’ report, eight of ninepatients (one died of leukemia) post-thoracoscopy requiredan average of 8.4 days of tube drainage and 13.4 days ofhospitalization. Our patients required only 3–7 days of chesttube drainage and averaged 8.2 days of hospitalization. Thisdifference may be explained by the fact that our patientswere treated by conventional means for 3–7 days and theirsfor 1–17 days.

In conclusion, if clinical signs do not steadily improveafter thoracentesis or tube thoracostomy, there is no ratio-nale for prolonging inadequate therapy, and thoracoscopyshould be performed. Thoracoscopic decortication, in well-trained hands, is a safe, effective, economical, and mini-mally painful treatment for pediatric empyema. It rapidlyimproves the clinical status and shortens both hospitaliza-tion and recovery time.

References

1. Campbell PW (1995) New developments in pediatric pneumonia andempyema. Curr Opin Pediatr 7: 278–282

2. Gustafson RA, Murray GF, Warden HE, Hill RC (1990) Role of lungdecortication in symptomatic empyemas in children. Ann Thorac Surg49: 940–947

3. Hoff SJ, Neblett WW, Edwards KM, Heller RM, Pietsch JB, HolcombGW Jr, Holcomb III GW (1991) Parapneumonic empyema in children:decortication hastens recovery in patients with severe pleural infec-tions. Pediatr Infect Dis J 10: 194–199

4. Kennedy AS, Agness M, Bailey L, White JJ (1991) Decortication forchildhood empyema. The primary provider’s peccadillo. Arch Surg126: 1287–1291

5. Kern JA, Rogers BM (1993) Thoracoscopy in the management ofempyema in children. J Pediatr Surg 9: 1128–1132

6. Kosloske AM, Cartwright KC (1988) The controversial role of decor-tication in the management of pediatric empyema. J Thorac Cardio-vasc Surg 96: 166–170

7. Kosloske AM, Cushing AH, Shuck JM (1980) Early decortication foranaerobic empyema in children. J Pediatr Surg 15: 422–429

8. Mayo P, Saha SP, McElvein RB (1982) Acute empyema in childrentreated by open thoracotomy and decortication. Ann Thorac Surg 34:401–407

9. Ridley PD, Braimbridge MV (1991) Thoracoscopic debridement andpleural irrigation in the management of empyema thoracis. Ann Tho-rac Surg 51: 461–464

10. Rosen H, Nadkarni V, Theroux M, Padman R, Klein J (1993) Intra-pleural streptokinase as adjunctive treatment for persistent empyema inpediatric patients. Chest 103: 1190–1193

11. Rothenberg SS (1994) Thoracoscopy in infants and children. SeminPediatr Surg 3: 277–282

94

Letters to the editor

A technique for laparoscopic cystic duct cholangiography

Cholangiography may be indicated during the course oflaparoscopic cholycystectomy. Many catheters for and tech-niques of performing cystic duct cholangiography havebeen developed. Catheters with balloons have the advantageof preventing leaks during dye injection, but balloon infla-tion may often result in displacement of the balloon eitherinto the common bile duct or out of the biliary tree alto-gether, resulting in unsatisfactory studies. Catheters withoutballoons, on the other hand, are prone to leak contrast dur-ing injection, again resulting in poor X-rays. We have de-veloped a technique for safe, rapid, and accurate cystic ductcholangiography without leakage of contrast.

Following identification and isolation of the cystic duct,Hartmann’s pouch is elevated with a grasper via the mid-subcostal port, exposing the entire length of the cystic duct.A clip is placed as high as possible on the cystic duct, andan anterior incision is made into the duct. The cystic duct isthen cannulated via the lateral subcostal port with an Ameri-cath (American Catheter) cholangiogram catheter, with thefirst hashmark just inside the duct. The operator places theendoclip applier around the cystic duct and begins to closethe clip while the assistant continuously injects saline intothe catheter. The operator continues to close the clip until itis just tight enough to prevent further injection of contrast

by the assistant. The operator releases the clip onto the duct,with the assistant still trying to inject. It has been our ex-perience that the clip will be tight enough to prevent leaksbut loose enough to permit flow of contrast. Cholangiogra-phy may then be performed without incident in the usualway. At its conclusion, the clip and catheter are removed,and the cystic duct may be clipped and divided. Alterna-tively, if the cholangiogram reveals choledocholithiasis, thecommon bile duct may be explored via the cystic duct ordirectly.

In summary, cholangiography is often indicated duringlaparoscopic cholecystectomy. A simple technique for leak-proof cystic duct cholangiography is described. This proce-dure provides an effective means for obtaining high-quality,accurate cholangiograms without leakage of contrast orcatheter displacement.

H. J. WinterC. J. Butler

2301 Evesham Ave., Suite 504Voorhees, NJ 08043USA

SurgicalEndoscopy

© Springer-Verlag New York Inc. 1997Surg Endosc (1997) 11: 163

Case reports

Bleeding esophageal ulcers caused by NSAIDs

C. Sugawa,1 Y. Takekuma,2 C. E. Lucas,1 H. Amamoto1

1 Department of Surgery, Wayne State University School of Medicine, 6-C, University Health Center, 4201 St. Antoine, Detroit, MI 48201, USA2 Showa University School of Medicine, Fujigaoka Hospital, Yokohama, Japan

Received: 5 April 1996/Accepted: 28 May 1996

Abstract. This report describes four patients with NSAID-induced esophageal ulcers documented by endoscopy. Thecause of injury was ibuprofen alone in two patients, aspirinin one patient, and a combination of aspirin and ibuprofen inone patient. The most common findings were anemia,retrosternal pain, and dysphagia. Three patients had bleed-ing esophageal ulcers requiring blood transfusions. One pa-tient had massive bleeding which was controlled by endo-scopic hemostasis. Three patients were followed up by en-doscopy, which showed healing in 3–4 weeks. TheseNSAID-induced ulcers had characteristic endoscopic fea-tures, namely, a large, shallow, discrete ulcer in the mides-ophagus near the aortic arch with normal surrounding mu-cosa. These findings suggest that the injury resulted frommucosal contact with NSAIDs. A precise history and imme-diate endoscopic examination were most important in es-tablishing the diagnosis of esophageal ulcer. Healing occursif drug-induced injury is recognized early and treatment isappropriately started with antacids and H2 blockade. Of-fending medication should be discontinued and patientsshould be counseled to take pills in an upright posture withliberal amounts of fluids well before retiring for the night.

