COLONIZATION AND BRONCHOPULMONARY INFECTION

Epidemiology of Klebsiella bacteraemia: a case control study using Escherichia coli bacteraemia as control

Epidemiological data from 117 episodes of Klebsiella bacteraemia were compared with those from matched controls with Escherichia coli bacteraemia. Cases and controls were obtained from 20,631 blood cultures taken from patients in Hvidovre Hospital between 1990 and 1992. The data studied included: sex and age, risk factors, portal of entry, outcome, nosocomial acquisition and distribution within the hospital. The incidence of Klebsiella bacteraemia was 9.3/10,000 admissions (76% Klebsiella pneumoniae; 24% Klebsiella oxytoca). Patients with Klebsiella and E. coli bacteraemia had many common features, including a high incidence of neoplastic disease, biliary tract disease, and renal failure. Many had undergone surgery or received therapy with steroids, antacids or antibiotics. Klebsiella bacteraemia was more often found in males, in patients with hospital contact within the previous month, and polymicrobial infection. Logistic regression analysis showed that use of invasive plastic devices and diabetes were significantly associated with Klebsiella bacteraemia. The urinary tract was the commonest source, followed by the biliary tract; 27% of patients had no obvious focus of infection, and in many of these an invasive device may have been involved. Forty-five K-serotypes were found--the largest number being nine strains of type K3; only a few strains had acquired resistance characters to antimicrobial agents. There were no differences between community- and hospital-acquired strains; indicating that our hospital does not have a resident strain of Klebsiella.

Adherence properties of an mrkD-negative mutant of Klebsiella pneumoniae

The role of the mrkD gene in attachment by a type 3 fimbriate Klebsiella pneumoniae strain was further characterized. A clinical isolate, K. pneumoniae IA565, was found to contain two copies of the gene encoding the fimbrial subunit, mrkA, and one copy of the gene encoding the adhesin subunit, mrkD. One copy of mrkA was located on the bacterial chromosome, and the other copy was associated with mrkD and located on a plasmid. The plasmid-borne mrk gene cluster was lost when K. pneumoniae IA565 was subcultured serially in broth at 44 degrees C. The resulting mrkD-negative strain, designated K. pneumoniae IApc35, did not exhibit the following adherence characteristics associated with K. pneumoniae possessing MrkD-positive fimbriae: agglutination of tannic acid-treated human erythrocytes and attachment to trypsinized human buccal cells. However, K. pneumoniae IApc35 produced type 3 fimbriae that were composed of the characteristic 21.5-kDa major fimbrial subunit, were reactive with specific serum, and were visualized specifically by immunoelectron microscopy. K. pneumoniae IApc35 retained a copy of the mrkA gene on its chromosome. This mrkA-containing gene cluster could be complemented by a recombinant plasmid carrying only the mrkD gene, resulting in restoration of the K. pneumoniae IA565-like adhesive phenotype and demonstration of type 3 filament-associated MrkD subunits by using colloidal gold labeling and immunoelectron microscopy. These data indicate that K. pneumoniae may contain multiple copies of the mrk genes which may be present simultaneously on both plasmid and chromosomal DNAs and which may encode fimbriae with different binding specificities.

Pneumonia in the intensive care unit setting

Nosocomial pneumonia is a common and serious occurrence in the ICU. It most often results from aspiration of oropharyngeal secretions that have become colonized with pathogenic enteric gram-negative bacilli. Colonization occurs in association with acute and chronic illness and particularly with therapy that includes nasogastric or endotracheal tubes, H 2 blocking antacid drugs, or antibiotics; aspiration is increased by anesthesia, sedative drugs, and upper airway instrumentation. The diagnosis of ICU-acquired pneumonia is complicated greatly by the nonspecificity of clinical and laboratory data, and the difficulty in distinguishing the organisms producing infection from those merely colonizing the airway when using routine culture techniques. Among specialized diagnostic techniques, quantitative culture of specimens obtained with the protected sampling brush offers the most promise in establishing a specific microbacteriologic diagnosis of nosocomial pneumonia. Empirical treatment with broad spectrum antibiotics is frequently necessary when a specific diagnosis cannot be made. The poor outcome associated with nosocomial pneumonia, regardless of treatment, suggests that methods to prevent dissemination and oropharyngeal colonization of the offending organisms should be emphasized.

Fimbrial types among respiratory isolates belonging to the family Enterobacteriaceae

Bacterial attachment is believed to be an early step in gram-negative nosocomial pneumonia. The frequency of fimbria-associated adhesins among respiratory pathogens has not been studied in detail. In this study isolates belonging to the family Enterobacteriaceae, prospectively obtained from intensive care unit patients who were suspected of having nosocomial pneumonia, were examined for fimbria-associated adhesins. Type 3, P, type 1, and other fimbrial phenotypes were identified by specific hemagglutination and electron microscopy. The Klebsiella type 3 fimbrial phenotype was further characterized by using a monoclonal antibody. Also, both type 3 and Escherichia coli P fimbrial genotypes were detected by using DNA colony blot assays. The frequencies of genera or species isolated were as follows: Enterobacter (38.6%), Klebsiella (26.8%), Serratia (17.7%), E. coli (13%), and Proteus (5.2%). Isolates of Klebsiella oxytoca, K. pneumoniae, and Enterobacter cloacae most commonly possessed the type 3 fimbrial phenotype and genotype. The phenotype and genotype for E. coli P fimbriae (46.2 and 50%, respectively), a known pathogenic determinant in the urinary tract, were detected more frequently than expected. In addition, a previously unspecified hemagglutinin that was specific for porcine erythrocytes was almost uniformly expressed among isolates of Enterobacter aerogenes. Finally, the expression of the type 1 fimbrial phenotype was widely detected among the isolates tested but notably absent among K. oxytoca and Proteus mirabilis isolates. The frequency of the various fimbrial types identified suggests a role for these bacterial organelles in adherence to respiratory epithelia.

