The aerobic gram-negative bacillary pneumonias

Contemporary issues with bacterial infection in the intensive care unit

Nosocomial infection in the critically ill results from defects in the intrinsic barriers to microbial invasion. The diagnosis is complicated by an inability to perform an adequate physical examination in a patient with several compounding findings, usually necessitating sophisticated technologies to aid in the diagnosis. Pneumonia, line sepsis, urosepsis, sinusitis, endocarditis, peritonitis, and acalculous cholecystitis are the more common infections that challenge the care of the critically ill. Antibiotic therapy is adjunctive to efforts to preserve the barrier, but should be started early, should be targeted as specifically as possible to the offending organisms, and should be dosed adequately to ensure an effective concentration in the infected tissue.

In vivo effects of endotoxin on DNA synthesis in rat nasal epithelium

Airway inflammation in bacterial infections is characterized by the presence of neutrophils and often epithelial injury and repair. Release of endotoxin from bacteria may contribute to these processes. The purpose of this study was to determine the in vivo effects of repeated endotoxin exposure on DNA synthesis in rat nasal epithelium in the presence and absence of neutrophilic influx. Rats were intranasally instilled, once a day for 3 days, with endotoxin or saline (controls). Before the first and third instillations, half of the saline and endotoxin-instilled animals were depleted of circulating blood neutrophils by administering a rabbit anti-rat neutrophil antiserum. Rats were sacrificed 6 or 24 h after the last instillation. Two hours prior to sacrifice, rats were intraperitoneally injected with bromodeoxyuridine (BrdU), an analog of thymidine that is incorporated in the nucleus of cells in the S-phase of the cell cycle. Nasal tissues were processed for light microscopy and immunohistochemical detection of BrdU in nasal epithelial cells. The numbers of nasal epithelial cells, BrdU-labeled epithelial nuclei, and neutrophils per millimeter of basal lamina in the epithelium lining the nasal turbinates in the proximal nasal passages were determined by morphometric analysis. We did not observe a neutrophilic influx in the nasal tissues of neutrophil-depleted rats at 6 or 24 h after the last endotoxin instillation; however, the numbers of nasal epithelial cells and the BrdU-labeling index were significantly increased compared to saline-instilled controls.(ABSTRACT TRUNCATED AT 250 WORDS)

The role of salivary function on oropharyngeal colonization

Aspiration of oropharyngeal flora is the most common route of acquiring Gram-negative pneumonia, a major nosocomial infection. Epidemiologic studies indicate that the elderly are at increased risk for developing these life-threatening pneumonias. The primary objective of this study was to determine the effects of salivary oral defenses against Gram-negative colonization. The study population consisted of 41 male outpatients, age 70 and older, seen at the Denver VAMC. The group included subjects with both diminished and normal salivary flow. Each subject answered a questionnaire regarding overall health, medication use, and symptoms of salivary dysfunction. We then collected whole saliva, unstimulated and stimulated parotid saliva, and performed a throat swab on each patient. For each throat culture, analysis was done to identify Gram-negative bacteria. Flow rates between colonized (n = 6) and noncolonized subjects (n = 34) were compared. While there were no significant differences in the flow rates between the two groups, a trend was noticed in that flow rates were lower for all three flow measures in the colonized group (whole, 22% decrease; unstimulated, 22% decrease; and stimulated, 28% decrease). These preliminary findings suggest that subjects with diminished salivary flow may possibly be at increased risk for oropharyngeal Gram-negative colonization. Research on the role that saliva plays in oropharyngeal bacterial colonization is continuing.

