Impaired clearance of escherichia coli bacteremia in early biliary obstruction

Impaired Clearance of Escherichia coli Bacteremia in Early Biliary Obstruction Carol E.H. Scott-Conner,

James B. Grogan, rho, Kenneth S. Scher, Jack Bernstein, MD, Jackson, Mississippi

MD, FACS,

Adult male rats underwent common bile duct ligation or sham celiotomy. At intervals of 7 and 14 days postoperatively, bacteremia was induced by intravenous injection of lo9 Escherichia coli or intraperitoneal injection of lo6 E. coli. Serial quantitative blood cultures and quantitative whole organ cultures were obtained. One week after surgery, clearance of bacteremia was impaired in all of the animals. Clearance of intraperitoneally injected E. cob was less efficient in the duct ligation rats. Fourteen days postoperatively, clearance of bacteremia induced by intravenous or intraperitoneal injection had improved iu the sham celiotomy rats but was still significantly impaired in the duct ligation rats. An increased number of viable E. coli were recovered from the lungs of duct ligation rats after intravenous administration. We found that rats with obstructive jaundice do not respond normally to a bacteremic challenge. This impairment in reticuloendothelial function can be noted as early as 1 week after common duct ligation.

bstructive jaundice is associated with depression of reticuloendothelial function and increased suscepti0 bility to infection. Early studies demonstrated decreased clearance of intravenously injected particulate substances such as carbon or macroaggregated serum albumin in animals with biliary obstruction [1,2]. More recent experiments using bacteria as a test substance have shown increased survival of viable organisms within the lungs of rats with biliary obstruction but have not clearly demonstrated whether or not clearance of bacteria from the bloodstream is impaired [J-4, unpublished data]. The present study was undertaken to determine the extent to which the clearance of Escherichia coli, a typical gramnegative enteric organism, is impaired in rats with early biliary obstruction. From the Department of Surgery, University of Mississippi School of Medicine, Jackson, Mississippi. Supported in part by Grant 2 SO7 RR05386 awarded by the Biomedical Research Support Grant Program, Division of Research Resources, National Institutes of Health, Bethesda, Maryland. Requests for reprints should be addressed to Carol E.H. ScottConner, MD, Department of Surgery, University of Mississippi School of Medicine, 2500 North State Street, Jackson, Mississippi 392164505. 210

THE AMERICAN JOURNAL OF SURGERY

MATERIAL

MD, FACS,

AND METHODS

Male Sprague-Dawley or Lewis strain rats weighing 250 to 300 grams were housed in groups of six, fed standard laboratory chow (Purina 5001@rodent laboratory chow, Ralston-Purina, Circleville, OH), and permitted water ad libitum until operation. National Research Council guidelines for animal housing and care were followed throughout the study. The animals were anesthetized with intramuscular ketamine hydrochloride. Celiotomy with ligation and complete transsection of the common bile duct or sham celiotomy with mobilization of the duct was performed using clean technique. During the postoperative period, the animals were given unrestricted access to food and water. All subsequent studies were performed on the 7th or 14th postoperative day. The animals were divided into four experimental groups: Groups I (18 duct ligation rats, 18 sham celiotomy rats) and II (5 duct ligation, 5 sham celiotomy) were challenged with intraperitoneal injection of lo6 E. coli 1 and 2 weeks after surgery, respectively; Groups III (5 duct ligation rats, 5 sham celiotomy) and IV (5 duct ligation, 5 sham celiotomy) were challenged with intravenous administration of lo9 E. coli 1 and 2 weeks after surgery, respectively. Group I was the first group studied. A larger number of animals were included in this group because initial results with a small number of animals suggested that a small difference between duct ligation and sham celiotomy rats might become significant with a larger number of animals. When this did not occur, subsequent experiments were limited to five duct ligation and five sham celiotomy animals per group. Intraperitoneal injection: Sprague-Dawley rats were used for these preliminary studies to permit comparison with previous experiments in splenectomized animals [5]. E. coli (ATCC 259212, American Type Culture Collection, Rockville, MD) were grown in tryptic soy broth. The final bacterial concentration was prepared to optical standards and confirmed by standard plate counts. Bacteremia was produced by intraperitoneal injection of lo6 E. coli. Tail vein blood samples were obtained 15, 30, 45, 60, 90, 120, and 240 minutes after bacterial injection according to a protocol previously described [5]. Quantitative cultures were obtained by spreading the blood on blood agar plates, and then the number of colony-forming units per milliliter of blood was determined. Intravenous injection: Subsequent experiments were performed on male Lewis strain rats. The strain was changed to facilitate study of immunologic parameters in later experiments. Bacterial strain and culture techniques were not changed. Bacteremia was produced by intravenous injection of lo9 E. coli. Serial quantitative blood