Risk factors for nosocomial pneumonia after coronary artery bypass graft operations

We attempted to determine risk factors for nosocomial pneumonia in patients undergoing a coronary artery bypass graft operation. We reviewed the microbiology and medical records for any patient with a sputum culture who had undergone a coronary artery bypass graft operation in 1988 to identify patients with pneumonia according to a standard clinical definition. We found 19 cases of pneumonia through our initial review; complete medical records were found on 15 cases. Gram-negative bacilli predominated as the most common etiologic agent causing pneumonia in this cohort. There were no clusters noted. Mortality was 26.6%. Pneumonia occurred approximately 4 days after the operation. Thirty-six controls were randomly selected from patients undergoing coronary artery bypass graft operations in 1988. Logistic regression analysis revealed that a history of chronic obstructive lung disease, duration of more than 2 days of mechanical ventilation after operation but before diagnosis of pneumonia, and receipt of gastric acid inhibitors (antacids or H2-blockers) were independent risk factors for nosocomial pneumonia. Only the last risk factor was amenable to intervention at the time of operation.

Adult respiratory distress syndrome in pediatric patients. II. Management

Adult respiratory distress syndrome, a clinical syndrome of respiratory failure that follows many kinds of insults, often in patients with no previous pulmonary disease, occurs in pediatric patients. This group of disorders has a typical clinical, pathologic, and pathophysiologic course, the hallmark of which is injury to the alveolar-capillary membrane with increased permeability of the pulmonary vasculature and pulmonary edema. Resolution may occur at any stage, but most patients die and many develop chronic lung disease requiring respiratory support for weeks or months. Multiple organ system failure, secondary infection, and irreversible respiratory dysfunction are responsible for the poor outcome. The underlying mechanisms that relate injury to the development of pulmonary disease are unclear. In some cases there may be direct injury to the lung, but in others, such as septic shock, there are mediators that link the initial insult to the subsequent lung injury. The leukocyte may have a central role in this process, although this is uncertain. Therapeutic measures needed to support the patient, especially increased inspired oxygen, are additional factors in the progression of lung disease. Current therapy, as summarized in Table II, is primarily supportive. Efforts to treat ARDS after it is clinically apparent have been disappointing. The pathogenic mechanisms that lead to ARDS are probably well advanced by the time the syndrome is diagnosed on the basis of the usual clinical signs. Therefore an emphasis on understanding the mechanisms of lung injury so that specific markers can be used to predict which patients will develop ARDS, allowing intervention in the early stages of the process, may prove rewarding.

Enteral nutrition in patients receiving mechanical ventilation. Multiple sources of tracheal colonization include the stomach

Nutritional therapy of patients receiving mechanical ventilation includes enteral feeding. To determine the frequency of gastric microbial colonization in patients receiving enteral nutrition, 18 patients with acute respiratory failure receiving ventilation were studied. Multiple sources of tracheal colonization were then evaluated to determine the frequency of tracheal transmission of gastric flora and the relationship of gastric colonization and transmission to nosocomial respiratory infection. The stomach was colonized in every patient who received enteral feeding. Tracheal colonization occurred in 89 percent (16 of 18) of patients. In 12 of these 16 patients, 14 organisms colonized the trachea after transmission from the stomach and/or oropharynx. Thirty-six percent (five of 14) of organisms transmitted to the trachea originated from the stomach, whereas another 36 percent (five of 14) were first recovered from the oropharynx. Four tracheally transmitted organisms were first recovered simultaneously from the stomach and oropharynx. Nosocomial respiratory infection developed in 11 patients (63 percent). Three infections occurred during the six-day study period, one of which was associated with transmission of a gastric organism. Multiple sources of tracheal colonization occur in patients receiving enteral nutrition. The stomach is an important source of tracheal colonization. Enteral nutrition can be associated with gastric flora colonizing the trachea and causing nosocomial respiratory infection.

A germ-free status does not protect from the lethal effects of acute lung damage caused by O,S,S,-trimethyl phosphorodithioate

To investigate whether a normal resident microbiological flora of conventional rats influences the lethality of chemical-induced lung damage, the pneumotoxin O,S,S-trimethyl phosphorodithioate (OSSMe, 75 or 100 mg/kg, s.c.) was administered to age-matched conventional and germ-free male F344 rats. Microbiological and serological examinations confirmed the germ-free state of the germ-free rats and showed that no specific lung pathogens were present in the conventional rats. As in conventional rats, clinical symptoms and death of OSSMe-treated germ-free rats resulted from respiratory failure. The germ-free rats were not more resistant, but rather more susceptible to OSSMe than conventional rats. Increases in lung weight and histological examination of lung tissue 3 days after dosing with OSSMe (75 mg/kg, s.c.) showed no differences between germ-free and conventional rats. Despite alterations in their nasopharyngeal flora, death in the conventional rats was probably not caused by bacterial superinfection. The higher susceptibility of germ-free rats to OSSMe can be partly attributed to pharmacokinetic differences, since plasma levels of OSSMe decreased more slowly in germ-free than in conventional rats. It is concluded that germ-free rats are not protected from the lethal consequences of acute chemical-induced lung damage.