A concise biotyping system for differentiating strains of Escherichia coli.

A six test biotyping system comprising fermentation of dulcitol, sorbose, raffinose, and 5 ketogluconate, motility and production of beta-haemolysis was used to obtain biotype profiles for 514 strains of Escherichia coli isolated from the urinary tract. This profile was suffixed to the API-20E code, recorded when the strains were originally identified. An expanded and reliable biotyping system was thus created giving a greater number of possible biotypes than with either system alone. The sensitivity in terms of distinguishing the organisms studied was therefore greatly increased. The use of O-serotyping in combination with biotyping is discussed. Biotyping can also be usefully supplemented by the determination of the eight commonest E coli O-serotypes. This is of value in many clinical situations where differentiation of organisms is vital to the proper analysis of results.

Resistance to trimethoprim in 1978-79 compared with 1973-75.

The incidence of resistance to trimethoprim among urinary isolates between October 1978 and November 1979 was 11.5%, more than double the figure found April 1973 and October 1975. Of the resistant strains, 60% had a minimum inhibitory concentration in excess of 1 mg/ml. Escherichia coli and Proteus mirabilis showed the greatest increase in resistance since the previous study. Rather wide fluctuations occurred in the incidence of resistance for various species when the figures were analysed over two-month periods, hence studies of short duration must be interpreted with caution. There are not yet enough data from this or other studies for the cause of the increased incidence of resistance to trimethoprim to be determined.

A comparison of the in vitro activity of metronidazole, tinidazole, and nimorazole against Gram-negative anaerobic bacilli.

The in vitro activities of metronidazole, nimorazole, and tinidazole were compared against 69 strains of obligately anaerobic Gram-negative bacilli. Geometric mean MICs were 0-34, 1-05, and 0-28 mug/ml respectively. Thirty-six strains were also tested by the disk method. Correlation between MIC and diameter of the zones of inhibition was poor.

Recurrent bacteriuria and primary biliary cirrhosis: ABO blood group, P1 blood group, and secretor status.

Patients with primary biliary cirrhosis have an abnormally high incidence of urinary tract infection (35%). Susceptibility to urinary infection and other infectious diseases has been linked with certain blood group antigens and secretor status. We have therefore studied these characteristics in patients with primary biliary cirrhosis. We were unable to show any abnormal distribution in blood groups or secretor status in patients with primary biliary cirrhosis (compared with a normal population) which might reflect their predisposition to urinary infection. The distribution of blood groups and secretor status in patients with primary biliary cirrhosis with a history of urinary infections was not significantly different from patients without such a history. Escherichia coli strains isolated from patients with primary biliary cirrhosis did not bind in any greater numbers to the uroepithelial cells of primary biliary cirrhosis patients than to the cells of a normal healthy control. We therefore conclude that blood group distribution, abnormal secretor status, and epithelial cell type are not important factors in the predisposition of primary biliary cirrhosis patients to urinary infections.

Phenomenon of resistance to Augmentin associated with sensitivity to ampicillin: occurrence and explanation.

Ampicillin was found to be some tenfold more active than amoxycillin against Enterobacter cloacae. This finding explains the observation that some Ent cloacae strains are sensitive to ampicillin in the disc test but resistant to Augmentin. Ampicillin was also found to be more active than amoxycillin against Citrobacter freundii and Serratia marcescens. In view of these findings, the practice of using ampicillin discs to predict sensitivity to amoxycillin should be reconsidered. The use of both ampicillin and amoxycillin discs is appropriate if errors are to be avoided.

Evaluation of the Microcult system for isolating and identifying Neisseria gonorrhoeae.

Specimens from 95 patients attending a venereal diseases clinic were examined for gonococci by three methods--a conventional culture technique using modified Thayer-Martin medium, microscopy of a Gram-stained direct smear, and the Microcult system. For 56% of the specimens the results by all three methods agreed. Assuming the results obtained by culture on Thayer-Martin medium to be correct, the largest source of error was due to false-positive results: microscopy gave 26 and Microcult gave 15 such results. False-negative results were less common: Microcult gave 14, microscopy six. Microcult gave positive results more quickly than the conventional Thayer-Martin cultural method, but the gonococci were difficult to isolate by subculture from the Microcult culture pads. The Microcult medium was not absolutely specific for Neisseria gonorrhoeae. Nevertheless, the Microcult test may well prove to be a useful adjunct to the diagnosis of gonorrhoea, especially when laboratory facilities are not readily available.

Antibacterial oxazolidinones. In vitro activity of a new analogue, E3709.

