Bacterial Contamination of the Biliary Tree after Choledochostomy
William Sllen, MD, Boston, Massachusetts lUllchael Werthelmer, MD, Worcester, Massachusetts Gerald Klrshenbaum, MD,’ Boston, Massachusetts
It is well known that bacterial infection of the urinary tract occurs commonly after urethral catheterization. Little attention has been paid, however, to the frequency or long-term consequences of bacterial colonization of the biliary tract after choledochostomy and T-tube drainage. Severe systemic sepsis not uncommonly follows T-tube cholangiography, in many instances secondary to organisms not originally present in the biliary tract [I]; but the fate of these organisms after removal of the T tube is not known and is rarely even considered. Chronic biliary infection, especially with gram-negative bacteria, is often associated with multiple biliary calculi of the pigment type (calcium bilirubinate) because of the production of a bacterial P-glucuronidase which deconjugates the bilirubin glucuronide of hepatic bile allowing the combination of free bilirubin with calcium [2]. Although it is generally conceded that bacteria will thrive in a biliary tree which is even partially obstructed or in which a foreign body is present, it is not known whether microorganisms will continue to thrive in an anatomically normal biliary tract or what factors influence the persistence or disappearance of the organisms. There is little question but that under certain clinical circumstances biliary infection is associated with multiple pigment calculi even when the sphincter of Oddi is patulous and no strictures are present [3]. Our own clinical experiences, albeit completely anecdotal, suggest that in some instances serious potentially preventable biliary infections, which may be caused by persistence of postoperative bactobilia, develop in the remote postoperative period after choledochostomy and T-tube drainage. Because appropriate cultures of bile are impossible Fromthe departmentsof Surgery, Harvard Medical School, and Seth Israel Mspital, Boston, Massachusetts. Reprint requests should be addressed to William Silen, MD. Department of Surgery, Seth Israel Hospital. 330 Srookllne Avenue, Boston, Massachusetts 02215. Resent address: 5863 S Adams Street, Fort Knox, Kentucky 40121. As of July 1077: V. A. Hospital, 4150 Clement Street. San Francisco, Callfomia 94121. l
Volume 135, March 1878
to obtain
at intervals after removal of a T tube, absolute proof for this hypothesis is not attainable with present technics. Yet, the high incidence of bacterial contamination after choledochostomy demonstrated in the current study suggests the adoption of a few simple, safe, and practical measures for prophylaxis against the possibility of lingering biliary infection in a manner analogous to that commonly employed in the genitourinary system. Material and Methods
A retrospective study was made of the hospital records of 201 patients who had undergone operations on the common bile duct in the period between July 1966 and January 1975. A bacteriologic study was made of the bile obtained from the common duct at operation and the results compared with specimens from the T tubes of the same patients obtained at various intervals from five days to six weeks postoperatively. Complete pre- and postoperative cultures were available from ninety-five patients, and these constituted the study group. Standard bacteriologic technics were employed for isolation, subculture, and identification of organisms. No attempt was made at selection of patients either in terms of clinical presentation (that is, bacteremia) or abnormalities found at operation, although the majority (91 per cent) in fact had common duct calculi. Results
Of the study group, 42 per cent (40 patients) who were initially sterile had bacteria in the common duct bile at the time of operation. In only thirteen of the remaining fifty-five patients (14 per cent) did the T-tube bile remain sterile throughout the period of observation, until the T tubes were removed. Forty-two patients (44 per cent of the total) had initially sterile bile which grew positive cultures during the period of observation. The enteric organisms klebsiella group and Escherichia coli were the most commonly found, &hough no less than nine groups of microorganisms were isolated. (Table I.)
