Levels Of Bacteremia Research Articles

Effect of methylprednisolone on bacterial clearance and endotoxin liberation during experimental sepsis induced by gram-negative bacteria

To determine the effect of methylprednisolone administration on the clearance of bacteremia and the release and clearance of endotoxin during antibiotic therapy of gram-negative bacterial sepsis, Escherichia coli K1 sepsis was induced in paired rabbits. Moxalactam and either methylprednisolone or placebo were administered to infected rabbits 1.5 h after intraperitoneal administration of live bacteria. Serial blood samples were obtained for quantitation of bacteremia and endotoxemia, arterial blood gases, and complete blood count. Arterial blood pressure, heart rate, and core body temperature were also monitored. There were no significant differences between the methylprednisolone-treated and placebo-treated groups in either the levels of bacteremia or endotoxemia or in the physiologic, metabolic, or hematologic parameters that were measured. We conclude that methylprednisolone administration has no acute effect on bacterial clearance or on the kinetics of endotoxin release and clearance during antibiotic therapy of gram-negative bacterial sepsis in this experimental model.

Haemophilus influenzae type b infection in childhood: history of bacteremia and antigenemia.

Groups of children (mean age, 31.4 months) with Haemophilus influenzae type b meningitis, epiglottitis, or septic arthritis were tested for the presence and levels of bacteremia, capsular polyribophosphate (PRP) antigenemia, and development of specific antibody in serum after the onset of acute illness. Although bacteremia cleared promptly after antibiotic therapy, circulating PRP could be detected in serum for relatively long periods, with 51% of the patients still having detectable antigen after 30 days postinfection. Even in the presence of specific antibody, antigenemia persisted for as long as 47 days after admission. It was observed that there was no statistically significant correlation between the persistence of antigenemia and age (P greater than 0.2), the initial antigen concentration (P greater than 0.50), or the development of antibody (P greater than 0.20). The presence of a low magnitude of bacteremia (less than 300 organisms per ml) was associated with a maximum concentration of 10 ng of PRP per ml. On the other hand, bacterial counts in excess of 10(4)/ml were associated with greater than 1,000 ng of PRP per ml (r = 0.98, r2 = 0.96, P less than 0.001). It was observed that the amount of circulating PRP in the acute phase of illness was related to whether a child developed convalescent-phase antibody. Invariably, the younger children, who primarily had meningitis, had a PRP concentration of greater than 10 ng/ml and failed to develop an antibody response in any isotype, whereas the older patients, who primarily had infections other than meningitis, had a PRP concentration of less than 10 ng/ml and a 45.5% success rate in developing an antibody response (P = 0.006). These findings suggest that there is a direct correlation between the magnitudes of bacteremia and antigenemia, that antigen may persist for long periods even in the presence of antibody, and that the level of antigenemia in addition to the patient age is significantly related to the nature of the convalescent-phase antibody response.

Routine Quantitative Blood Cultures in Children With Haemophilus influenzae or Streptococcus pneumoniae Bacteremia

The potential clinical value of quantitative blood cultures determined by a commercially available lysis-direct plating method was studied in 50 children with either Haemophilus influenzae or Streptococcus pneumoniae bacteremia. The magnitude of bacteremia correlated with the severity of the infection; patients with greater than or equal to 100 colony-forming units per milliliter were significantly more likely to have meningitis (P less than .01, chi 2 = 7.5). On the other hand, all patients with S. pneumoniae bacteremia with colony counts lower than 15 colony-forming units per milliliter had "occult bacteremia" with no focus of infection. The data suggest that patients with higher levels of bacteremia have more severe disease. Quantitative blood culture results may be helpful in identifying which children are at risk for invasive disease.

Role of antibiotic class in the rate of liberation of endotoxin during therapy for experimental gram-negative bacterial sepsis.