The oxazolidinone compound E3709, which contains a 4-pyridyl group, was found to be more active in vitro than other members of this series, such as DuP 721. MIC90 for staphylococci(including methicillin-resistant isolates), streptococci (including Enterococcus faecalis), Clostridia, and diphtheroids was less than 0.5 micrograms/ml. Haemophilus influenzae, Moraxella catarrhalis, and Bacteroides fragilis were less susceptible, with an MIC90 between 2 and 8 micrograms/ml. E3709 MICs of Gram-negative species ranged from 100 to greater than 1000 micrograms/ml. At a concentration of 10 micrograms/ml, E3709 was bactericidal for selected Gram-positive species. A postantibiotic effect of 3 hr was observed against staphylococci. Resistance to E3709 was not detected.

Importance of methodology in determining bactericidal and bacteriostatic activities of azlocillin and ticarcillin against Pseudomonas aeruginosa.

The activities of azlocillin and ticarcillin against Pseudomonas aeruginosa were compared by estimating minimum inhibitory and bactericidal concentrations (MIC and MBC) in liquid and solid media, and by constructing killing curves from sequential viable counts. In MIC studies, azlocillin was about three times more active than ticarcillin in solid medium (agar dilution test) and in liquid media (tube and microdilution tests). When the MBC was measured, however, results varied according to the technique used. On agar and in microdilution tests, both azlocillin and ticarcillin were bactericidal, the MBC being 1.3-3 MIC. In the tube test, the MBC for ticarcillin was again about 3 MIC, but azlocillin appeared not to be bactericidal (MBC greater than 1 mg/ml). However, sequential viable counts of four clinical isolates showed that at 4 MIC both antibiotics reduced viable counts by a factor of 10(4) in 8 h. Our results stress the importance of methodology when assessing the antibacterial activity of an antibiotic.

Pre-opsonisation of Escherichia coli induces resistance to neutrophil killing in serum and urine: relationship to growth phase.

Tests of phagocytosis and killing by polymorphonuclear neutrophil leucocytes (PMNL) are usually done with pre-opsonised organisms. Phagocytosis of 11 strains of Escherichia coli, pre-opsonised, and in the stationary phase, resulted in the killing of only one strain although all the organisms were phagocytosed. However, when the same strains were added unopsonised to a PMNL-serum mixture, eight were killed after phagocytosis. With two of these strains, the amount of killing was inversely proportional to the time of pre-oposonisation. E. coli incubated for 30 min in dilute peptone water in Hanks's Balanced Salts Solution before phagocytosis also became resistant to killing; bacterial division did not occur during this period. Experiments with bacteria in urine confirmed these findings and showed that E. coli exposed to serum or urine before phagocytosis became resistant to killing by PMNL. E. coli rapidly changes its sensitivity to phagolysosome killing during transition from stationary to lag phase in a nutrient medium. This resistance is retained through the exponential phase but is lost during the stationary phase. The killing of Pseudomonas, Enterobacter, and Acinetobacter by PMNL was unaffected by varying the method of opsonisation or the phase of growth. If this phenomenon occurs in vivo it may affect the outcome of infections caused by strains of E. coli that survive killing by PMNL.

Relationship between adhesion of Escherichia coli to uro-epithelial cells and the pathogenesis of urinary infection: problems in methodology and analysis.

Escherichia coli strains isolated from patients with urinary infections were tested for their ability to adhere to human uro-epithelial cells. In any single experiment, the numbers of bacteria adhering to individual uro-epithelial cells showed great variations; some cells had hundreds of bacteria adhering to them whereas other cells had few or none. This non-Normal distribution of bacterial attachment must be taken into account when carrying out statistical analyses of the results. The wide discrepancies reported in the literature regarding bacterial adhesion to uro-epithelial cells must, in part, be related to the type of statistical analysis used. In many cases, a Normal rather than a non-Normal distribution has been assumed. We found that even when all variables were kept constant, the experiment was still not reproducible. Therefore the technique shows a high degree of both inter- and intra-experimental error. Adhesion depended on such factors as the type of fimbriae produced by the bacteria, differing viability of uro-epithelial cells and varying pH of the medium used for a particular experiment. It is concluded that the results of in-vitro experiments demonstrating adhesion of E. coli to uro-epithelial cells are difficult to relate to bacterial adhesion in vivo but better results could be obtained if more attention were paid to standardisation of methods and their analysis.

Anaerobic periurethral flora of healthy women and women susceptible to urinary-tract infection.

The anaerobic periurethral microbial flora of 25 healthy women was compared with that of 29 women attending the urinary-tract-infection clinic at the Royal Free Hospital. The latter group consisted of 19 patients receiving long-term prophylactic antimicrobial therapy and 10 with proven recurrent urinary-tract infection not receiving prophylactic treatment. The numbers and species of anaerobes isolated from each group were similar. Lactobacillus spp. were the most frequently isolated organisms in each group and the most numerous. Bacteroides spp. were the next most frequently isolated. In any one subject, the anaerobic flora varied considerably during the study period of approximately 6 months. Thus, the anaerobic flora is not affected by recurrent urinary-tract infection in the past nor by the use of prophylactic chemotherapy. It does not appear to exert a protective role against the initiation of urinary-tract infection.