Key words: Bleeding esophageal ulcers — NSAIDs —NSAID-induced esophageal ulceration — Endoscopic he-mostasis

Pemberton [19], in 1970, first described drug-inducedesophageal injury. Since this initial report, the incidence ofthese injuries has progressively risen; antibiotics are themost common injurious agent [9, 13, 14]. Many other medi-

cations, however, can cause esophageal injury. Nonsteroidalantiinflammatory drugs (NSAIDs) can damage every levelof the gastrointestinal tract from the esophagus through thelarge intestine [12]. NSAID-induced esophageal ulcerationmay cause pain, difficulty swallowing, and anemia [3, 13,14]. Overt gastrointestinal bleeding from NSAID-relatedesophageal injury has rarely been reported. This report de-scribes four patients with NSAID-induced esophageal ul-cers which caused significant bleeding, requiring bloodtransfusion in three patients.

Case report

Case 1

This 63-year-old black female presented to the emergency room followinga sudden onset of painless massive hematemesis. The patient described a‘‘sticking’’ sensation and retrosternal discomfort at the xiphoid level aftertaking her morning crushed dose of 800 mg Motrin, which was being takentwice daily for degenerative joint disease. The sensation passed after con-suming some water. Approximately 3 h later, the patient suddenly devel-oped hematemesis. Following resuscitation, endoscopy revealed a long,shallow, confluent esophageal ulcer (7 × 2.5 cm) at 25–32 cm from theincisors; there was active pulsatile bleeding near the edge of the ulcer base(Fig. 1). The bleeding was controlled by endoscopic hemostasis, using aheater probe (Fig. 2). A large hiatal hernia with a small epiphrenic diver-ticulum was also noted. There was no evidence of acid reflux esophagitis.Both the stomach and duodenum were normal.

The patient’s hemoglobin fell to 8.2 g/dl with hydration, for which shereceived two units of packed red blood cells. She was given nothing bymouth and was treated with antacids and H2 blockade. Repeat endoscopy5 days later showed significant healing of the ulcer (Fig. 3). Biopsy speci-mens revealed no evidence of neoplasm, fungi, or viral inclusions. Uppergastrointestinal x-rays confirmed the hiatal herniation with a small epi-phrenic diverticulum, but no lesion or stricture could be detected. Hersymptoms improved in 5 days, and she was discharged on the 6th hospitalday. Follow-up endoscopy 4 weeks later showed a completely healed ulcer.

Case 2

This 41-year-old black female with a history of heavy ethanol intake pre-sented to the emergency room with hematemesis, chest and midepigastricpain, and shortness of breath. She had been taking 8–10 tablets of aspirin


Correspondence to:C. Sugawa

SurgicalEndoscopy


and Anacin daily for 3 weeks to treat her abdominal and chest pain. Shechewed some pills before the hematemesis. She denied peptic ulcer diseaseor prior hematemesis. The initial hemoglobin was 7.8 g/dl.

Endoscopy revealed a large, shallow, confluent ulcer measuring 5 × 5cm, with clots on the base at the midesophagus (25–30 cm from theincisors) (Fig. 4), and a few 1 × 1 cmsatellite ulcers at 3–4 cm distally.Duodenitis and a few gastric erosions were also noted. Biopsy specimensfrom the esophageal ulcer showed no evidence of neoplasm, fungi, or viralinclusions.

The patient’s symptoms improved after treatment with gastric rest,antacids, and H2 blockade. She was discharged on the 6th day. The patientdid not keep her appointment for repeat endoscopy but, based on a tele-phone interview, was asymptomatic 2 weeks following discharge.

Case 3

This 20-year-old male with a 2-month history of back pain presented to theemergency room with abdominal pain of 2 weeks’ duration, anorexia, andweight loss. For 2 days prior to admission, he took six doses of ibuprofen(Motrin) with aspirin for his back pain. He denied hematemesis or hema-tochezia. Physical examination revealed guaiac-positive stools. The bloodcount showed a microcytic anemia (Hb 7.8 g/dl).

Endoscopy revealed a 6 × 6 cmshallow midesophageal (25–30 cm)ulcer oozing blood from the base (Fig. 5). Antral gastritis and mild duo-

denitis were also found. Biopsy specimens from the esophageal ulcershowed no neoplasm. Special viral and fungi stains were negative.

The patient was transfused with two units of packed red cells. Hissymptoms resolved in 3 days after treatment with antacids and H2 block-ade. He went home on the 5th day. Repeat endoscopy 3 weeks later showedcomplete healing.

Case 4

This 18-year-old black female, with no significant past medical history,presented to the emergency room with a 3-day history of midepigastricpain and nausea, which was worse with eating and drinking. She deniedhematemesis and hematochezia. She began taking two Motrin tablets threetimes daily for headache 4 days prior to admission. Her hemoglobin was12.3 g/dl. Endoscopy revealed a shallow 4 × 3 cmesophageal ulcer at themidesophagus (27 cm) (Fig. 6) without hiatal herniation or esophagitis.Biopsy specimens from the lesion showed no neoplasm. Special viral andfungi stains were negative. The patient’s symptoms resolved in 2 days aftertherapy with antacids and H2 blockade. Repeat endoscopy 3 weeks latershowed complete healing.

Discussion

Over 70 medications purportedly cause pill-induced esoph-ageal injury [4]. Medications with the most damaging topi-

Fig. 1. A long, shallow, confluent ulcer (7 × 2.5 cm) is seen at 25–32 cm, with active pulsatile bleeding near the edge of the ulcer base.

Fig. 2. This shows active pulsatile bleeding being controlled by heater probe.

Fig. 3. This shows significant healing of the ulcer 5 days after hemostasis.

Fig. 4. A large, shallow ulcer with clots on the base is seen at the midesophagus.

Fig. 5. A large, shallow ulcer is seen on the midesophagus oozing blood from the base.

Fig. 6. A shallow 4 × 3 cmulcer is seen at the midesophagus.