325
Silen. Wertheimer, and Kirshenbaum
Illustrative
Cases
Case I. VK (B.I.H. #38-lo-77), in 1958 at the age of forty-five years, underwent an uneventful cholecystectomy and common bile duct exploration at another hospital for symptomatic biliary tract disease. At operation, multiple calculi of unknown type were removed from the common bile duct, which was 9 mm in diameter. The papilla of Vater readily accepted a 7 mm Bakes dilator. Postoperative Ttube cholangiography showed a normal biliary tract without residual calculi, and cultures of the bile grew klebsiella and Escherichia coli. The patient remained well until April 1973, when she began having repeated episodes of fever, chills, epigastric pain, and jaundice. In June 1973 she underwent operation at the Beth Israel Hospital. The common bile duct was 14 mm in diameter and contained hundreds of 1 to 2 mm pigment calculi. The papilla of Vater easily admitted a 5 mm Bakes dilator, but because of the myriad of stones, wide sphincteroplasty was carried out. The common duct bile at operation and postoperatively was found to have klebsiella and Escherichia coli in large numbers. A course of tetracycline to which both organisms were sensitive was given for ten days postoperatively. The patient has remained well to date. Comment. Our review of the records concerning the 1958 operation allowed comparison of the bile cultures taken in 1958 and 1973. Proof is lacking that these bacteria continued to grow in an anatomically normal common bile duct, but the circumstances are suggestive.
Case II. IS (B.I.H. #46-33-74), in 1962 at the age of sixty-nine years, underwent cholecystectomy and common bile duct exploration at the Beth Israel Hospital for symptomatic biliary tract disease. At operation, multiple calculi of unknown type were removed from the common bile duct, which was 2 cm in diameter. The papilla of Vater readily accepted a 7 mm Bakes dilator. Postoperative Ttube cholangiography showed the common bile duct to be without residual calculi, and cultures of the T-tube bile grew klebsiella and Escherichia coli. The patient remained well until June 1972, when she began having repeated episodes of fever, chills, and right
TABLE I
Organisms Pound in Initially Sterile Bile 5 Days to 6 Weeks after T-Tube Drainage No. of
Patients Klebsiella aerobacter, enterobacter 25 11 Escherichia coli 7 Enterococcus 6 Proteus Staphylococcus aureus, coagulase-positive 5 Staphylococcus epidermitis 5 3 Pseudomonas 3 Serratia 1 Clostridia Note: More than 1 organism was found in 14 patients.
326
upper quadrant pain. Exploration was performed at the Beth Israel Hospital in July 1972, at which time the common bile duct measured 3 cm in diameter and contained innumerable 1 to 2 mm pigment calculi. A 5 mm Bakes dilator could be passed easily into the duodenum, but because of the multiple calculi, wide sphincteroplasty was performed. Cultures of the common duct bile at operation grew klebsiella and Escherichia coli with the same antibiotic sensitivity pattern as was found in 1962. The patient was given a ten day course of tetracycline to which both organisms were sensitive after removal of the T tube. She has remained completely well, except that in 1976 a tender fluctuant mass appeared in her incision site. Incision and drainage was carried out; a single silk suture was removed, culture of which grew E coli. Comment. The biliary tree had been cleared of calculi completely at the original operation in 1962 but the bactobilia was not treated after removal of the T tube. The circumstantial evidence is reasonably strong that the bacteria continued to proliferate in this dilated and thick-walled common bile duct even though the papiIla of Vater was normal.
Comments
Bacterial colonization of the biliary tree develops frequently after choledochostomy with T-tube drainage of the common bile duct. The circumstances are parallel to those of the patient subjected to urethral catheterization for a similar length of time, an analogy previously suggested by Flemma et al [4]. However, the long-term consequences of colonization of the biliary tract are not so clearly defined as in the urinary system, Whereas the bacterial flora of the urinary tract can be monitored easily by repeated examination of readily available urine, the same cannot be said of the biliary tract once the T tube has been removed. Obviously, if a stricture or calculus is present, it is unlikely that the biliary tree will be cleared of bacteria until the lesion is corrected. The unanswerable question remains, however, as to the fate of the colonized but unobstructed biliary tree. Our findings in this investigation are remarkably similar to those of Keighley et al [I] who also demonstrated superinfection with new organisms in patients after choledochostomy with T-tube drainage. They in fact emphasized that early postoperative infective complications were more closely related to the bacteria found in cultures taken postoperatively from the T tube than to the findings of cultures taken at operation. When postoperative cholangiography demonstrates that the biliary tree is free of calculi or strictures and has free drainage into the duodenum, it is likely that chronic infection is uncommon, since most patients who undergo choledochostomy and T-tube