Effect of bacterial products including endotoxin on neutrophil function in infected urine.

Experiments were performed to determine the effects of products of bacterial growth (including endotoxin) on phagocytosis and intracellular killing by polymorphonuclear leucocytes (PMNL) in urine. Bacteriologically filtered supernates of two strains of Escherichia coli grown in urine were added in varying amounts to mixtures of PMNL and E. coli, also in urine. Phagocytosis of the two strains was reduced from > 90% in controls to 66% and 48%, respectively, in the presence of undiluted culture filtrate (containing endotoxin 2-2.5 micrograms/ml). Intracellular killing was also decreased and was abolished by dilutions corresponding to endotoxin concentrations of 0.6 and 0.75 micrograms/ml. When PMNL exposed to these inhibitory dilutions were resuspended in fresh urine, their phagocytic ability was fully restored and 13-24% of their killing activity was regained. A minimum concentration of commercially purified E. coli endotoxin of 200 micrograms/ml was required to abolished PMNL killing, with phagocytosis uninhibited. The results strongly suggest that bacterial growth metabolites, not endotoxin, are responsible for the depression of phagocytosis and intracellular killing in infected urine. A moderate dilution of the bacterial products in urine permits good PMNL function. Extrapolating this to the clinical situation, diluting the urine by water loading (as recommended for patients with urinary infections) should ensure efficient activity of PMNL under in-vivo conditions providing urinary pH and osmolality are not adversely affected.

Teicoplanin vs. vancomycin for the treatment of serious infections: a randomised trial.

A prospective, randomised study of 56 patients comparing teicoplanin with vancomycin for suspected or proven severe Grampositive infection was conducted. The majority of infections were soft tissue infections (8 teicoplanin; 16 vancomycin) and by chance a significantly higher number of Hickman catheter-related infections occurred in the vancomycin arm (4 vs. 14, P < 0.01). Teicoplanin was administered as a single daily dose of 400 mg iv or im; 5 patients received 200 mg following the initial dose of 400 mg. Vancomycin was given 1 g every 12 h. Fifty-four patients were evaluable for efficacy (26 teicoplanin, 28 vancomycin). Of these, 18 episodes in 17 patients (teicoplanin) and 19 episodes in 18 patients (vancomycin) gave an evaluable clinical response, the success rates being similar (76% teicoplanin; 68% vancomycin). Staphylococcus aureus was the most common pathogen isolated; all pathogens were susceptible to both glycopeptides with MICs < 4 mg/l. Bacteriological elimination rates were similar in both groups (71% teicoplanin; 78% vancomycin). Significantly more patients given vancomycin experienced adverse events (7 teicoplanin; 16 vancomycin; P = 0.03). This caused treatment to be discontinued in 4 cases, compared with only one receiving teicoplanin. The most common vancomycin-related events were histamine-associated reactions (15 patients), including 2 cases of Red Man Syndrome, and nephrotoxicity (5 patients). There were no histamine-mediated events and only one case of nephrotoxicity with teicoplanin. Teicoplanin and vancomycin show similar clinical and bacteriological efficacy and teicoplanin is significantly less toxic and easier to use in patients with severe infection.

Clinical study of cefuroxime in the treatment of lower respiratory tract infections.

Twenty-three hospital in-patients with severe lower respiratory tract infections were treated with cefuroxime sodium. The drug was given intramuscularly in a dose of either 750 mg or 1000 mg at 8-hourly intervals for 5 days. Of the 21 patients who could be assessed, the response to treatment was highly satisfactory and there were no treatment failures. Eight patients had failed to respond to a course of oral antibiotics before being seen. Most of the patients were elderly and all were very ill. The sputum became mucoid in all but 1 patient. There was no change in tests of liver or renal function. Cefuroxime appears to be an effective and well-tolerated drug for the treatment of patients with severe chest infections.

Detective work in continuous ambulatory peritoneal dialysis.

We report five cases of continuous ambulatory peritoneal dialysis in which the mechanisms and sources of infection were established. We show how diligent enquiry and environmental investigation can explain the pathogenesis of infection and help in prevention by motivation of the patient.

Reassessment of the rationale for the combinations of sulphonamides with diaminopyrimidines.