144

cal effects include Quinaglute, tetracycline, iron prepara-tions, potassium tablets, and NSAIDs, including aspirin [9,13, 15]. Recently NSAIDs have been shown to cause esoph-ageal injury. Although NSAIDS commonly injure the gas-tric mucosa, esophageal injury has been considered a rareadverse reaction to NSAIDS [1, 7, 8, 9, 13, 14, 16, 21].Among 49 reported cases of NSAID-induced esophagealinjury, one-third have been complicated by hemorrhage (16cases) or stricture (4 cases) [13]. The likelihood of hemor-rhage or stricture is highly in the NSAIDs group than inpatients with antibiotic-induced injury, which is the mostcommon cause of esophageal injury [13].

The characteristic features of the 11 documented casesof NSAID-induced confluent esophageal ulcers, seen withendoscopy, are shown in Tables 1 and 2. Six of the 11patients presented with G.I. bleeding. All four cases hereinhad a definite confluent ulcer in the midesophagus. Fivepatients needed blood transfusion (Table 2). The patientwith active pulsatile bleeding from the esophageal ulcerrequired emergency endoscopy with endoscopic hemostasisby heater probe. Endoscopic hemostasis for massive bleed-ing has never been described before in patients withNSAID-induced esophageal injury.

The typical clinical features of NSAID-induced esoph-ageal injury include retrosternal pain, odynophagia, anddysphagia [3, 9, 13, 14, 16, 21]. Hematemesis or melenamay occur without chest pain. A sticking sensation afteringesting a pill or even crushed medication, as was noted inone patient described herein, occurs when inadequate peri-stalsis impedes passage of the medication into the stomach.The mean duration from starting medication until the onsetof symptoms is reported as 10.2 days [18]; this pattern wasnoted for three patients in this series.

A detailed medication history obtained in all patientswith esophagitis or distal esophageal stricture helps identifydrug-induced esophageal injury. When assessing possiblepill-induced esophageal injury, endoscopy is more sensitivethan a barium study [18], which will be only 50% sensitivefor diagnosing these entities [3]. The barium swallow, how-ever, may be helpful in detecting extrinsic compression oresophageal stricture. A definitive diagnosis requires endo-scopic examination, which can detect slight mucosalchanges not visible on barium studies, and permits biopsy to

rule out other diseases, such as cytomegalovirus (CMV)esophagitis or peptic ulcer disease [17]. The differentialdiagnosis includes monilial, herpetic, or CMV esophagitis,Crohn’s disease, Behçet’s disease [10], and ulcerations as-sociated with reflux disease [7]. The NSAID-inducedesophageal ulcers have characteristic endoscopic features—namely, large, shallow, discrete ulcers in the midesophagusnear the aortic arch, surrounded by normal mucosa with noinflammation. These traits were present in all four patientsin this series. The only patient with a hiatal hernia had noevidence of reflux esophagitis.

Most patients with pill-induced esophageal injury haveno history of swallowing difficulties [13, 14]. Tablets andcapsules can adhere to the esophageal wall and dissolvelocally even in the normal esophagus [11, 13, 14]. The mostcommon site of injury is near the level of the aortic arch [9,13, 14], as was seen in all four patients reported herein. Thecombination of external compression by the arch and aphysiologic reduction in the amplitude of the peristalticwave appear to be contributory. The sharp demarcation ofesophageal injury from normal mucosa, as was seen in eachpatient, suggests that injury results from mucosal contactwith a caustic agent (NSAIDs). This hypothesis is supportedby the fact that about 25% of reported patients sensed thata swallowed pill had become stuck [13, 14], as was de-scribed by our patient who felt a crushed ibuprofen tablet‘‘stick’’ in a retrosternal location. The chemical nature ofthe drug affects mucosal injury [9, 13]. Some drugs, such asdoxycycline or ferrous sulfate, have caustic effects [6],whereas other drugs have other mechanisms of mucosalinjury. In experimental studies acetylsalicylic acid and ibu-profen have been documented to induce esophageal muco-sal damage [4, 6]. NSAIDs may impair esophageal mucosalcytoprotection by inhibition of prostaglandin synthesis [5,17]. The precise mechanism for injury, however, remainscryptic.

NSAIDs often cause gastric or intestinal bleeding [15],but cases of NSAID-related esophageal bleeding have beenreported [3, 8, 13, 20]. Three of the four patients with largeulcers reported here required blood transfusions because ofthe extent of hemorrhage. Pill-induced hemorrhage may oc-cur due to full-thickness penetration into the left atrium [22]or a major blood vessel. The only prior patient reported with

Table 1.NSAID-Induced esophageal injury

No.Ref.No. Age Sex Chief complaints Past history Medication Duration

1. * 63 F Hematemesis; sticking sensationin esophagus

Osteoarthritis Ibuprofen(crushed tab)

1 day

2. * 41 F Hematemesis; chest pain Heavy ethanol intake Aspirin 2/day3. * 21 M Back pain Heavy ethanol intake Ibuprofen + aspirin 2 days4. * 18 F Midepigastric pain Heavy ethanol intake Ibuprofen 3/day5. 7 61 F Melena; burning sensation Low back pain Piroxicam 7 days6. 7 24 ? Dysphagia; retrosternal pain Long history of intermittent

dysphagia (? esophagitis)Ibuprofen 1 hour

7. 17 47 F Odynophagia; retrosternal pain Lumbar sprain Ibuprofen + Excedrin+ meclofenamate

4 days

8. 16 36 M Odynophagia; substernal pain Arthralgia Clinoril 1/2 day9. 8 87 F Melena Polymalgia rheumatica Naproxen + aspirin 14 days

10. 20 52 M Hematemesis Peptic ulcer disease; CVA Aspirin 1 day11. 1 28 M Dysphagia; heartburn Arthritis Indomethacin 7 days

* Our case

145

massive bleeding from aspirin-induced esophageal ulcer re-quired oversewing of the ulcer and a Nissen fundoplication[20]. The patient, herein, had successful endoscopic hemo-stasis without surgical intervention.

Most uncomplicated cases of pill-induced esophagealinjury heal if recognized early and treated [9, 13, 14]. Theoffending medication should be discontinued, or if not pos-sible, should be given as a liquid or parenterally. Treatmentshould include antacids, H2 blockade, and sucralfate in thesuspension form. Generally, patients improve within a fewdays after discontinuing medication and starting treatment.Endoscopic follow-up 3–4 weeks later proved that the ul-cers had completely healed in all three of our patients whopermitted follow-up endoscopy. Persistent symptoms maybe caused by complications, mostly esophageal stricture [2].Severe odynophagia necessitates parenteral hydration or ali-mentation. Acute inflammatory strictures often resolvespontaneously. Patients who develop chronic strictures mayrequire repeated dilatation or even surgical resection if re-fractory to dilatation [2]. Most minor hemorrhage stopsspontaneously but massive hemorrhage may require endo-scopic or surgical hemostasis.