The American Journalof Surgery
Contaminationafter Choledochostomy
drainage have an excellent ultimate result. Yet enough patients, similar to the two described herein, are encountered to suggest that chronic bacterial infection may supervene in the anatomically normal biliary tract. Once chronic infection is present, especially with glucuronidase-producing gram-negative bacteria, precipitation of deconjugated bilirubin as calcium bilirubinate is enhanced [2]. The surgeon has only sphincteroplasty or choledochoduodenostomy to offer the patient with thousands of tiny calculi throughout the ducts. The more frequent use of sphincteroplasty and choledochoduodenostomy in recent years in patients with recurrent choledochblithiasis, even in the presence of a normal papilla of Vater, is testimony to the need for an operation that allows ready passage of calculi into the gastrointestinal tract. Since neither of these operations is without problems, it seems desirable to examine those factors that might lead to the establishment of bacterial infections responsible for producing calculi. Although the mechanisms for propulsion of bile in the common bile duct are poorly understood, the presence of an extremely dilated common bile duct or a duct with a very thick wall could lead to sluggish flow and the development of secondary infection. A vicious cycle could be set up by bacterial deconjugation of bile salts within the biliary tree, since it is now known that certain bile salts themselves inhibit the growth of anaerobic intestinal bacteria [5]. Finally, since bacterial contamination is common after choledochostomy with T-tube drainage, it would appear prudent to attempt to minimize the frequency and seriousness of this event. In our opinion, several steps might prevent chronic infection in a biliary tree devoid of anatomic abnormalities. We recognize that their efficacy will be difficult to prove because bile is not available for culture after removal of the T tube and because it will require careful observation and results of cultures of bile of patients for many years. Yet, these measures are simple, safe, and logically sound. First, a closed system should be employed for biliary drainage, to effectively prevent contamination of the urinary tract for approximately one week. Second, cultures of T-tube bile should be taken a few days before the T tube is removed. If bacterial contamination is present, appropriate antibiotics should be given for seven to ten days after removal of the T tube, since eradication of the organisms in the presence of this foreign body is unlikely. Although primary closure of the duct [6] would eliminate the need
Volume 136. March 1978
for the T tube and thus avoid contamination, most surgeons hesitate to use this technic as a routine procedure. Summary
Secondary exogenous contamination of bile after choledochostomy could be an important cause of long-term morbidity if infection persists after removal of the T tube. Surprisingly, documentation of the frequency of conversion of sterile to infected bile after choledochostomy has rarely been recorded. Patients undergoing exploration and T-tube drainage of the common bile duct between July 1966 and January 1975, in whom intraoperative and postoperative cultures of bile were available, were studied. Postoperative cultures were obtained from five to forty days after operation. Of ninety-five patients available for study, contamination developed postoperatively in previously sterile bile in 44 per cent. The common duct bile contained bacteria at operation in 42 per cent and remained sterile throughout in 14 per cent. The most common secondary contaminants were klebsiella group and Escherichia coli. The fate of the contaminating organisms in a biliary tree without anatomic abnormalities is unknown. Two cases are presented which suggest that persistent infection may linger in the biliary tree for many years and give rise to pigment calculi. Since bacteria cannot be eradicated so long as a foreign body remains in the common duct and because such bacteria may cause symptoms even in an anatomically normal biliary tree, we suggest that a closed system of biliary drainage be employed and that appropriate antibiotic therapy be instituted for seven to ten days after removal of the T tubes. References 1. Keighley MRB, Lister DM, Jacobs SI, Giles GR: Hazards of surgical treatment due to microorganisms in the bile. Surgery 75: 578, 1974. 2. Maki T: Pathogenesis of calcium bilirubinate gallstones: role of E. coli, fl glucuronidase and coagulation by inorganic ions, polyelectrolytes and agitation. Ann Surg 164: 90, 1966. 3. Wen C-C, Lee H-C: lntrahepatic stones: a clinical study. Ann Surg 175: 166, 1972. 4. Flemma RI, Flint LM, Osterhout S, Shingleton WW: Bacteriologic studies of biliary tract infection. Ann Surg 166: 563, 1967. 5. Williams RC, Showalter R, Kern R Jr: In vivo effect of bile salts and cholestyramine on intestinal anaerobic bacteria. Gastroenterology 69: 483, 1975. 6. Collins PG, Redwood CRM, Wynne-Jones G: Common bile-duct suture without intraductal drainage following choledochotomy. Br J Surg 47: 661, 1960.