Antibiotic choices in surgical intensive care unit patients

Antibiotics have dramatically changed the care of the critically ill patient over the last 60 years. Patients with complex physiological conditions present with infectious processes requiring the effective use of antimicrobial drugs. In many situations, the inability to eradicate the infectious process is complicated by the progressive development of resistance among the causative organisms. Systemic antibiotic prophylaxis is warranted only for the prevention of wound infections. Regimens in these cases should use large doses of nontoxic antibiotics covering the spectrum of organisms likely to contaminate the wound. The duration of wound prophylaxis should be short, essentially covering only the period of active wound closure; this is usually less than 24 hours. Prevention of most other infections in the ICU depends on the recognition and correction of the various disturbances of host defenses. Topical antibiotic therapy may reduce the level of colonization for a few specific types of infection. Initial empiric antibiotic therapy should be started for clear indications. The antibiotics chosen should be those most likely to be effective against the probable organisms, those which have the lowest toxicity, and those with the smallest likelihood of inducing multiresistance. They must be adjusted promptly based on the microbiologic sensitivities observed. The realization that the physiology of critical illness may alter the normal relations between drug dosages and the tissue antibiotic levels obtained mandates a different approach to the treatment of these patients. The drug volumes of distribution are generally markedly expanded in these patients. Furthermore, these patients require high tissue antibiotic concentrations to improve the chances for successful therapy. Thus, the antibiotics selected must be capable of providing these levels without significant toxicity to the host. Therapy should be continued based on the clinical response observed. Premature cessation of effective therapy often results in relapse.

In vivo effects of endotoxin on nasal epithelial mucosubstances: quantitative histochemistry

Airway inflammation induced by gram-negative bacteria is often characterized by an influx of neutrophils and hypersecretion of mucus. The purpose of this study was to determine how endotoxin, a component of gram-negative bacteria and a chemotaxinogen for neutrophils, affects the amount of stored intraepithelial mucosubstances in the rat nasal airway. Rats were intranasally instilled, once a day for 3 days, with endotoxin or saline (controls). Before the first and third instillation, half of the animals were depleted of circulating blood neutrophils by administering a rabbit anti-rat neutrophil antiserum. Rats were sacrificed 6 or 24 h after the last instillation. Nasal tissues were processed for light microscopy and histochemical detection of stored intraepithelial mucosubstances. The numbers of nasal epithelial cells and intraepithelial neutrophils per millimeter of basal lamina in the anterior nasal septum and the amounts of intraepithelial mucosubstances in the same nasal tissue were determined by image analysis. We did not observe a neutrophil influx in the nasal tissues of neutrophil-depleted rats at 6 or 24 h after the last endotoxin instillation; however, we did observe a significant increase in intraepithelial mucosubstances, compared to saline-instilled controls. In contrast, nonneutrophil-depleted animals had a marked neutrophilic influx and a concomitant decrease in stored mucosubstances, compared to saline-instilled controls. There was no significant difference in the number of nasal epithelial cells per millimeter of basal lamina among any of the experimental groups. These results indicate (1) that endotoxin induces an increase in the amount of intraepithelial mucosubstances only when intraepithelial neutrophils are absent, and (2) that the endotoxin-induced, neutrophil influx probably triggers mucous hypersecretion.

Klebsiella pneumoniae bacteraemia at an urban general hospital

Of 47 patients with Klebsiella pneumoniae bacteraemia admitted to the Hillbrow Hospital, Johannesburg during a period of 18 months, 31 were males and 16 were females. Features predisposing to illness were found in 89.4% patients, chronic alcoholism, neoplastic disease and diabetes mellitus being the most common. Twenty-five infections were acquired in hospital and 22 in the community. Most patients (59.6%) had pneumonia. All isolates of K. pneumoniae were resistant to ampicillin (100%); several (42.6%) were resistant to other antibiotics also. The overall mortality rate was 55.3%. A higher mean initial blood pressure and lower concentrations of serum urea and bilirubin were found in survivors. None of the 28 patients, surviving more than 48 h who received combined therapy with an aminoglycoside and a beta-lactam antibiotic (to which the organism was susceptible) died. Among the remaining patients treated with either an appropriate beta-lactam agent alone, an appropriate aminoglycoside alone or ciprofloxacin the combined mortality rate was 83.3% (P = 0.007).