VOLUME 157 FEBRUARY 1989

cultures were obtained 1,5,10,15,20,25,30,60, and 240 minutes after injection. Clearance of intravenously injected organisms was assumed to fol!ow first-order kinetics during the initial phase, as described by Benacerraf et al [q and Biozzi et al [ 71. The loglo bacterial concentration was plotted as a function of time. The phagocytic index (K), defined as the slope of the initial rapid clearance phase, was calculated by least squares analysis applied to the first three points of each curve. The corrected phagocytic index (a) was calculated from the following equation [2,q:

where Wa was the body weight of the rat in grams and WE was the combined liver and spleen weight in grams. Quantitative organ cultures were obtained by a method previously described, in which weighed homogenates of lung, liver, and spleen were plated on blood agar plates and the number of colony-forming units per gram of tissue and per organ determined [4,8,unpublished data]. Serum was obtained for direct and indirect bilirubin determinations, and selected organs were studied histologically. Clearance curves were plotted for each animal. The phagocytic and corrected phagocytic indices were calculated for each clearance curve obtained in Groups III and IV. For each experimental group, the geometric mean, standard error of the mean, and standard deviation were calculated for each point on the curve. Groups were compared using the Student’s t test for independent samples. Mortality rates among groups were compared using the &i-square test. Statistical significance was defined at the p <0.05 level. The statistical analysis was performed using a commercially available statistical package (Abstat 4.08, Anderson-Bell, Canon City, CO) on an IBMPC.

RESULTS All rats that underwent common duct ligation were visibly jaundiced by 1 week. The serum bilirubin level at 1 week (duct ligation: Groups I and III) was 6.5 f 1.5 mg/ dl compared with 0.4 f 0.3 mg/dl for sham celiotomy control rats (p
IQQO

Sham

100

Ia

Time Aiter InJection (min) We 1. Clear_anceof bactemfnla hhcad by htrapwttonea inof lo6 Eschehhlacollinratslweekaftercommon bile duct l&Won (CBM) or sham cektomy (SC). A SlgnlfbM dlrfwnw(p<0.005)Isnotedbetweenthetw0 groups 240 minutes postoperatively;CFU = colony-forming units.

6.2 f 2.1 mg/dl in duct ligation rats compared with 0.3 f

0.4 mg/dl in sham celiotomy rats (p <0.05). Liver weight in the duct ligation rats was 17 f 0.8 g compared with 8.4 f 1 g in sham celiotomy rats (p <0.05). Histologic examination of duct ligation livers revealed more pronounced changes in biliary obstruction. Weights of duct ligation spleens, lungs, and kidneys were not different from sham celiotomy rats, and no histologic abnormalities were apparent in these organs. Five of 15 duct ligation rats (33 percent) died by 2 weeks after surgery compared with none of the 10 sham celiotomy rats (p = NS). Postmortem examination revealed similar findings of widespread patchy hepatic necrosis. The body weight of the duct ligation rats was 293.8 f 25.9 g compared with 351 f 63.1 g (p = NS) for the sham-operated control rats. Group I. One week after surgery, clearance of bacteremia produced by intraperitoneal injection was delayed significantly in both duct ligation and sham celiotomy rats (Figure 1). Although there appeared to be a shift in the curves, the difference between the duct ligation rats and the sham celiotomy rats was not statistically significant with the exception of 240 minutes after injection. Group II. After 2 weeks, clearance of bacteremia in the sham celiotomy control rats had returned to baseline values and was comparable to results obtained previously in intact rats [5,9]. In contrast, clearance was significantly retarded in duct ligation rats after 2 weeks of jaundice