To evaluate the role of antibiotic class in the rate of liberation of endotoxin during therapy for sepsis caused by gram-negative bacteria, we obtained serial blood samples from rabbits with sepsis caused by Escherichia coli and treated with chloramphenicol, gentamicin, or moxalactam. The concentrations of viable bacteria, free endotoxin, and total endotoxin in each blood sample were measured. In rabbits treated with chloramphenicol, the geometric mean levels of free endotoxin remained proportional to the geometric mean levels of bacteremia, a result indicating the absence of antibiotic-induced endotoxin liberation. In contrast, levels of free endotoxin increased rapidly while levels of bacteremia declined after treatment with gentamicin or moxalactam, a result indicating antibiotic-induced release of endotoxin. Despite similar rates of bacterial killing, mean levels of free endotoxin were as much as 20-fold higher in rabbits treated with moxalactam than in paired rabbits receiving gentamicin (P less than .05). These results indicate that endotoxin liberation during therapy for sepsis caused by gram-negative bacteria is dependent upon the class of antibiotic administered and is not necessarily correlated with the rate of bacterial killing.

Lack of antibacterial activity after intravenous hydrogen peroxide infusion in experimental Escherichia coli sepsis

The intravenous administration of hydrogen peroxide has been reported to benefit patients with pneumonia and to reduce Plasmodium parasitemia in experimentally infected mice. We assessed the antibacterial activity of intravenously infused hydrogen peroxide against hydrogen peroxide-susceptible Escherichia coli (MBC of hydrogen peroxide, 0.23 mM) in experimentally infected rabbits. No decrease in the level of bacteremia was detected at the maximum intravenous infusion rate of hydrogen peroxide physiologically tolerated by the rabbits (2.0 mumol/h). Moreover, the addition ex vivo of greater amounts of hydrogen peroxide to human or murine blood containing E. coli resulted in no detectable antibacterial action. In contrast, ethyl hydrogen peroxide, which is not affected by catalase, was bactericidal when added ex vivo to human blood containing E. coli. These results suggest that extracellular hydrogen peroxide, whether of exogenous or endogenous origin, does not have antibacterial activity in the blood of animals having even low levels of catalase.

1129 HAEMOPHILUS INFLUENZAE TYPE B (HIB) INFECTION IN CHILDHOOD: NATURAL HISTORY OF BACTEREMIA AND ANTIGENEMIA

Groups of children, up to 7½ years of age, with meningitis, epiglottitis or septic arthritis due to HIB were tested for the presence and levels of bacteremia, capsular polyribophosphate (PRP) antigenemia and development of HIB specific antibody in serum following onset of acute illness. The quantitative direct plating method and indirect ELISA were used to determine the number of organisms, PRP, and immunoglobulin class of HIB specific antibody. Bacteremia and antigenemia were detected in all patients immediately after onset of illness. Although bacteremia cleared promptly after antibiotic therapy, PRP could be detected in serum up to 96 days after onset of illness. The presence of a low magnitude bacteremia ( 1000 ng/ml of PRP. Despite this correlation, persistence of PRP antigenemia appeared to be independent of a) magnitude of antigenemia observed during the acute phase, b) development of antibody, or c) patient's age at time of illness. However, serum antibody response was observed in convalescent samples only in subjects who exhibited peak PRP concentrations <10 ng/ml (p=0.006). These findings suggest a direct correlation between the magnitudes of bacteremia and antigenemia and that antigen may persist for long periods even in the presence of antibody and the level of antigenemia significantly influences the convalescent antibody response.

211 ENDOTOXIN LIBERATION DURING THERAPY FOR GRAM-NEGATIVE BACTERIAL SEPSIS

To determine if plasma levels of free endotoxin increase during the initial hours of antibiotic therapy for gram-negative bacterial sepsis, serial blood samples were obtained from 24 patients presenting with suspected sepsis. For each sample the level of bacteremia was determined by quantitative culture, and the levels of free (filterable) and bacterial cell-bound plasma endotoxin were measured by Limulus lysate assay. Fourteen patients had no detectable bacteremia or endotoxemia. Five patients with constant or declining levels of free endotoxemia had transient gram-negative bactereraia (2), no detectable bacteremia (2), or overwhelming gram-positive bacterial sepsis (1). The remaining five patients had gram-negative bacterial sepsis ranging from 30 to 107 cfu/ml blood. Plasma free endotoxin increased in all of these patients during the initial hours of antibiotic therapy, while levels of bacteremia and cell-bound endotoxin were decreasing. Plasma free endotoxin subsequently cleared in four of these patients, but continued to rise in the other patient, associated with progressive purpura fulminans and irreversible shock. These results are consistent with the hypothesis that antibioticinduced endotoxin release may be responsible for the deterioration observed in some patients with gram-negative bacterial sepsis during antibiotic therapy.