Nosocomial pneumonia. A multivariate analysis of risk and prognosis

One hundred and twenty consecutive episodes of nosocomial pneumonia (NP) in 118 nonneutropenic adults admitted to a 1,000-bed teaching hospital were studied in order to investigate the prognosis and risk factors. The overall fatality rate was 36.6 percent. The identification of a "high-risk" microorganism (Pseudomonas aeruginosa, Enterobacteriaceae, and other Gram-negative bacilli, Streptococcus faecalis, Staphylococcus aureus, Candida sp, Aspergillus sp, and episodes of polymicrobial pneumonia), bilateral involvement on chest x-ray examination, the presence of respiratory failure, inappropriate antibiotic therapy, and age older than 60 years or an underlying condition ultimately or rapidly fatal were those factors selected by a stepforward logistic regression analysis as independently worsening the prognosis. A series of variables frequently quoted as predisposing to NP was determined to be either present or absent in the same 120 cases of NP and in an equal number of randomly selected control subjects. After adjusting for confounding, factors significantly predisposing to NP were tracheal intubation, depressed level of consciousness, underlying chronic lung disease, thoracic or upper abdominal surgery, prior episode of a large volume aspiration, and age older than 70 years. Since some of the factors influencing the risk or the prognosis of NP are amenable to medical intervention, a percentage of NP might be prevented and its prognosis can be improved.

Penetration of cefotaxime into respiratory secretions

The objective of this study was to quantitate cefotaxime and its active metabolite, desacetyl cefotaxime, in the distal airways and to compare these levels to concentrations in plasma. Respiratory secretions were obtained from the subsegmental level in 17 adult patients undergoing fiber-optic bronchoscopy within 2 h after receiving four doses of cefotaxime (2 g intravenously every 6 h). In 11 patients, cefotaxime levels measured by high-pressure liquid chromatography in bronchial secretions were below detectable limits (less than 0.5 mg/liter); however, levels of desacetyl cefotaxime exceeded 1.5 mg/liter in 9 of these 11 patients (range, 1.6 to 10 mg/liter). Concentrations of desacetyl cefotaxime in lung secretions (6.9 +/- 0.85 [standard error] mg/liter) was 77% of mean levels of desacetyl cefotaxime in plasma (8.9 +/- 1.26 mg/liter). In summary, concentrations of desacetyl cefotaxime in bronchial secretions are markedly higher than those of cefotaxime.

Pneumonia caused by gram-negative bacilli

Gram-negative bacillary pneumonia has become an increasingly important disease in immunosuppressed, elderly, and hospitalized patients. The clinical features, etiologic agents, population at risk, treatment, and outcome in patients with well-documented gram-negative pneumonia were compared in two groups of patients: those with bacteremic pneumonia and those with nonbacteremic pneumonia documented by transtracheal aspiration. Clinical features were frequently subtle in both groups. A wide range of gram-negative bacilli were implicated as pathogens and pneumonias documented by transtracheal aspiration were frequently mixed infections. Pseudomonas aeruginosa and Serratia marcescens were the most common pathogens causing bacteremic pneumonias, whereas Escherichia coli and Klebsiella were more common in the nonbacteremic group. Gram-negative bacillary pneumonia was frequently a lethal disease despite two-drug therapy, particularly in bacteremic patients.

Ceftazidime versus tobramycin-ticarcillin in the treatment of pneumonia and bacteremia

A prospective, randomized study was undertaken to compare the efficacy and safety of ceftazidime with those of a combination of ticarcillin and tobramycin in the treatment of 40 nonneutropenic patients with pneumonia or bacteremia. Altogether, 93% of the patients receiving ceftazidime for pneumonia were cured, and 87% of those with bacteremia responded favorably. Of the subjects who were treated with ticarcillin and tobramycin ceftazidime developed significant superinfection, and one individual treated with the aminoglycoside and carboxypenicillin developed reversible azotemia. Ceftazidime appears to be as efficacious as the ticarcillin-tobramycin combination and is probably safer with regard to oto-and nephrotoxicity; however, superinfections did occur more frequently in the group treated with ceftazidime.