JOURNAL

OF SURGERY

VOLUME

157

FEBRUARY

1989

211

1000 CBDL * I’ ./

*

l’

i

,

/’

100

IO

I I 1 I 1530 45 60

I 90

‘~,Q+ztrn ’ 240

120

Time after Injection (min) Figure 2. Clearance of bacteremla Induced by intraperlioneal InJectIonof 10’ Escherlchla toll In rats 2 weeks after common bile duct llgallon (CBDL) or sham cellotomy. The differences between the two curves are slgnlflcant at 120 mlnutes (p <0.02) and 240 mlnutes (p
(Figure 2), and an increasing number of organisms were recovered 2 and 4 hours after injection. Croup III. Clearance curves obtained 1 week after surgery showed delayed clearance of bacteremia in both duct ligation and sham celiotomy rats. Clearance curves obtained for duct ligation and sham celiotomy animals were not significantly different. The phagocytic index (K) was 0.23 f 0.007 for duct ligation rats compared with 0.21 f 0.03 for sham celiotomy rats (p = NS); however,

the corrected phagocytic index (a) was 9.4 f 0.9 for duct ligation rats and 17.1 f 2 for sham celiotomy rats (p
TABLE I Whole-Organ Baclerlal Cultures Obtained 4 Hours After Induction of Bacteremla by Intravenous Injection of lOa Escherlchla toll’

Organ

Viable Organisms (colony-forming SC (n = 5)

CBDL (n = 5)

One week postop Liver Spleen Lungs Two weeks postop Liver Spleen Lungs

7.3 x IO’ f 2.1 x 10’ f 8.9 X 10s f

1.2 x 107 1.4 x 107 2.1 X 10s

10.2 x 10’ f 3.7 x IO’ f 9.2 x 10s f

2.1 x 10’ 2.2 x 10’ 3.4 x 10s

5.3 x 10’ f 7.5 X lo6 f 2.5 x 107 f

1.5 x 10’ 2.3 X 10s 7.0 x 106

9.9 x 107 f 1.2 X 10’ i 4.8 X lo6 f

1.9 x 10’ 2.5 X 10s 3.3 X 10s

s Values expressed as the mean f SE. CBDL = common bile duct @ation; SC = sham celiotomy.

212

THE AMERICAN

JOURNAL

units/organ)

OF SURGERY

VOLUME

157

FEBRUARY

1989

p Value

NS NS NS NS NS p <0.05

BACTEREMIACLEARANCEINEARLYBILIARYOBSTRUCTION

from the bloodstream appeared unaffected, increased numbers of viable organisms were recovered from the lungs of jaundiced rats. The kinetics of bacterial clearance from the bloodstream of normal animals was investigated and described in detail by Benacerraf et al [q after earlier studies into the mechanism of clearance of nonbacterial particulates [ 71. The intact animal rapidly clears induced bacteremia unless an overwhelming inoculum or a highly virulent organism is used. An initial rapid clearance phase is characterized by an exponential decrease in the number of organisms recovered from the bloodstream. First-order kinetics can be used to describe the rate of removal of bacteria from the blood during this phase. The phagocytic index is defined as the slope of the straight line obtained when log10bacterial concentration is plotted as a function of time. This index is highly dependent upon the total weight of the liver and spleen relative to body weight [6,7]. The corrected phagocytic index incorporates a correction factor that allows comparison of animals with differing liver and spleen weightto-body weight ratios. In our experiments, the phagocytic index was not significantly different from control values 1 week after bile duct ligation. However, because of changes in liver and spleen weight relative to body weight, the corrected phagocytic index showed some deterioration in phagocytic function even as early as 1 week postoperatively. Two weeks postoperatively, both the crude and the corrected phagocytic indices were significantly less in duct ligation rats than in sham celiotomy rats, indicating less rapid initial clearance of bacteremia in the jaundiced animals. After the initial rapid clearance phase, subsequent removal of bacteria occurs at a slower rate [q. When a sufficiently large inoculum or a sufficiently virulent organism is used, bacteria may reappear in the bloodstream in increasing numbers after a variable lag phase. Uncontrolled bacterial proliferation may result in death. In our experiments, no difference in the later phase of clearance was noted 1 week after surgery. Two weeks postoperatively, duct ligation rats showed consistently higher levels of bacteria in the blood until sacrificed 4 hours after administration (Figure 3). Sham-operated rats exhibited some impairment 1 week postoperatively compared with shams 2 weeks after operation, suggesting that nonspecific responses to celiotomy may still be observed at this early time period. Data from the sham celiotomy rats at 2 weeks were very similar to those previously obtained after 1 month, confirming that the animals had returned to preoperative condition at that time [5]. Route of administration is also a significant factor. Slow, sustained bacteremia is a characteristic response to intraperitoneal injection in control as well as in experimental animals [&b,unpublished data]. The blood bacterial level after intraperitoneal injection of E. coli represents a balance between continued absorption from the peritoneal cavity and removal of organisms by the reticuloendothelial system. Previous studies utilizing the intraperitoneal injection route demonstrated sequestration of large numbers of viable organisms in the liver, lung, and THE AMERICAN