Prevention of recurrent pill-induced esophageal injurydemands the joint effort of patients, physicians, nurses, andpharmacists [13]. All patients should be advised to takemedications only when sitting or standing, to take at least100 cc of water to wash the drug down, and to remainupright for at least an hour after ingestion [2, 3]. Cautionshould be exercised in prescribing ulcerogenic medications,especially in elderly patients with dysphagia, known ob-structive symptoms, or impaired mobility, such as the bed-ridden patient.

References

1. Bataille C, Soumagne D, Lory J, Brassinne A (1982) Esophageal ul-ceration due to Indomethacin. Digestion 24: 66–68

2. Bonavina L, DeMeester TR, McChesney L, Schwizer W, AlbertucciM, Bailey RT (1986) Drug-induced esophageal strictures. Ann Surg206: 173–183

3. Bott S, Prakash C, McCallum RW (1987) Medication-induced esoph-ageal injury: survey of the literature. Am J Gastroenterol 82: 758–763

4. Brewer AR, Smyrk TC, Bailey RB, Bonavina L, Eypasch EP, De-Meester TR (1990) Drug-induced esophageal injury: histopathologicalstudy in a rabbit model. Dig Dis Sci 35: 1205–1210

5. Brooks PM, Day RO (1991) Nonsteroidal antiinflammatory drugs—differences and similarities. N Engl J Med 324: 1716–1725

6. Carlborg B, Sensert O (1980) Esophageal lesion caused by orallyadministered drugs. An experimental study in the cat. Eur Surg Res 12:270–282

7. Coates AG, Nostrant TT, Wilson JAP (1986) Esophagitis caused bynonsteroid anti-inflammatory medication: case reports and review ofthe literature of pill-induced esophageal injury. South Med J 79: 1094–1097

8. Ecker GA, Karsh J (1992) Naproxen-induced ulcerative esophagitis. JRheumatol 19: 646–647

9. Eng J, Sabanathan S (1991) Drug-induced esophagitis. Am J Gastro-enterol 86: 1127–1133

10. Fiedorek SC, Casteel HB (1988) Pediatric medication-induced focalesophagitis. Clin Pediatr 27: 455–456

11. Hey H, Jorgensen F, Sorensen K, Hasselbalch H, Wamberg T (1982)Oesophageal transit of six commonly used tablets and capsules. BrMed J 285: 1717–1719

12. Hirschowitz BI (1994) Nonsteroidal antiinflammatory drugs and thegastrointestinal tract. Gastroenterologist 2: 207–223

13. Kikendall JW (1991) Pill-induced esophageal injury. GastroenterolClin North Am 20: 835–846

14. Kikendall JW, Friedman AC, Morakinyo AO, Fleischer D, Johnson LF(1983) Pill-induced esophageal injury. Dig Dis Sci 28: 174–182

15. Lanas A, Chandra Sekar M, Hirschowitz BI (1992) Objective evidenceof aspirin use in both ulcer and nonulcer upper and lower gastrointes-tinal bleeding. Gastroenterology 103: 862–869

16. Levine MS, Rothstein RD, Laufer I (1991) Giant esophageal ulcer dueto Clinoril. AJR Am J Roentgenol 156: 955–956

17. Minocha A, Greenbaum DS (1991) Pill-esophagitis caused by nonste-roidal anti-inflammatory drugs. Am J Gastroenterol 86: 1086–1089

18. Ovartlarnporn B, Kulwichit W, Hiranniramol S (1991) Medication-induced esophageal injury: report of 17 cases with endoscopic docu-mentation. Am J Gastroenterol 86: 748–750

19. Pemberton J (1970) Oesophageal obstruction and ulceration caused byoral potassium therapy. Br Heart J 32: 267–268

20. Schreiber JB, Covington JA (1988) Aspirin-induced esophageal hem-orrhage. JAMA 259: 1647–1648

21. Semble EL, Wu WC, Castell DO (1989) Nonsteroidal anti-inflam-matory drugs and esophageal injury. Semin Arthritis Rheum 19: 99–109

22. Sumithran E, Lim KH, Chiam HL (1979) Atrio-oesophageal fistulacomplicating mitral valve disease. Br Med J 15: 1552–1553

Table 2.NSAID-Induced esophageal injury

No.Ref.no. Endoscopic findings

G.I.bleeding

Bloodtransfusion Follow-up

1. * Long, shallow 7 × 2 cmulcer, active bleeding(25–30 cm)

Yes Yes Endoscopic hemostasis 5 days:healing 4 weeks: healed

2. * Large, shallow 5 × 5 cmulcer, with clots(28 cm)

Yes Yes

3. * Large, shallow 6 × 6 cmulcer, midesophagus Yes Yes 3 weeks:healed4. * Shallow ulcer, 4 × 3 cm (27cm–) 3 weeks: healed5. 7 Confluent ulceration (30–40 cm) Yes 3 months: linear erosion6. 7 Ulceration with impacted tablet, narrowing

(34 cm–)3 months: superficial UL

7. 17 Linear ulcer (27–32 cm)8. 16 Large, shallow 7 × 2 cmulcer

(below the level of the carina)2 months: residual UL 4 months:

depressed mucosa9. 8 Large, confluent esophageal ulcer with oozing

(distal esophagus)Yes Yes

10. 20 Large esophageal ulcer with impacted aspirin tablet Yes Yes Operation (oversewn)11. 1 Linear lcer (30 cm–) 4 weeks: healed

* Our case

146

The acute abdomen in the pregnant patient

Is there a role for laparoscopy?