327
William Sllen, MD, Boston, Massachusetts lUllchael Werthelmer, MD, Worcester, Massachusetts Gerald Klrshenbaum, MD,’ Boston, Massachusetts
It is well known that bacterial infection of the urinary tract occurs commonly after urethral catheterization. Little attention has been paid, however, to the frequency or long-term consequences of bacterial colonization of the biliary tract after choledochostomy and T-tube drainage. Severe systemic sepsis not uncommonly follows T-tube cholangiography, in many instances secondary to organisms not originally present in the biliary tract [I]; but the fate of these organisms after removal of the T tube is not known and is rarely even considered. Chronic biliary infection, especially with gram-negative bacteria, is often associated with multiple biliary calculi of the pigment type (calcium bilirubinate) because of the production of a bacterial P-glucuronidase which deconjugates the bilirubin glucuronide of hepatic bile allowing the combination of free bilirubin with calcium [2]. Although it is generally conceded that bacteria will thrive in a biliary tree which is even partially obstructed or in which a foreign body is present, it is not known whether microorganisms will continue to thrive in an anatomically normal biliary tract or what factors influence the persistence or disappearance of the organisms. There is little question but that under certain clinical circumstances biliary infection is associated with multiple pigment calculi even when the sphincter of Oddi is patulous and no strictures are present [3]. Our own clinical experiences, albeit completely anecdotal, suggest that in some instances serious potentially preventable biliary infections, which may be caused by persistence of postoperative bactobilia, develop in the remote postoperative period after choledochostomy and T-tube drainage. Because appropriate cultures of bile are impossible Fromthe departmentsof Surgery, Harvard Medical School, and Seth Israel Mspital, Boston, Massachusetts. Reprint requests should be addressed to William Silen, MD. Department of Surgery, Seth Israel Hospital. 330 Srookllne Avenue, Boston, Massachusetts 02215. Resent address: 5863 S Adams Street, Fort Knox, Kentucky 40121. As of July 1077: V. A. Hospital, 4150 Clement Street. San Francisco, Callfomia 94121. l
Volume 135, March 1878
to obtain
at intervals after removal of a T tube, absolute proof for this hypothesis is not attainable with present technics. Yet, the high incidence of bacterial contamination after choledochostomy demonstrated in the current study suggests the adoption of a few simple, safe, and practical measures for prophylaxis against the possibility of lingering biliary infection in a manner analogous to that commonly employed in the genitourinary system. Material and Methods
A retrospective study was made of the hospital records of 201 patients who had undergone operations on the common bile duct in the period between July 1966 and January 1975. A bacteriologic study was made of the bile obtained from the common duct at operation and the results compared with specimens from the T tubes of the same patients obtained at various intervals from five days to six weeks postoperatively. Complete pre- and postoperative cultures were available from ninety-five patients, and these constituted the study group. Standard bacteriologic technics were employed for isolation, subculture, and identification of organisms. No attempt was made at selection of patients either in terms of clinical presentation (that is, bacteremia) or abnormalities found at operation, although the majority (91 per cent) in fact had common duct calculi. Results
Of the study group, 42 per cent (40 patients) who were initially sterile had bacteria in the common duct bile at the time of operation. In only thirteen of the remaining fifty-five patients (14 per cent) did the T-tube bile remain sterile throughout the period of observation, until the T tubes were removed. Forty-two patients (44 per cent of the total) had initially sterile bile which grew positive cultures during the period of observation. The enteric organisms klebsiella group and Escherichia coli were the most commonly found, &hough no less than nine groups of microorganisms were isolated. (Table I.)