10’

?

10’

k I a

10"

3

0

T

z a

P

ID I

10’

.-.._.*---__ l__CfDL i Shorn

r’

J

103 ,

‘:A./

,_ SO

5 1015202530

240

Time After InJoctlon (min)

Figure 3.Clearancd of bacteremia hc&ced by intravenous InJecth of 10s Eschwlchla CONIn rats 2 weeks after common bile duct ligation (CBDL) ix sham ceibtomy. The curves are sigMcaMly dihrenl for all points (p
spleen of rats with biliary obstruction when compared with sham-operated control rats [I,unpublished data]. Impaired clearance was suggested from these earlier studies by the fact that heart blood cultures were positive 4 hours after injection in a significant number of animals with biliary obstruction, The present study details the decrease in clearance of bacteremia induced by intraperitoneal injection and demonstrates that nonspecific effects of celiotomy are still detected 1 week postoperatively. When bacteremia persists after intraperitoneal injection, delayed absorption from the peritoneal cavity, persistence of intraabdominal organisms, intravascular or intraperitoneal proliferation of organisms, impaired reticuloendothelial system clearance, and sequestration of viable organisms in other organs, such as the lungs, may be involved. Persistent bacteremia after intravenous injection is more likely to be directly related to either impaired clearance by the reticuloendothelial system, defective phagocytosis, intravascular proliferation, or sequestration in other organs [10,12], Impaired function of neutrophils and tissue macrophages obtained from icteric rats has been demonstrated in vitro [13,14]. Persistence of viable organisms in the lungs after induced bacteremia has been noted by other observers and may be one of the mechanisms for pulmonary dysfunction in sepsis [ 23171. Large numbers of viable organisms were recovered from the lungs of animals with biliary obstruction in the present study, as well as in our previous studies [I,unpubfished data] and those of Katz et al [3]. These sequestered bacteria may serve as a reservoir for continued bacteremia. Although weight loss is frequently noted in bile ductligated rats, studies of nutritional status have revealed no