Cefmenoxime efficacy, safety, and pharmacokinetics in critical care patients with nosocomial pneumonia

Nephrotoxicity frequently complicates the use of aminoglycosides in severely compromised acute care patients. Therefore, an open clinical trial was initiated to determine if cefmenoxime alone is useful in serious nosocomial pneumonias. Thirty consecutive patients were entered in the trial, and 28 patients with an average age of 66 years were evaluable. Most were malnourished at entry, with serum albumin averaging 2.8 g/dl and prognostic nutritional index values over 70 percent (normal less than 40 percent). One-half the patients had severe chronic obstructive pulmonary disease and 68 percent required ventilators. Fifty-seven percent had concomitant cardiac disease, and 79 percent had previously been treated with antibiotics. Pneumonia was proven to be present by new infiltrate on chest x-ray, new fever, elevated white blood cell count, and gram-negative rods on gram stain and in cultures of tracheal aspirates or sputum. Patients were given cefmenoxime 1 to 2 g every six hours an average of 12 days. Cefmenoxime peak (one hour) and trough concentrations were measured by high pressure liquid chromatography and averaged 58 and 7 micrograms ml, respectively. Pharmacokinetic data in 18 patients were determined from serum profiles. Gram-positive organisms, Escherichia coli, Klebsiella, and Hemophilus influenzae were usually eradicated. Persistence was noted for some Enterobacter, Pseudomonas, Serratia, and Acinetobacter. Persistence in patients with good clinical response was considered colonization rather than superinfection. Overall, a satisfactory clinical response rate was noted in 78.6 percent of evaluable patients, whereas four patients responded satisfactorily with recurrence and two treatments had an unsatisfactory response. No serious adverse effects were observed. Cefmenoxime is a promising agent in the treatment of susceptible pneumonias in critical care patients.

Experimental bacterial pneumonia in rabbits: polymorphonuclear leukocyte margination and sequestration in rabbit lungs and quantitation and kinetics of 51Cr-labeled polymorphonuclear leukocytes in E. coli-induced lung lesions

A relationship between the circulating and marginal polymorphonuclear leukocyte (PMN) pools was documented using 51Cr-labeled leukocytes as a marker. 51Cr-leukocytes marginating in the lungs were found to decrease following a first-order exponential decline, while 51Cr radioactivity accumulated in the liver and the spleen. Intravenously administered endotoxin caused a rapid selective disappearance of PMNs from the circulation. The percentage of infused 51Cr cells disappearing was equal to the percentage of disappearance of host cells. The PMNs were found to sequester in the lungs, with peak sequestration of labeled cells occurring 5 min after an endotoxin challenge. Over the next 25 min the 51Cr radioactivity in the lungs declined. Large numbers of PMNs, probably newly derived from the bone marrow, were observed histologically to be sequestered in the lung vasculature 90 min after an endotoxin dose, while the early sequestration of circulating leukocytes could not be assessed histologically. Pulmonary inflammatory lesions were induced selectively with Escherichia coli in the left lower lobes of rabbits, leaving the right lower lobes as intrinsic controls. PMN-accumulation into the lesions was quantitated using 51Cr-labeled blood leukocytes. With the aid of 125I-labeled E. coli, a logarithmic dose-response relationship was found between the number of E. coli and of PMNs. Over a 6-hr period circulating PMNs were found to accumulate in a lesion in the left lower lobe, whereas in the control right lower lobe, leukocyte radioactivity declined. These findings were confirmed with the aid of lavages of the right and left lungs. Two peaks of PMN-accumulation were found by studying leukocyte kinetics: a larger peak between 0 and 6 hr and a smaller peak 18-24 hr after instillation of the microorganisms. Histologic studies confirmed the accumulation of leukocytes, and by 3 weeks showed a complete resolution of the lesions.

Escherichia coli pneumonia in the elderly with reference to the role of E. coli K1 capsular polysaccharide antigen

Escherichia coli pneumonia occurred in 17 elderly patients over a two-year period. All cases were documented by transtracheal aspiration. Six of the 17 E. coli organisms causing pneumonia contained K1 capsular polysaccharide. E. coli K1 organisms did not appear to be more virulent than non-K1 strains as measured by incidence of bacteremia, shock and death. E. coli K1 pneumonia was usually community- acquired (five of six cases) and occurred in patients with a prior history of urinary infection. Non-K1 E. coli pneumonia occurred in the hospital setting in eight of 11 patients, and a gastrointestinal source of infection was likely in five patients. All patients were treated with appropriate initial antibiotic therapy, and 71 percent of patients survived.