325
Silen. Wertheimer, and Kirshenbaum
Illustrative
Cases
Case I. VK (B.I.H. #38-lo-77), in 1958 at the age of forty-five years, underwent an uneventful cholecystectomy and common bile duct exploration at another hospital for symptomatic biliary tract disease. At operation, multiple calculi of unknown type were removed from the common bile duct, which was 9 mm in diameter. The papilla of Vater readily accepted a 7 mm Bakes dilator. Postoperative Ttube cholangiography showed a normal biliary tract without residual calculi, and cultures of the bile grew klebsiella and Escherichia coli. The patient remained well until April 1973, when she began having repeated episodes of fever, chills, epigastric pain, and jaundice. In June 1973 she underwent operation at the Beth Israel Hospital. The common bile duct was 14 mm in diameter and contained hundreds of 1 to 2 mm pigment calculi. The papilla of Vater easily admitted a 5 mm Bakes dilator, but because of the myriad of stones, wide sphincteroplasty was carried out. The common duct bile at operation and postoperatively was found to have klebsiella and Escherichia coli in large numbers. A course of tetracycline to which both organisms were sensitive was given for ten days postoperatively. The patient has remained well to date. Comment. Our review of the records concerning the 1958 operation allowed comparison of the bile cultures taken in 1958 and 1973. Proof is lacking that these bacteria continued to grow in an anatomically normal common bile duct, but the circumstances are suggestive.
Case II. IS (B.I.H. #46-33-74), in 1962 at the age of sixty-nine years, underwent cholecystectomy and common bile duct exploration at the Beth Israel Hospital for symptomatic biliary tract disease. At operation, multiple calculi of unknown type were removed from the common bile duct, which was 2 cm in diameter. The papilla of Vater readily accepted a 7 mm Bakes dilator. Postoperative Ttube cholangiography showed the common bile duct to be without residual calculi, and cultures of the T-tube bile grew klebsiella and Escherichia coli. The patient remained well until June 1972, when she began having repeated episodes of fever, chills, and right
TABLE I
Organisms Pound in Initially Sterile Bile 5 Days to 6 Weeks after T-Tube Drainage No. of
Patients Klebsiella aerobacter, enterobacter 25 11 Escherichia coli 7 Enterococcus 6 Proteus Staphylococcus aureus, coagulase-positive 5 Staphylococcus epidermitis 5 3 Pseudomonas 3 Serratia 1 Clostridia Note: More than 1 organism was found in 14 patients.
326
upper quadrant pain. Exploration was performed at the Beth Israel Hospital in July 1972, at which time the common bile duct measured 3 cm in diameter and contained innumerable 1 to 2 mm pigment calculi. A 5 mm Bakes dilator could be passed easily into the duodenum, but because of the multiple calculi, wide sphincteroplasty was performed. Cultures of the common duct bile at operation grew klebsiella and Escherichia coli with the same antibiotic sensitivity pattern as was found in 1962. The patient was given a ten day course of tetracycline to which both organisms were sensitive after removal of the T tube. She has remained completely well, except that in 1976 a tender fluctuant mass appeared in her incision site. Incision and drainage was carried out; a single silk suture was removed, culture of which grew E coli. Comment. The biliary tree had been cleared of calculi completely at the original operation in 1962 but the bactobilia was not treated after removal of the T tube. The circumstantial evidence is reasonably strong that the bacteria continued to proliferate in this dilated and thick-walled common bile duct even though the papiIla of Vater was normal.
Comments
Bacterial colonization of the biliary tree develops frequently after choledochostomy with T-tube drainage of the common bile duct. The circumstances are parallel to those of the patient subjected to urethral catheterization for a similar length of time, an analogy previously suggested by Flemma et al [4]. However, the long-term consequences of colonization of the biliary tract are not so clearly defined as in the urinary system, Whereas the bacterial flora of the urinary tract can be monitored easily by repeated examination of readily available urine, the same cannot be said of the biliary tract once the T tube has been removed. Obviously, if a stricture or calculus is present, it is unlikely that the biliary tree will be cleared of bacteria until the lesion is corrected. The unanswerable question remains, however, as to the fate of the colonized but unobstructed biliary tree. Our findings in this investigation are remarkably similar to those of Keighley et al [I] who also demonstrated superinfection with new organisms in patients after choledochostomy with T-tube drainage. They in fact emphasized that early postoperative infective complications were more closely related to the bacteria found in cultures taken postoperatively from the T tube than to the findings of cultures taken at operation. When postoperative cholangiography demonstrates that the biliary tree is free of calculi or strictures and has free drainage into the duodenum, it is likely that chronic infection is uncommon, since most patients who undergo choledochostomy and T-tube