Role of nasopharyngeal colonization with and without bacteremia in the protection of infant rats against Haemophilus influenzae type b challenge.

Nasopharyngeal colonization of infant rats with Haemophilus influenzae type b was investigated by two methods of intranasal inoculation. After traumatic instillation of the bacteria, 100% of the animals became colonized, compared with 75.5% of animals after atraumatic instillation. Among colonized rats, significantly more animals in the traumatic group developed bacteremia compared with those in the atraumatic group. Rats in the traumatic group had an onset of bacteremia at a mean of 2.6 days after inoculation compared with 7.3 days in the atraumatic group. The duration of colonization and bacteremia was the same in both groups. The majority of heavily colonized rats developed bacteremia compared with none of the lightly colonized rats. Thus, the development of bacteremia appeared to be related independently to both heavy colonization and traumatic instillation. Protection against intraperitoneal bacterial challenge with H. influenzae type b developed in rats that had been bacteremic; in the majority of animals, this correlated with the development of serum bactericidal activity. Protection and bactericidal activity were only rarely observed in nonbacteremic rats that had been either heavily or lightly colonized. The development of serum bactericidal activity was not related to either the duration or peak level of bacteremia. Thus, in this rat model, H. influenzae type b nasopharyngeal colonization without bacteremia did not appear to stimulate circulating antibodies that protected the animals against challenge.

Improved detection of significant pathogens in blood culture

As we have treated children with malignancies for combination chemotherapy of multiple anticancer agents, severe neutropenia has been often recognized. Once they become severe neutropenia, they often have high fever, but we cannot find out the etiology of their fever. Most researchers recognize that the etiology of this fever is septicemia, but they cannot almost detect pathogenic microorganisms by conventional blood culture. In this paper, I discussed that the possibilities of improvement in blood culture method, and I obtained next results.1) I evaluated whether in vitro antibiotics, which were present in blood, have influence on the detection of microorganisms by blood culture. In artificial bacteremia 3.0×10°FU/ml of S. aureus was inhibited by 1.0μg/ml of gentamicin. Therefore, if bacteremia level is less than 3.0×10°FU/ml of S. aureus, blood culture, which was done within 6 hours after drip infusion of 2 mg/kg of gentamicin, might not be positive. As in bacteremia of E. coli the growth of bacteria was inhibited by more than 10.0μg/ml of gentamicin, the results of blood culture were not influenced by conventional dosage of gentamicin. I evaluated whether in vivo antibiotics which were present in patients' blood have influence on the detection of pathogens by blood culture. Blood culture which was done after 2 hours after administration of antibiotics did not have influence on the growth of microorganisms.2) I evaluated that blood volume from 5% to 30% inoculated into media has influence on the detection of microorganisms. At both blood of leukemic patients with leukopenia and normal healthy adults, the detection rates of bacteria made no difference. Namely it took more than 24 hours to recognize the growth of bacteria at 5% inoculated blood volume, and even 30% inoculated blood volume did not inhibit the growth of bacteria.3) I tried to separate bacterial pathogens from all antibacterial properties of the blood, including antibiotics, complement, phagocytes and antibodies by centrifugation. But controlled culture with no treatment had a trend to growth more rapidly, I could not obtain more sensitive culture method by centrifugation.4) I tried to make the layer in which almost all of microorganisms were present by centrifugation. At the speed of 3, 000 rpm, bacterial pathogens remained in the layer of blood cells. All microorganisms could not be sedimented at the bottom of blood cells layer.

A minor high-molecular-weight outer membrane protein of Haemophilus influenzae type b is a protective antigen.