A. T. Gurbuz,1,2 M. E. Peetz1

1 Department of Surgery, North Colorado Medical Center, 1801 16th Street, Greeley, CO 80631, USA2 Saint Joseph Hospital Medical Center, Denver, CO, USA

Received: 1 April 1996/Accepted 15 July 1996

AbstractBackground:The acute abdomen in the pregnant patientposes a difficult diagnostic and therapeutic challenge to thesurgeon. Appendicitis, cholecystitis, and bowel obstructionaccount for the majority of the abdominal pain syndromeswhich require surgical intervention. Laparoscopy is beingused increasingly in the diagnosis and operative manage-ment of these disorders.Methods:We examine our experience over the last 3 yearswith 47 women who developed significant abdominal painduring pregnancy. Thirty-four patients had symptomaticgallstone disease, nine had appendicitis, two had incarcer-ated inguinal hernias, and two had pelvic masses. Twenty-two patients with biliary colic and two patients with acutecholecystitis were managed conservatively during preg-nancy. Twenty-three of these underwent laparoscopic cho-lecystectomy in the postpartum period. A total of 23 womenrequired surgical intervention during pregnancy and 15 un-derwent a variety of laparoscopic procedures. Ten patientsunderwent laparoscopic cholecystectomy, and five had lap-aroscopic appendectomy. The remaining five patients hadopen appendectomy. Among the 15 laparoscopic proce-dures, four were performed in the first trimester, seven wereperformed in the second trimester, and four were performedin the third trimester.Results:Laparoscopy didn’t result in increased maternalmorbidity. There were no congenital malformations, fetallosses, or premature deliveries in the pregnant patients whounderwent laparoscopy.Conclusions:Laparoscopy can be a useful means of diag-nosis and in addition a therapeutic tool in selected pregnantpatients with abdominal pain. Close maternal and fetalmonitoring is essential during and after the procedure. Lap-aroscopic cholecystectomy is safe and can be performed

without additional risk to the fetus for those who requiresurgical intervention during pregnancy.

Key words: Pelvioscopy — Laparoscopy — Pregnancy —Symptomatic cholelithiasis — Appendicitis

Abdominal pain is a common complaint during pregnancy.It is reported by nearly all pregnant females at some timeduring the course of gestation. Pregnancy-related causes,—namely, nonpathological physiological changes, ectopicpregnancy, and placental abruption make up the majority ofabdominal pain syndromes. Fortunately general surgicalemergencies are less common. Among those, appendicitis isnumber 1 in frequency, followed by intestinal obstructionand cholecystitis [19].

The pregnant patient presents several diagnostic andtherapeutic difficulties. Some of the signs and symptoms ofacute abdomen such as nausea, vomiting, and anorexia maysimply be pregnancy induced. Anatomical changes due toenlarging uterus and displacement of the organs away fromthe anterior abdominal wall may distort the clinical picture.The uterus can also inhibit the migration of omentum to thearea of inflammation. The leukocytic response due to infec-tion may be attributed to pregnancy-induced leukocytosis[28]. The surgeon not only has to take care of the mother butalso has to ensure fetal well-being. Delay in diagnosis andtreatment of the surgical abdomen in the pregnant patientbecause of the fear of unnecessary laparotomy contributesto the high complication rate in this patient population. Aswritten by Balber in his 1908 paper, ‘‘the mortality of ap-pendicitis complicating pregnancy is the mortality of de-lay’’ [2]. Early diagnosis with aggressive treatment is there-fore essential for a successful pregnancy and fetal outcome.

Materials and methodsDuring the period from October 1992 to September 1995, 47 pregnantfemales were treated for abdominal pain of surgical etiology; 23 patientsunderwent various surgical procedures during pregnancy: ten had laparo-scopic cholecystectomies, five had laparoscopic appendectomies, four had


Correspondence to:A. T. Gurbuz, Memphis Health Sciences Center, Uni-versity of Tennessee, 956 Count Avenue A-318, Memphis, TN 38163,USA

Surg Endosc (1997) 11: 98–102

SurgicalEndoscopy


open appendectomies, two had incarcerating inguinal hernia repairs, andtwo were explored for pelvic masses.

General inhalational anesthesia with end-tidal CO2 monitoring wasemployed routinely. Sequential compression stockings were placed on allpatients, Foley catheters were placed in the appendectomy and pelvioscopygroup, and orogastric tubes were placed in the cholecystectomy group.Patients were rotated to the left to displace the uterus from the inferior venacava. The peritoneal cavity was entered with the open method (Hassontechnique) in all patients and the blunt-tip trocar was inserted into theabdomen under direct vision. CO2 was insufflated to a pressure of no morethan 12 mmHg. Fetal heart rate monitoring was employed in all patients.Intraoperative cholangiograms were performed using fluoroscopy and ad-equate fetal shielding with lead aprons.

Laparoscopic cholecystectomy was performed with the standard four-port technique (Fig. 1). The periumbilical port was placed above the uterinefundus; placement was modified according to stage of pregnancy. Thepatients that underwent laparoscopic cholecystectomy had an average ageof 29.1 (range 21–40); eight were multigravidas and two were nullipara. Atthe time of surgery two patients were in the first trimester (7, 9 weeks),seven were in the second trimester (12, 14, 15, 18, 21, 22, 22 weeks), andone was in the last trimester (27 weeks) of pregnancy. The average fetalage was 16.8 weeks. Four of the ten patients had acute cholecystitis un-responsive to conservative medical management; six were operated on asemielective basis for recurrent episodes of biliary colic interfering withmother’s nutritional status. The patients with biliary colic had an averageof 3.2 attacks (2–4) per patient. All four patients with acute cholecystitishad only one attack. Abdominal ultrasound was used for diagnosis in allpatients. Leukocytosis (WBC count 12,000–18,000) was observed in four;mild elevations in liver enzymes were observed in six patients. Two pa-tients had elevated bilirubin levels (>1.5 mg/dl); no patients had an el-evated amylase. Intraoperative cholangiograms were performed in twopatients because of elevated bilirubin (>1.5 mg/dl) and liver enzymes(twice the normal range) with or without demonstration of a dilated com-mon bile duct (>5 mm) with abdominal ultrasound examination preopera-tively.

Twenty-two pregnant patients with biliary colic and two patients withacute cholecystitis were managed conservatively during the same period.Symptoms started in the first trimester in one patient, in the second tri-mester in four patients, and in the third trimester in 19 patients. Nineteenpatients had only one attack, one had two attacks, and two had three attacksof biliary colic. Both patients with acute cholecystitis had only one singleattack. Diagnosis was verified with abdominal ultrasound, and cholelithi-asis was demonstrated in all 24 patients. There was ultrasound evidence of

acute cholecystitis (gallbladder wall thickening and pericholecystic fluid)in two patients. All patients were given intravenous fluids and an intrave-nous second-generation cephalosporin.