The American Journalof Surgery
Contaminationafter Choledochostomy
drainage have an excellent ultimate result. Yet enough patients, similar to the two described herein, are encountered to suggest that chronic bacterial infection may supervene in the anatomically normal biliary tract. Once chronic infection is present, especially with glucuronidase-producing gram-negative bacteria, precipitation of deconjugated bilirubin as calcium bilirubinate is enhanced [2]. The surgeon has only sphincteroplasty or choledochoduodenostomy to offer the patient with thousands of tiny calculi throughout the ducts. The more frequent use of sphincteroplasty and choledochoduodenostomy in recent years in patients with recurrent choledochblithiasis, even in the presence of a normal papilla of Vater, is testimony to the need for an operation that allows ready passage of calculi into the gastrointestinal tract. Since neither of these operations is without problems, it seems desirable to examine those factors that might lead to the establishment of bacterial infections responsible for producing calculi. Although the mechanisms for propulsion of bile in the common bile duct are poorly understood, the presence of an extremely dilated common bile duct or a duct with a very thick wall could lead to sluggish flow and the development of secondary infection. A vicious cycle could be set up by bacterial deconjugation of bile salts within the biliary tree, since it is now known that certain bile salts themselves inhibit the growth of anaerobic intestinal bacteria [5]. Finally, since bacterial contamination is common after choledochostomy with T-tube drainage, it would appear prudent to attempt to minimize the frequency and seriousness of this event. In our opinion, several steps might prevent chronic infection in a biliary tree devoid of anatomic abnormalities. We recognize that their efficacy will be difficult to prove because bile is not available for culture after removal of the T tube and because it will require careful observation and results of cultures of bile of patients for many years. Yet, these measures are simple, safe, and logically sound. First, a closed system should be employed for biliary drainage, to effectively prevent contamination of the urinary tract for approximately one week. Second, cultures of T-tube bile should be taken a few days before the T tube is removed. If bacterial contamination is present, appropriate antibiotics should be given for seven to ten days after removal of the T tube, since eradication of the organisms in the presence of this foreign body is unlikely. Although primary closure of the duct [6] would eliminate the need
Volume 136. March 1978
for the T tube and thus avoid contamination, most surgeons hesitate to use this technic as a routine procedure. Summary
Secondary exogenous contamination of bile after choledochostomy could be an important cause of long-term morbidity if infection persists after removal of the T tube. Surprisingly, documentation of the frequency of conversion of sterile to infected bile after choledochostomy has rarely been recorded. Patients undergoing exploration and T-tube drainage of the common bile duct between July 1966 and January 1975, in whom intraoperative and postoperative cultures of bile were available, were studied. Postoperative cultures were obtained from five to forty days after operation. Of ninety-five patients available for study, contamination developed postoperatively in previously sterile bile in 44 per cent. The common duct bile contained bacteria at operation in 42 per cent and remained sterile throughout in 14 per cent. The most common secondary contaminants were klebsiella group and Escherichia coli. The fate of the contaminating organisms in a biliary tree without anatomic abnormalities is unknown. Two cases are presented which suggest that persistent infection may linger in the biliary tree for many years and give rise to pigment calculi. Since bacteria cannot be eradicated so long as a foreign body remains in the common duct and because such bacteria may cause symptoms even in an anatomically normal biliary tree, we suggest that a closed system of biliary drainage be employed and that appropriate antibiotic therapy be instituted for seven to ten days after removal of the T tubes. References 1. Keighley MRB, Lister DM, Jacobs SI, Giles GR: Hazards of surgical treatment due to microorganisms in the bile. Surgery 75: 578, 1974. 2. Maki T: Pathogenesis of calcium bilirubinate gallstones: role of E. coli, fl glucuronidase and coagulation by inorganic ions, polyelectrolytes and agitation. Ann Surg 164: 90, 1966. 3. Wen C-C, Lee H-C: lntrahepatic stones: a clinical study. Ann Surg 175: 166, 1972. 4. Flemma RI, Flint LM, Osterhout S, Shingleton WW: Bacteriologic studies of biliary tract infection. Ann Surg 166: 563, 1967. 5. Williams RC, Showalter R, Kern R Jr: In vivo effect of bile salts and cholestyramine on intestinal anaerobic bacteria. Gastroenterology 69: 483, 1975. 6. Collins PG, Redwood CRM, Wynne-Jones G: Common bile-duct suture without intraductal drainage following choledochotomy. Br J Surg 47: 661, 1960.
327