Cell surface-exposed antigenic determinants of several high-molecular-weight outer membrane proteins of Haemophilus influenzae type b (Hib) have been shown to be consistently immunogenic in human infants convalescing from Hib meningitis. A monoclonal antibody (mab), 6G12, directed against one of these cell surface-exposed outer membrane proteins that has an apparent molecular weight of 98,000 (98K) was identified by radioimmunoprecipitation analysis. Of 120 clinical isolates of Hib, 83 were found to possess antigenic determinants which reacted with mab 6G12 in a colony blot-radioimmunoassay procedure, indicating that the antigenic determinant recognized by mab 6G12 is present in the majority of Hib strains. A different radioimmunoassay, which uses whole Hib cells as antigen, confirmed that strains reactive with mab 6G12 in the colony blot-radioimmunoassay procedure possessed a cell surface-exposed and antibody-accessible antigenic determinant recognized by this mab. Hib strains which did not react with mab 6G12 were found to lack a 98K protein. Passive immunization with mab 6G12 reduced the level of bacteremia that developed in infant rats challenged with the homologous Hib strain against which this mab was raised. In contrast, no protection was observed when the challenge strain was one which lacks the antigenic determinant recognized by mab 6G12. Radioimmunoprecipitation analysis of sera from human infants convalescing from Hib meningitis detected an antibody response directed against the 98K protein. The protection against experimental Hib disease provided by antibody to the 98K protein, the immunogenicity of this protein in human infants, and its presence in a majority of Hib strains indicate that the 98K outer membrane protein may have potential for vaccine development.

Kinetics of endotoxin release during antibiotic therapy for experimental gram-negative bacterial sepsis.

For study of the kinetics of endotoxin release from bacterial cells during therapy for gram-negative bacterial sepsis, serial blood samples were obtained from rabbits with Escherichia coli sepsis that were treated with either antibiotic or placebo. The concentrations of viable bacteria, free endotoxin, and total endotoxin in each blood sample were quantitated. In animals treated with placebo the concentration of free endotoxin was proportional to the level of bacteremia. In contrast, in animals treated with antibiotic the plasma levels of free endotoxin increased 10- to 2,000-fold, in spite of decreasing levels of bacteremia. Free endotoxin that was present in the plasma following antibiotic treatment appeared to be derived in part from the breakdown of circulating bacteria and in part from the disintegration of bacteria in tissues other than the blood. The results of this study demonstrate that significant amounts of endotoxin are released from bacterial cells following administration of antibiotics in vivo.

COMPARATIVE STUDY OF ENDOTOXIN RELEASE FOLLOWING THERAPY WITH AN AMINOGLYCOSIDE VERSUS A THIRD GENERATION CEPHALOSPORIN

In order to determine if the rate of endotoxin Liberation following antibiotic therapy of gram-negative bacterial sepsis is related to the antibiotic class administered, sepsis was induced in 12 rabbits by intraperitoneal injection of E. coli K1. Blood was sampled just prior to and 1/2 hour and hourly following antibiotic administration. Bacteremia was measured by quantitative culture. Plasma endotoxin levels before and after passage through a 0.45 micron filter were quantitated by Limulus lysate assay. Two hours after infection rabbits received either 2.5 mg/kg of gentamicin or 100 mg/kg of moxalactam intravenously. Both groups of rabbits developed similar mean levels of bacteremia and free endotoxin prior to antibiotic therapy. However, beginning within 1/2 hour after antibiotic administration and continuing for 4 hours, two to ten-fold higher mean levels of free (filterable) endotoxin were present in the moxalactam-treated rabbits as compared to the gentamicin-treated rabbits (p <.05 for each sample time). These higher mean levels of free endotoxin occurred in the moxalactam group although there was no significant difference in the mean rates of bacterial killing. Assessment of the clinical significance, if any, of the observed difference in rates of endotoxin liberation following aminoglycoside therapy versus third generation cephalosporin therapy will require further studies employing biochemical and physiological measurements.

Quantitation of bacteria in cerebrospinal fluid and blood of children with meningitis and its diagnostic significance.

Cerebrospinal fluid (CSF) specimens from pediatric patients with meningitis were examined for their concentration of microbes and the relationship of this count to the bacteremia levels, microscopy results, and polymorphonuclear leukocyte concentration. A total of 2,031 consecutive CSF specimens were analyzed, of which 63 (3.1%) were positive by culture from the same number of patients. We observed that 85% of the total CSF specimens positive for Haemophilus influenzae type b, Streptococcus agalactiae, Streptococcus pneumoniae, and Neisseria meningitidis had counts in excess of 10(3) CFU/ml, with 56% of the specimens exceeding 10(5) CFU/ml. A correlation existed between the number of organisms present in the CSF and blood. For example, from a total of 22 patients who had counts of H. influenzae greater than 10(3) CFU/ml in the CSF, 16 or 73% had levels of bacteremia greater than 10(3) CFU/ml. It was also noted that the bacterial concentration had a profound effect on the sensitivity of microscopy. The percentage of positive results increased from 25% with less than or equal to 10(3) CFU/ml to 60% in the range of greater than 10(3) to 10(5) CFU/ml and to 97% at concentrations of greater than 10(5) CFU/ml. Furthermore, a significant correlation (P less than 0.01) was noted between the concentration of bacteria in the CSF and the number of polymorphonuclear leukocytes observed on microscopy.