JOURNAL

OF SURGERY

VOLUME

157

FEBRUARY

1989

213

SCOTT-CONNER ET AL

evidence of malnutrition [ 191. Body weight was not significantly different between jaundiced and sham-operated rats in our studies. Thus, it is unlikely that nutritional depletion contributes significantly to the observed defect in phagocytosis. The decrease in rate of clearance of bacteria from the bloodstream after intravenous injection in the present study confirmed the decreased reticuloendothelial system efficiency noted with nonbacterial particulates in earlier studies. This impairment in phagocytosis could be noted as early as 1 week postoperatively, when the phagocytic index was corrected for changes in liver and spleen weight associated with bile duct ligation. In addition, when bacteremia was induced by intraperitoneal injection, signilicantly greater numbers of bacteria could be recovered from the blood of duct ligation animals 4 hours after injection, also suggesting that clearance is impaired 1 week after surgery. Two weeks after surgery, bacterial clearance curves obtained from sham-operated rats returned to baseline values similar to those of intact animals. In contrast, clearance of bacteremia induced by either the intravenous or intraperitoneal route remained impaired 2 weeks after common duct ligation, This depression in bacterial clearance, noted as early as 1 week after bile duct ligation, may explain some of the septic, renal, and pulmonary complications noted in early obstructive jaundice. It also suggests that patients with jaundice of less than 2 weeks’ duration may be unable to handle a bacteremic challenge and therefore may be at increased risk of complications. REFERENCES 1. Holman JM, Rikkers LF. Reticuloendothelial function and biliary obstruction. Curr Surg 1980; 37: 366-7. 2. Holman JM, Rikkers LF. Biliary obstruction and host defense failure. J Surg Res 1982; 32: 208-13. 3. Katz S, Grosfeld JL, Gross K, et al. Impaired bacterial clearance and trapping in obstructive jaundice. Ann Surg 1984; 199: 14-20. 4. Scott-Conner CEH, Bernstein JM, Scher KS, Mack ME. The effect of biliary obstruction on a gram-negative bacteremic challenge: a preliminary report. Surgery 1986; 99: 679-83. 5. Scher KS. Wroczvnski F. Coil JA. The effect of ~r~~~~~~ solenectomv ~~~~,on ._.

214 THE AMERICAN JOURNAL OF SURGERY

gram-negative bacteremia. J Trauma 1982; 22: 407-9. 6. Benacerraf B, Sebestyen MM, Schlossman S. A quantitative study of the kinetics of blood clearance of P32-labelled Escherichia coli and staphylococci by the reticuloendothelial system. J Exp Med 1959; 110: 27-48. 7. Biozzi G, Benacerraf B, Halpern BN. Quantitative study of the granulopectic activity of the reticuloendothelial system: II. A study of the kinetics of the granulopectic activity of the RES in relation to the dose of injective. Relationship between the weight of the organs and their activity. Br J Exp Pathol 1953; 34: 441-57. 8. Scher KS. Prevention of wound infection: The comparative effectiveness of topical and systemic cefazolin and povidone-iodine. Am Surg 1982; 48: 268-70. 9. Scher KS, Wroczynski AF, Jones CW. Protection from postsplenectomy sepsis: Effect of prophylactic penicillin and pneumococcal vaccine on clearance of type 3 Pneumococcus. Surgery 1983; 93: 792-7. 10. Rogers DE. Studies on bacteriemia. I. Mechanisms relating to the persistence of bacteriemia in rabbits following the intravenous injection of staphylococci. J Exp Med 1956; 103: 713-42. 11. Rogers DE. Host mechanisms which act to remove bacteria from the blood stream. Bact Rev 1960; 245: 50-66. 12. Wood WB, Smith MR, Perry WD, Berry JW. Studies on the cellular immunology of acute bacteriemia. I. Intravascular leucocytic reaction and surface phagocytosis. J Exp Med 1951; 94: 521-33. 13. Roughneen PTM, Drath DB, Kulkarni AD, Rowlands BJ. Extrahepatic cholestasis alters neutrophilic and tissue macrophage phagccytic and metabolic activity. Surg Forum 1986; 37: 148-50. 14. Roughneen PT, Drath DB, Kulkarni AD, et al. Impaired nonspecific cellular immunity in experimental cholatasis. Ann Surg 1987; 206: 578-82. 15. Crccker SH, Lowery BD, Eddy DO, et al. Pulmonary clearanceof blood-borne bacteria. Surg Gynecol Obstet 1981; 153: 84551. 16. Lanser ME, Saba TM. Neutrophil-mediated lung localization of bacteria: A mechanism for pulmonary injury. Surgery 198 1; 90: 473-81. 17. Niehaus GD, Schumacker PR, Saba TM. Reticuloendothelial clearance of blood-borne particulates. Relevance to experimental lung microembolization and vascular injury. Ann Surg 1980; 191: 479-87. 18. Harrow EM, Jakab GJ, Brody AR, Green GM. The pulmonary response to a bacteremic challenge. Am Rev Resp Dis 1975; 112: 7-16. 19. Gouma DJ, Roughneen PT, Kumar S, et al. Changes in nutritional status associated with obstructive jaundice and biliary drainage in rats. Am J Clin Nutr 1986; 44: 362-9.

VOLUME 157 FEBRUARY 1989