Corneal ulcer due to Achromobacter xylosoxidans.

We report a case of corneal ulcer caused by the opportunistic organism Achromobacter xylosoxidans which developed during chronic topical steroid treatment of an eye with neovascular glaucoma. A. xylosoxidans has probably been underreported as a cause of ocular infection because of confusion between this organism and other Gram-negative organisms, particularly pseudomonas. A. xylosoxidans is resistant to aminoglycosides and some cephalosporins but not carbenicillin. This difference in antibiotic sensitivity patterns between A. xylosoxidans and pseudomonas makes an accurate differentiation between the 2 organisms important. This case was successfully treated after substituting topical carbenicillin for topical gentamicin and amikacin.

Clostridium difficile. Colonization and toxin production in a cohort of patients with malignant hematologic disorders.

We examined 45 (80%) of 56 consecutive adult patients with malignant hematologic disorders who were hospitalized during a 15-week period at Emory University Hospital, Atlanta. Stool samples for Clostridium difficile culture and cytotoxin assay were obtained on admission and then weekly during each patient's hospitalization. On admission, four patients had detectable C difficile in their stool samples, which was associated with prior antimicrobial use but not with prior cancer chemotherapy. One of the four patients with positive stool samples also had toxin present in the stool sample and was the only one with diarrhea. Eight (36%) of 22 patients hospitalized for one or more weeks had C difficile isolated from at least one stool specimen. The positive cultures showed no clustering in time, and no risk factors were identified for colonization. Only seven of 15 culture-positive stool samples and three of seven toxin-positive samples were associated with diarrhea.

Numbers and types of anaerobic bacteria isolated from clinical specimens since 1960.

Between 1960 and 1974, 826 specimens, excluding stool, urine, sputum, and blood, yielded 689 (83%) positive cultures, of which 403 (58.5%) contained anaerobic bacteria. This represents 48.8% of the total specimens cultured. Isolates from 153 specimens obtained and stocked from 1965 to 1974 were reidentified by current criteria. Gram-negative bacilli, primarily bacteroides, were the most frequently isolated anaerobes, being found in 70% of 153 anaerobe-positive specimens and accounting for 42% of the total anaerobes isolated. Gram-positive cocci were second in occurrence, being found in 66% of 153 specimens and accounting for 40% of the total isolates. Bacteroides fragilis was by far the most frequently isolated species. Compairson of 14 years of cumulative data with data from current studies covering 1- to 2-year periods indicated that the anaerobes isolated from clinical material have not changed significantly in type or relative numbers.

Indirect fluorescent antibody procedure for the rapid detection and identification of Bacteroides and Fusobacterium in clinical specimens.

An indirect fluorescent antibody (IFA) technique was evaluated as a procedure for rapid detection and identification of members of the Bacteroidaceae. Antisera were prepared against 31 members of this family, including species of Bacteroides and Fusobacterium commonly isolated from human infections. The antisera had demonstrated species and/or subspecies specificity. Thirty clinical specimens were studied. Of 13 specimens yielding Bacteroidaceae, for which antisera were available, 23 were presumptively diagnosed by IFA to contain subspecies of B. fragilis and/or Fusobacterium species. Of 17 specimens yielding negative culture results, two were positive by IFA on direct smear. Frequently the in vivo morphology of cells detected in direct smears by this procedure closely mimicked that of cellular debris, tissue cells, and leukocytes. Polyvalent antisera pools facilitated use of the IFA procedure as a practical tool for rapid diagnosis of infections involving the Bacteroidaceae.

Bacteroidaceae in thromboembolic disease: effects of cell wall components on blood coagulation in vivo and in vitro.

The effects of Bacteroides sp., Fusobacterium mortiferum, Bacteroides fragilis, and Sphaerophorus necrophorus on various parameters of blood coagulation in vivo and in vitro were determined and compared to the coagulation effects of Escherichia coli and Salmonella minnesota, wild type and R595. Intravenous injection of washed cells, culture filtrate, lipopolysaccharide, or lipid A of the anaerobic gram-negative microorganisms into mice resulted in acceleration of coagulation. Lipopolysaccharide and lipid A of the anaerobic microorganisms had no apparent effect on circulating platelets in mice or rabbits and did not cause aggregation of human platelets in vitro. Washed cells, lipopolysaccharide, and lipid A of Bacteroides sp. and F. mortiferum also significantly accelerated the clotting time of recalcified platelet poor normal human plasma and C6-deficient rabbit plasma. Lipid A, but not lipopolysaccharide, of E. coli and washed cells of S. minnesota R595 accelerated coagulation by a similar mechanism. These results indicated that Bacteroides sp. and F. mortiferum can accelerate blood coagulation in vivo and in vitro by a mechanism which does not involve platelets or terminal components of complement.

Control of surface wound infection: skin versus synthetic grafts.

Auto-, iso-, or xenografts of skin and synthetics placed on surface wounds freshly contaminated with Pseudomonas aeruginosa stabilizes the wound bacterial population in rats over a 24-h period. When these wounds contained a bacterial contamination established for 24 h prior to grafting, only skin and the synthetic polyhydroxyethylmethacrylate were effective in lowering the initial bacterial concentration. Polyurethane foam and nylon velour were not effective in the established infection model. Skin placed on a contaminated wound for 2 h or longer appeared to equilibrate with the underlying muscle so that the bacterial count per milligram of skin was similar to that of the muscle. It was suggested that this preparation would be useful to obtain an estimate of surface contamination without biopsy of the infected muscle. Skin grafts in place for 2 h significantly lowered the bacterial count in a wound with an established infection. A second decrease occurred between 4 and 24 h after grafting. Histological studies of contaminated and exposed panniculus muscle showed that leukocytes tend to migrate from the muscle surface to its base. Skin grafts and polyhydroxyethylmethacrylate appear to reverse the white cell migration so that the cells move toward the surface of the muscle with preservation of normal staining characteristics in the muscle. It is suggested that this alteration in cell movement after graft application might modify the white cell function and result in a greater bactericidal activity. Apparently, grafts lower bacterial levels in an established infection by modifying the host response to the surface contamination.

Infected surface wound: an experimental model and a method for the quantitation of bacteria in infected tissues.

Methods for the quantitation of bacteria in infected tissues must be rigidly standardized to insure uniformity of results. In this communication we report on a laboratory animal model for the study of surface wound infection and the development of a standardized method for the quantitative estimation of bacteria in infected surface wound tissue by mechanical tissue homogenization and serial dilution. Parallel comparative studies demonstrated that a moist-swab sampling procedure detected only 10% of the bacteria recoverable by a surface-wash procedure. Either tissue homogenization or surface-wash procedures recovered significantly more bacteria from contaminated surface wounds than were obtained by surface-swab sampling techniques.