Trimethoprim is the best known inhibitor of bacterial dihydrofolate reductase. Initially, this was always combined with sulphamethoxazole. It was later combined with other sulphonamides (eg. sulphamoxole, sulphadiazine or sulfametopyrazine), but the sulphonamide moiety as a contributor to clinical efficacy was increasingly questioned. Thus, in 1979 (in UK) trimethoprim alone was introduced. Justification for the combination was based on: (a) synergy occurs in vitro (b) bactericidal activity, while the two components are bacteriostatic (c) the emergence of resistance was claimed to be lower. However, these claims were not substantiated by studying the microbiological and pharmacokinetic properties of trimethoprim and the sulphonamides, but most importantly by the results of clinical trials. These show that in most indications, trimethoprim alone is as good, cheaper and causes considerably fewer adverse events than use with a sulphonamide. For urinary infections most agree that monotherapy is best. In respiratory infections diaminopyrimidines have relatively poor activity against important pathogens, eg pneumococci and especially Moraxella catarrhalis. It could be argued in these case that the addition of a sulphonamide may increase therapeutic efficacy. This can only be resolved by large clinical trials. In brucellosis and gonorrhoea, where sulphonamides are more microbiologically active than diaminopyrimidines, it is likely that combination with another antibiotic is needed. However, too much reliance must not be placed on extrapolating from trimethoprim to other diaminopyrimidines; conclusions should be drawn from work using the other compounds.

Limitations of and indications for the use of co-trimoxazole.

Co-trimoxazole is still widely used for indications where trimethoprim alone is equally effective. Microbiological and pharmacokinetic considerations reveal that trimethoprim alone provides adequate anti-microbial activity for treatment of conditions for which co-trimoxazole is often given. Synergy may be shown in vitro, but in clinical practice is an unusual occurrence. There is no evidence from clinical studies that the sulphonamide moiety fo co-trimoxazole prevents the development of resistance to trimethoprim. The adverse event profile of co-trimoxazole is a summation of that of sulphonamide and of trimethoprim. Thus, using trimethoprim alone should reduce both the incidence and potential severity of adverse events seen when co-trimoxazole is used. Clinical trials have shown trimethoprim to be as effective as co-trimoxazole in many of the common bacterial infections of the urinary and respiratory tracts. However, there are a few specific varieties of infection for which co-trimoxazole can be shown to be superior to trimethoprim: these include toxoplasmosis, brucellosis, nocardiosis, chancroid and pneumonia caused by Pneumocystis carinii. For many common infections, scientific, rational, economic and clinical reasons dictate that trimethoprim is preferable to co-trimoxazole.

Combinations of sulphonamides with diaminopyrimidines: how, when and why?

Co-trimoxazole is still widely used for indications where trimethoprim alone is equally effective. The pharmacological rationale of the combination of trimethoprim and sulphamethoxazole involves synergistic action of the two drugs. This is true only from a laboratory point of view; several considerations have led to the conclusion that the synergism between the two components is of only in vivo marginal importance in determining the clinical efficacy of co-trimoxazole. This is due to a greater tissue affinity of trimethoprim compared to that of sulphamethoxazole and, therefore, to the different tissue concentration ratios obtained in vitro and in vivo. Another claim sustaining the combination is the prevention of developing resistance to trimethoprim; however, there is no substantial clinical evidence to support this claim. It does seem likely that trimethoprim has protected against the emergence of that sulphonamide resistance. This slight benefit is outweighed by the disadvantages of the combination, mainly consisting of the occurrence of adverse events due to the sulphonamide moiety. Consequently, the incidence and severity of the adverse events seen with co-trimoxazole should be reduced by using trimethoprim alone. There are only a few cases where co-trimoxazole is better than trimethoprim: toxoplasmosis, brucellosis, nocardiosis, chancroid and pneumonia due to Pneumocystis carinii. For the other and many common infections, scientific rationale, economic and clinical reasons dictate that trimethoprim is superior to co-trimoxazole.

Cefaclor into the millennium.

We review the discovery and development of the cephalosporins and subsequently cefaclor. Cefaclor is active against a wide range of commonly encountered bacterial pathogens, acting by inhibiting cell wall synthesis. Its in vitro activity compares favourably with other beta-lactam antibiotics. Its pharmacokinetic properties indicate that an 8-hourly dosing schedule is appropriate. In addition a delayed release formulation allowing twice daily dosage has been developed. The efficacy of both formulations of cefaclor has been verified by many clinical trials. Cefaclor has been widely used in infections of the respiratory tract (including otitis media), urinary tract and soft tissues. The results of therapy are summarized. The low incidence of adverse events is highlighted and the beneficial influence of this on compliance is described. Finally, the pharmaco-economics of cefaclor are considered.

A review of the antimicrobial activity of the fluoroquinolones.

The evolution of the fluoroquinolones is described, and structure-activity relationships outlined. The in-vitro antimicrobial activities of ciprofloxacin, enoxacin, norfloxacin, ofloxacin and pefloxacin against a wide range of organisms are critically reviewed. In-vitro factors influencing fluoroquinolone activity are discussed. Reports of the acquisition of resistance to the fluoroquinolones are evaluated. Finally, possible future directions for this group of antibiotics are discussed.