Experimental pneumococcal meningitis: role of leukocytes in pathogenesis.

Two groups of rabbits with experimental meningitis induced by direct intracisternal inoculation of Streptococcus pneumoniae cells were studied. One group was rendered profoundly leukopenic by nitrogen mustard, and the other had normal leukocyte counts. The two groups had comparable bacterial growth rates (mean generation time, 60 versus 67 min) and ultimate bacterial populations in the cerebrospinal fluid (CSF) (mean log10 CFU, 9.1 versus 8.7); therefore leukocytes did not effectively slow or limit the growth of pneumococci in the CSF in vivo. Increased CSF protein, decreased CSF glucose, and increased CSF lactate levels were similar in both groups, suggesting that leukocytes are not essential for these changes to occur. Quantitative blood cultures revealed identical levels of pneumococcal bacteremia until 13 to 14 h after the initiation of infection, when the leukopenic rabbits showed a larger number of pneumococci in the blood, ultimately exceeding the number reached in nonleukopenic rabbits by 100-fold. Leukocytes therefore limit the extent of pneumococcal bacteremia after infection of the CSF despite their lack of effect on the course or the CSF manifestations of experimental meningitis.

Bacteraemia with upper gastrointestinal endoscopy--a reappraisal.

To assess the risk of infective endocarditis for susceptible patients having upper gastrointestinal (GI) endoscopy, we have prospectively studied the incidence, level, duration and source of endoscopy-related bacteraemia and the microorganisms involved. Blood was drawn for aerobic and anaerobic culture and pour plate estimation from 50 patients undergoing upper GI endoscopy. Blood cultures were positive in two patients with the isolation of Streptococci and other oropharyngeal organisms during and 5 min after endoscopy. Pre-endoscopy salivary cultures in both patients grew Streptococci identical to those isolated from the blood; pour plates were negative. We conclude that though the risk of infective endocarditis after upper GI endoscopy is probably very low, the finding of a 4% incidence of transient low-level bacteraemia of oropharyngeal origin argues for antibiotic prophylaxis for susceptible patients.

Protection from infection with Haemophilus influenzae type b by monoclonal antibody to the capsule.

Monoclonal antibodies to Haemophilus influenzae type b were produced by fusing splenic lymphocytes from immunized C57BL/6 mice with the mouse myeloma line P3-X63-Ag8.653. The antibody produced by one hybridoma (5M1H9) bound the capsular polysaccharide, as determined by a radiolabeled antigen-binding assay. The antibody was of the IgM class and was bactericidal in vitro with complement. The protective and therapeutic capacity of antibody 5M1H9 was examined in the infant rat model of H. influenzae type b disease. Antibody (0.1 ml), either undiluted or diluted 1:2, 1:10, or 1:100, administered 4 hr before intraperitoneal injection of 10(4)-10(5) H. influenzae type b organisms protected 100%, 90%,. 55% and 10% of the animals, respectively. To determine the efficacy of antibody 5M1H9 in treating established infection, antibody was given at 24-hr intervals after intraperitoneal injection of bacteria. Delayed administration of antibody 5M1H9 was effective in reducing both the level and incidence of bacteremia.

Enhanced susceptibility of mice infected with murine cytomegalovirus to intranasal challenge with Escherichia coli: pathogenesis and altered inflammatory response.