Appendectomies were performed in nine patients. The patients had anaverage age of 24.5 (range 21–27). Appendectomy was performed laparo-scopically in five patients and four patients had open appendectomies. Themean gestational age was 26 weeks (9, 11, 28, 30, 32 weeks) in thelaparoscopic group and 17 weeks (11, 14, 22, 24 weeks) in the open group.All patients had right lower quadrant pain, eight had elevated WBC count(>12,000), and seven were febrile (99.6°F). Laparoscopic appendectomywas performed with three ports. The first port was placed in the perium-bilical area and the second port was placed laterally in the right upperabdomen. The insertion site of the suprapubic port was modified accordingto the stage of the gestation. In the early stages of the pregnancy (less then3 months) it was placed in the lower midline under direct vision betweenumbilicus and pubic symphysis above the uterine fundus (Fig. 2). In themore advanced stages of gestation the third port was placed in the rightupper quadrant between the first two ports but in a more cranial location(Fig. 3). The periumbilical port was again placed above the uterine fundus.The upward and lateral displacement of appendix with advancing gesta-tional age makes visualization of this organ easier through the right upperabdominal approach (Fig. 3). Appendectomy was performed with the helpof endolaparoscopic stapling instruments. The appendix was placed in aplastic bag prior to extraction from the peritoneal cavity.

Two patients underwent laparoscopy for lower abdominal–pelvic pain.Abdominopelvic ultrasound demonstrated pelvic masses in both patients.The first patient was in the first trimester (8 weeks) and the second patientwas in the early second trimester (13 weeks) of gestation. A three-portexploratory laparoscopy is performed with placement of one periumbilicaland right and left midabdominal ports placed on the midclavicular line.Simple ovarian cyst aspiration was performed with resolution of the cystsin both cases.

Results

Twenty-three pregnant patients had a variety of surgicalprocedures. Fifteen of these procedures were completedwith laparoscopy. Of these 15 laparoscopic procedures, fourwere performed in the first trimester, seven were performedin the second trimester, and four were performed in the thirdtrimester.

Fig. 1. Placement of ports for a laparoscopic cholecystectomy in a 6-month-pregnant female.

Fig. 2. The port sites for laparoscopic appendectomy in the early stages of gestation.

Fig. 3. Position of appendix vermiformis in different stages of gestation and port sites for laparoscopic appendectomy at or after 6th month of pregnancy.

99

Ten patients underwent laparoscopic cholecystectomies.The average operating time was 68 min (52–102 min). Allpatients were tolerating regular diet on the 1st postoperativeday except one patient who stayed in the hospital for 3 daysbecause of prolonged nausea. None of the patients weregiven narcotics after discharge, which limited the narcoticuse for analgesia to less than 2 days. There was no maternalmorbidity or mortality. Eight patients were dischargedhome on day 1, one patient on day 2, and one patient on day3 following surgery. Average hospital stay was 1.3 days.Patients were fully ambulatory on the 1st day followingsurgery. Pathological examination showed severe acutecholecystitis in four specimens and chronic cholecystitis insix specimens. All pregnancies were followed to term andresulted with delivery of healthy babies.

Twenty-four patients were followed conservatively forsymptomatic cholelithiasis. There was one spontaneousabortion due to fulminant acute cholecystitis and sepsis.This patient’s condition worsened rapidly over severalhours and she went on to have an abortion while she wasbeing prepared for surgery. The average hospital stay was 3days. All patients were followed to term and deliveredhealthy babies; 23 patients underwent standard laparoscopiccholecystectomy with intraoperative cholangiograms aftercompletion of pregnancy. The procedures were performedin the first 8 weeks of delivery in 14 patients and after 8weeks of delivery in nine patients. The mean operative timewas 68.2 min (54–109 min) and average postoperative hos-pital stay was 1.3 days. One patient developed mild post-operative pancreatitis (peak serum amylase 420 IU/ml)probably due to reflux of radiologic contrast medium intothe pancreatic duct at the time of intraoperative cholangio-gram. She was discharged home on the 3rd postoperativeday after resolution of abdominal pain and hyperamylase-mia. Pathological examination of specimens confirmedchronic cholecystitis in 23 and acute cholecystitis in onepatient.

Nine patients were operated for right lower abdominalpain and various combinations of leukocytosis and fever.Five had appendectomy via laparoscopy while four hadopen appendectomies. Mean operating time was 64 min inthe laparoscopic group and 58 min in the open group. Av-erage hospital stay was 1.2 days in the laparoscopic groupand 1.8 days in the open group. Pathology confirmed thediagnosis of acute appendicitis in seven specimens, and noappendicial inflammation was found in two specimens. Allpatients were followed to term and all nine of them deliv-ered healthy babies.

Two patients underwent pelvioscopy for ovarian cysts.Mean operating time was 47 min. Both patients had simpleovarian cysts. Mean hospital stay was 2 days. Pregnancyresulted with delivery of healthy babies in both patients.Overall, there was no maternal morbidity related to the lap-aroscopic surgery. Laparoscopy did not result in prematurelabor or spontaneous abortion.

Discussion

Abdominal pain of surgical etiology is relatively uncommonduring pregnancy. The surgeon is faced with the dilemma ofchoosing between unnecessary operation and increased ma-ternal morbidity or fetal loss due to delay in diagnosis and

treatment. The change in the signs and symptoms of com-mon acute abdominal disease due to pregnancy-induced al-teration in the anatomy and physiology adds further to theconfusion and causes a delay in diagnosis.

Acute appendicitis is the most common nonobstetricalabdominal emergency requiring surgery during pregnancy.The incidence is one in 1,500 pregnancies and is evenlydistributed throughout the trimesters. Perforation occurs in10% of these cases and almost one-third of the removedappendices are normal pathologically. Fetal loss is reportedin 1.5% of pregnancies with uncomplicated appendicitis,but the risk increases to 35% if rupture of the appendix hasoccurred [19].

The incidence of cholecystectomy is reported to be be-tween three and eight per 10,000 pregnancies [14]. Theincidence of spontaneous abortion is 5% with uncompli-cated cholecystectomy but increases to 60% with associatedpancreatitis [7]. The management of symptomatic gallstonedisease during pregnancy is still controversial, althoughmost agree that the initial treatment should be conservative[13]. If this proves unsuccessful with recurrent severe bili-ary colic, cholecystectomy should be performed. Surgeryshould also be considered for patients with acute severecholecystitis rather early during the course, especially if thecondition has not improved rapidly with conservative treat-ment or has deteriorated with medical management.

