Gram-negative pathogens in septicaemic infections.

For the last two decades, gram-negative rod bacteremia has been the major infectious disease problem developing in patients hospitalized in the United States and many European countries. Mortality rates are high and have exceeded 30% in patients with serious underlying diseases. The most commonly isolated organisms are Escherichia coli, Klebsiella-Enterobacter-Serratia, and Pseudomonas aeruginosa but the latter organism has consistently been associated with the highest case fatality ratios. Stool sampling surveys in normal individuals reveal low carriage rates of both K. Pneumoniae and Pseudomonas, but prospective studies in neutropenic patients using the surveillance culture technique has indicated a higher prevalence of gastrointestinal colonization. In neutropenic patients asymptomatic colonization may occur weeks prior to onset of gram-negative rod bacteremia. The patients most likely to develop systemic infection are neutropenic subjects who have P. aeruginosa colonizing the gastrointestinal tract. Microbial pathogenetic factors associated with gram-negative bacillary infections are poorly understood. Antigens such as the O and K antigens appear to confer the property of resistance to phagocytosis. Both antibody deficiency and restricted activation of the alternative complement pathway are responsible for impaired phagocytosis of some bacteremic strains in both fresh normal human serum as well as in the serum or plasma of bacteremic patients. A number of studies have shown a direct correlation between type-specific and cross-reacting serum antibody levels with survival. Additionally, Exotoxin A of P. aeruginosa is a likely virulence factors of this species and levels of humoral antitoxic antibodies measured at onset of bacteremia correlate with improved prognosis. Antibodies to capsular antigens like K-1 antigen of E. coli enhance phagocytosis of strains with different O types. Despite a substantial amount of circumstantial evidence, a role for circulating toxins or other candidate virulence factors has not bee convincingly established in human sepsis. Bioassays like those involving the limulus lysate gelation have identified some patients with endotoxemia, but this test may be negative if circulating immune complexes are present. Additionally, other factors in serum may inactivate endotoxins of complex with endotoxins. Several human and experimental studies suggest that prophylaxis or treatment of gram-negative infections may be effective with antiserum against endotoxins antigens. The high mortality from gram-negative bacteremia in some patients strongly justifies further research in this area.