Three-week-old mice inoculated intraperitoneally with murine cytomegalovirus (MCMV) and then challenged intranasally with Escherichia coli strain K1 demonstrated enhanced mortality (70%-90%) as compared with control animals infected with either pathogen alone (0-20%, P less than 0.05). Mortality was greatest when animals were challenged with E. coli on days 1 or 3 after MCMV inoculation. On day 3 of infection with MCMV, clearance of E. coli from blood and tissues was impaired, and there was a decreased inflammatory response to an E. coli-inoculated sponge implanted subcutaneously (geometric mean of 830 leukocytes in sponge fluid/mm3 in MCMV-infected animals vs. 8,510 leukocytes/mm3 in controls, P less than 0.01). On days 1 and 3 of MCMV infection, decreased leukocyte counts in sponge fluid correlated with increased levels of bacteremia (P less than 0.05). These results indicate that MCMV enhances susceptibility to an intranasal challenge with E. coli. A decrease in the inflammatory response may be one mechanism by which MCMV increases susceptibility to bacterial infections.

Nonspecific enhancement of resistance to local bacterial infection in starved mice.

We studied the effect of muramyl dipeptide (MDP) on enhancing resistance to local bacterial challenge in starved mice. The challenge consisted of the intramuscular insertion of a suture laden with Klebsiella pneumoniae. Mice that had been pretreated with MDP had a statistically lower rate of bacterial recovery at the site of the challenge, had consistently fewer bacteria in the blood, and had improved short-term survival. By enhancing local bacterial containment, MDP pretreatment of immunocompromised animals reduces the level of bacteremia and subsequent mortality.

Comparison of the quantitative direct plating method and the BACTEC procedure for rapid diagnosis of Haemophilus influenzae bacteremia in children.

The efficacy and reproducibility of the quantitative direct plating (QDP) method and the semi-automated BACTEC radiometric system (BBL Microbiology Systems, Cockeysville, Md.) were analyzed for the rapid diagnosis of Haemophilus influenzae bacteremia on the basis of 41 positive cultures from 35 patients. The QDP method detected 61% and BACTEC only 19% of the positive cultures within the first 12 h. Similarly, the QDP procedure yielded growth of H. influenzae in 56% of the cultures before a positive growth index reading was obtained with the BACTEC method. The observation that the quantitative procedure recovered H. influenzae in 88% of all positive cultures, using only 1 ml or less of blood, is attributed to the fact that 68% of the cultures had counts in excess of 100 colony-forming units per ml. The reproducibility of the QDP method was documented by the fact that duplicate blood cultures taken within a few hours of each other yielded comparable results on the number of bacteria at low (1 to 100/ml), moderate (100 to 1,000/ml), and high (greater than 1,000/ml) levels of bacteremia. We concluded that the QDP method is a valuable, simple, and inexpensive supplementary technique to the semi-automated BACTEC procedure for the rapid diagnosis of H.influenzae bacteremia in children.

Vancomycin prophylaxis of experimental Streptococcus sanguis. Inhibition of bacterial adherence rather than bacterial killing.

Using a strain of Streptococcus sanguis tolerant to vancomycin to infect aortic vegetations in rats, we found that prophylactic intravenous vancomycin given 30 min before bacterial challenge decreased the incidence of endocarditis from 88 to 8% (P less than 10(-5)). Because peak vancomycin serum levels were below the minimal bactericidal concentration, mechanisms of protection other than bacterial killing were investigated. S. sanguis were incubated with inhibitory concentration of vancomycin (50 microgram/ml) for 10 h and washed. 85% of rats (73/86) inoculated with control bacteria developed endocarditis, whereas only 42% (33/78) of those inoculated with vancomycin-exposed bacteria did so (P less than 10(-5)). When rats were killed 30 min after bacterial challenge, S. sanguis were detected by culture of the vegetations in 44% of rats injected with control bacteria, but in only 13% of those challenged with vancomycin-exposed bacteria (P less than 0.03). Enhanced clearance of vancomycin-exposed streptococci was not responsible for this protection because blood cultures showed no difference in the level and duration of bacteremia after injection of control or vancomycin-exposed S. sanguis. Moreover, this protection was not abolished in neutropenic rats injected with vancomycin-exposed bacteria, despite more prolonged bacteremia. These results suggest that vancomycin exerted its protection by lowering adherence of tolerant S. sanguis to vegetations rather than through bactericidal activity or enhanced clearance of bacteria by phagocytic cells. In the choice of antibiotics for prophylaxis of endocarditis, reduction of bacterial adhesion may be a criterion as important as bacterial killing.