Laparoscopy has added a new perspective to the diag-nosis and treatment of a variety of abdominal and pelvicdiseases. Owing to the success of laparoscopy in the non-pregnant patient, indications are now extending to the preg-nant patient population as well. Once an absolute contrain-dication in pregnancy [11], laparoscopy has been safelyused in the last several years for the treatment of abdomi-nopelvic pathology during pregnancy. In the series ofMazze and Kallen it constituted the largest number of sur-gical procedures performed during the first trimester [20].Laparoscopy is also helpful in defining the etiology of ab-dominal pain, therefore avoiding the delay in diagnosis [30].The pregnant patient would benefit even more from theadvantages of laparoscopic surgery—namely, easier post-operative ambulation, minimal postoperative pain, a lessernarcotic requirement, and an earlier ability to tolerate foodin the postoperative course. Average hospital stay is short-ened to less than 3 days in most patients. This is signifi-cantly shorter than the average hospitalization of 14 days forsymptomatic gallstone disease during pregnancy reportedearlier [8].

There are a total of 47 reported cases of laparoscopicprocedures performed during pregnancy in the literature [1,4–6, 9, 10, 12, 17–19, 21–23, 25–27, 29, 31–33]. Forty ofthese are cholecystectomies, six are appendectomies, andone is a laparoscopic ovarian cystectomy. The majority ofthe procedures were performed in the second trimester. In-dications for cholecystectomy were biliary colic unrespon-sive to conservative treatment in 35 patients, progressiveweight loss in two patients, acute cholecystitis in two pa-tients and acute gallstone pancreatitis in one patient. Indi-cations for appendectomy were right lower quadrant ab-dominal pain in all cases. The open technique (Hasson) waspreferred for entry into the peritoneal cavity in 29 proce-dures, whereas the Veress needle was used in 18 cases. Fourpatients were in the first trimester, 39 patients were in sec-ond trimester, and three patients were in the third trimester.

100

Cholangiograms were obtained in 13 of 40 cholecystecto-mies. Perioperative fetal monitoring was employed in 30cases and tocolytic therapy was instituted in two patientsprophylactically. No intra- or postoperative complicationsare described. Twenty-seven of 47 women were followed toterm with delivery of healthy babies. One infant developedhyaline membrane disease shortly after birth. The remaining20 women were reported to be continuing uncomplicatedgestation.

The enlarged uterus and the fetus are two points ofconcern during laparoscopy in a pregnant patient. The fun-dus of the uterus is at the pelvic brim by 12 weeks andreaches the umbilicus by the 20th week of gestation. Thus,there is a chance of injuring the uterus, the overlying en-larged uterine vessels, or the fetus during blind attempts ofinsufflation using the Veress needle, which has been previ-ously reported [3]. This can potentially result in uterineinsufflation and CO2 embolism. Others have recommendedinserting the Veress needle through the seventh intercostalspace on the left side [24] or using the umbilicus as the pointof entry but aiming the needle toward the right upper quad-rant to prevent uterine puncture. An alternative technique ofabdominal wall lifting is also described [9]. Opening theabdominal wall in layers and inserting the blunt trocar underdirect vision—therefore, the Hasson technique—should bethe preferred method of entry. The gravid uterus can alsointerfere with the visualization and instrumentation duringthe procedure. Minimal manipulation of the uterus is essen-tial. Pushing, pulling, or retracting the uterus should beavoided. Placement of trocars should be modified accordingto the height of the uterus. The periumbilical port has to bemoved cranially after the 20th week. Preexisting high in-traabdominal pressure due to the pregnant uterus and itscontents may cause difficulty in insufflation and in achiev-ing an adequate pneumoperitoneum in the advanced stagesof pregnancy. The latest gestational age at which one canperform laparoscopy safely is uncertain. There may not beenough room in the upper abdomen to accommodate theinstruments, and exposure of the lower abdomen and pelvismay be difficult in the late third trimester. Although initialreports had recommended that laparoscopy be performedduring the second trimester only [21], laparoscopic chole-cystectomy has been reported during the 25th, 28th, [10],and 31st weeks [23] of gestation. We also performed alaparoscopic appendectomy in the 32nd week of pregnancy.Each patient requires an individualized judgment by theoperating surgeon as to the feasibility of laparoscopy.

The potential untoward effects of intraperitoneal CO2and high intraperitoneal pressures are another point of con-cern. The fetus absorbs CO2 from maternal blood andthrough the uterine circulation [15]. It is known that CO2pneumoperitoneum causes an acute but mild hypercarbiaand respiratory acidosis in the mother. Since the pH of thefetal blood parallels the mother’s pH, a slight acidosis isalso detected in the fetus. Fortunately, the fetal acidosis ispartially reversible with maternal hyperventilation. Despitethe changes in the arterial CO2 and pH, the fetal PO2changes only minimally. It has also been demonstrated thatthere is a significant lag time between development of ma-ternal acidosis and increase in the end-tidal CO2 values [16].Therefore, close hemodynamic monitoring of the mother

with capnography and preferably an arterial line is neces-sary.

Considering the high pressures generated by the uterusat term, intraabdominal pressures up to 15 mmHg shouldnot be harmful to the fetus. No significant hemodynamiceffect of increased intraabdominal pressure is demonstratedin the pregnant ewe model if CO2 is replaced with N2O [16].

Fetal heart-tone monitoring should be an essential partof the procedure. The demonstration of fetal bradycardia ortachycardia should result in immediate exsufflation. Again,close communication with the obstetrician is necessary, es-pecially in the late second and the third trimester of preg-nancy in case of premature labor. The presence of an expe-rienced obstetrics nurse in the operating room at the time offetal monitoring would also be helpful.

To minimize possible damage to the fetus during theorganogenesis period, the first trimester should not be cho-sen for elective or semielective procedures. Similarly, ab-dominal operations and general anesthesia can cause pre-mature labor in the third trimester. Thus, it would be saferto defer any elective procedure to the second trimester whenthe organogenesis is complete and the uterus is not bigenough to impair the visual field. If the mother can becarried to term with conservative management, these riskswill be eliminated. The indications for intraoperative chol-angiography are not different from the nonpregnant pa-tient—namely, an elevated serum bilirubin (1.5 mg/dl), adilated common bile duct on the preoperative ultrasound(>8 mm), or gallstone pancreatitis. For the patients in whomlaparoscopic cholecystectomy and laparoscopic appendec-tomy were performed, results from the literature and ourpatients show no spontaneous abortions despite the reportedabortion rate of 5% with open cholecystectomy in the lit-erature [7].

Conclusions

Laparoscopy can be utilized in the pregnant patient for di-agnosis and treatment of selective abdominal emergencies.The procedure has to be carefully planned and conducted byan experienced laparoscopic surgeon. Close monitoring ofthe mother and the fetus is essential. Modifications of theport placement and surgical technique are required to pre-vent injury to the gravid uterus and the fetus. Laparoscopyin the pregnant patient does not result in fetal loss. Lapa-roscopic procedures do not increase maternal morbidity andactually may decrease hospital stay and recovery time com-pared to open procedures.

References

1. Arvidsson D, Gerdin E (1991) Laparoscopic cholecytectomy in preg-nancy. Surg Laparosc Endosc 1: 193–194

2. Balber EA (1908) Perforative appendicitis complicating pregnancy.JAMA 51: 1310–1313

3. Barnett MB, Lui DTY (1974) Complications of laparoscopy duringearly pregnancy. Br Med J 23: 328

4. Bennett TL, Estes N (1993) Laparoscopic cholecystectomy in the sec-ond trimester of pregnancy. J Reprod Med 38: 833–834

5. Camitalo JB, Lynch D (1994) Laparoscopic cholecystectomy in thepregnant patient. Surg Laparosc Endosc 4: 268–271

6. Constantino GN, Vincent GJ, Mukalian GG, Kliefoth WL (1994) Lap-

101

aroscopic cholecystectomy in pregnancy. J Laparendosc Surg 4: 161–164

7. Dixon NP, Green J, Rogers A, Rubin L (1983) Fetal loss after chole-cystectomy during pregnancy. Can Med Assoc J 88: 576–577

8. Dixon NP, Faddis DM, Silberman H (1987) Aggressive managementof cholecystitus during pregnancy. Am J Surg 154: 292–294

9. Edelman DS (1994) Alternative laparoscopic technique for cholecys-tectomy during pregnancy. Surg Endosc 8: 794–796

10. Elerding SC (1993) Laparoscopic cholecystectomy in pregnancy. AmJ Surg 165: 625–627

11. Gadacz TR, Talamini MA (1991) Traditional versus laparoscopic cho-lecystectomy. Am J Surg 161: 336–338

12. Hart RO, Tamadon A, Fitzgibbons RJ, Fleming A (1993) Open lapa-roscopic cholecystectomy during pregnancy. Surg Laparosc Endosc 3:13–16

13. Hiatt JR, Hiatt JG, Williams RA, Klein SR (1986) Biliary disease inpregnancy: a strategy for surgical management. Am J Surg 151: 263–265

14. Hill LM, Johnson CE, Lee RA (1975) Cholecystectomy in pregnancy.Obstet Gynecol 46: 291–293

15. Hunter JG (1992) 1991 Scientific session and postgraduate course:advances in laparoscopic general surgery. Surg Endosc 6: 47–56

16. Hunter JG, Swanstrom L, Thornburg K (1995) Carbon dioxide pneu-moperitoneum induces fetal acidosis in a pregnant ewe model. SurgEndosc 9: 272–279

17. Iarfati MD, Yarnell R, Schwaitzberg SD (1995) Gasless laparoscopiccholecystectomy in pregnancy. J Laparendosc Surg 5: 127–130

18. Jackson SJ, Sigman HH (1993) Laparoscopic cholecystectomy inpregnancy. J Laparendosc Surg 3: 35–39

19. Kammerer WS (1979) Non-obstetric surgery during pregnancy. MedClin North Am 63: 1157–1163

20. Mazze RI, Kallen B (1989) Reproductive outcome after anesthesia andoperation during pregnancy. Am J Obstet Gynecol 161: 1178–1185

21. Morrell DG, Mullins JR, Harrison PB (1992) Laparoscopic cholecys-tectomy during pregnancy in symptomatic patients. Surgery 112: 856–859

22. Nezhat F, Nezhat C (1991) Laparoscopic ovarian cystectomy duringpregnancy. J Laparendosc Surg 1: 161–164

23. Pucci RO, Seed RW (1991) Case report of a laparoscopic cholecys-tectomy in the third trimesterof pregnancy. Am J Ostet Gynecol 165:401–402

24. Reich H, Hunter JG (1992) Is it safe to perform laparoscopic appen-dectomy during pregnancy? Laparosc Focus 1: 11–12

25. Schorr RT (1993) Laparoscopic cholecystectomy and pregnancy. JLaparendosc Surg 3: 291–293

26. Schreiber JH (1990) Laparoscopic appendectomy in pregnancy. SurgEndosc 4: 100

27. Shaked G, Twena M, Charuzzi I (1994) Laparoscopic cholecystectomyfor empyema of gallbladder during pregnancy. Surg Laparosc Endosc4: 65–76

28. Smoleniec J, James D (1990) General surgical problems in pregnancy.Br J Surg 77: 1203–1204

29. Soper NJ, Hunter JG, Petrie RH (1992) Laparoscopic cholecystectomyduring pregnancy. Surg Endosc 6: 115–117

30. Spirtos NM, Eienkop SM, Spirtos TW, Poliakin RI, Hibbard LT(1987) Laparoscopy—a diagnostic aid in cases of suspected appendi-citis. Am J Obstet Gynecol 156: 90–94

31. Weber AM, Bloom GP, Allan TR, Curry SL (1991) Laparoscopiccholecystectomy during pregnancy. Obstet Gynecol 78: 958–959

32. Williams JK, Rosenmurgy AS, Albrink MH, Parsons MT, Stock S(1995) Laparoscopic cholecystectomy in pregnancy: a case report. JReprod Med 40: 243–245

33. Wilson RB, McKenzie RJ, Fisher JW (1994) Laparoscopic cholecys-tectomy in pregnancy: two case reports. Aust NZ J Surg 64: 647–649

102

surgical endoscopy 3

Documents

incidence of duct injury

laparoscopic technique

cases ofbile duct injury

laparoscopic cholecystectomyd

laparoscopic ap

impending injury

injury rate

open cholecystectomy