Non-antibiotics reverse resistance of bacteria to antibiotics.

BACKGROUND: Most clinical isolates that exhibit a multi-drug resistant phenotype owe that resistance to over-expressed efflux pumps. Compounds that are efflux pump inhibitors (EPIs) reduce or reverse resistance to antibiotics to which the bacterial strain is initially resistant. We have evaluated non-antibiotics to reduce resistance of commonly encountered bacterial pathogens to antibiotics. MATERIALS AND METHODS: The effect of non-antibiotics on the susceptibility of bacteria to antibiotics was conducted by minimum inhibition concentration determinations of the antibiotic in the absence and presence of the non-antibiotic. RESULTS: Non-antibiotics such as chlorpromazine, amitryptiline and trans-chlorprothixene are shown to reduce or reverse resistance of a variety of bacteria to antibiotics. CONCLUSION: The results suggest that non-antibiotics may serve as adjuncts to conventional antibiotics for the therapy of problematic antibiotic infections caused by bacteria that owe their resistance to over-expressed efflux pumps.

[Resistance of gram-negative bacteria to beta-lactams, aminoglycosides and quinolones]. / Gramnegative bakteriers resistens over for beta-laktamer, aminoglykosider og kinoloner.

Antibiotic treatment of infections has traditionally been restrictive, in that use of broad-spectrum antibiotics has been avoided as far as possible. A change in therapeutic principles has occurred in recent years for example with respect to treatment of patients that are immunocompromised, suffering from multiple traumas, recovering from major surgery or have implanted foreign bodies. The use of antibiotics has become more complicated, as one more frequently has to consider the possibility of development of antibiotic resistance in different bacterial species because of long-term treatment regimes in such patients. However, it is important to avoid the spread of broad-spectrum antibiotic regimes such as those for neutropenic patients to less vulnerable patient groups, where narrow-spectrum therapy is usually sufficient. In later years an increased resistance to broad-spectrum antibiotics among enterobacteria has been described abroad. Resistance mechanisms for the most commonly used intravenous antibiotics (beta-lactams, aminoglycosides and quinolones) are described. Resistance mechanisms for the individual antibiotics, cross-resistance and the most important epidemiological aspects are discussed. It is suggested that surveillance of antibiotic use and development of resistance as well as identification of resistance mechanisms be increasingly used so as to continuously ensure optimum treatment regimes.

The antibacterial activity of a siderophore. 3. The activity of deferoxamine in vitro and its influence on the effect of antibiotics against Escherichia coli, Proteus mirabilis and coagulase-negative staphylococci.

The in vitro activity of deferoxamine (DFO) both per se and in combination with the reductant ascorbic acid (AA) was determined against 10 E. coli strains, 5 P. mirabilis strains, and 10 coagulase-negative staphylococci. In terms of interaction, the influence of DFO on the activities of cephalothin and gentamicin was furthermore investigated against the same panel of strains employing a macrobroth dilution technique and killing-curve kinetics. The MICs of cephalothin and gentamicin were lowered for one half of the strains. Moreover, DFO prolonged the generation times of logarithmic growth phase considerably, especially when the reductant AA was present. The interactions between DFO or DFO+AA and subinhibitory concentrations of antibiotics were established by the application of growth constants, and resulted in synergy for 15 out of 25 strains with cephalothin and 9 out of 25 strains with gentamicin.

Impact of the agar medium and disc type on disc diffusion susceptibility testing against teicoplanin and vancomycin.

Susceptibility to teicoplanin and vancomycin was assessed by three disc types: two commercially available discs (NeoSensitabs and PDM disc (30 micrograms)) and one locally prepared 30 micrograms disc (SS disc) on four different medium types: Mueller-Hinton agar (MH medium), MH medium and PDM agar II supplemented with 5% horse blood (HMB medium and PDM medium, respectively), and Danish blood agar (DBA medium). Two previously studied groups of Gram-positive bacteria were tested: group B (N = 75) comprised miscellaneous cocci, and group C (N = 59) mostly rods. With NeoSensitabs, mean zone diameters were larger than with PDM and SS discs on all medium types, and mean zone diameters were larger on DBA medium than on MHB and PDM medium with all disc types. The impact of the medium type on the zone diameter was evaluated for 121 strains growing on MHB medium, PDM medium, and DBA medium. Bacterial groups B and C each divided into three MIC groups were analysed separately. We compared mean zone diameters for each specific group with the average zone diameter, i.e. the mean value for all zone diameters obtained. The smallest deviations from the average zone diameters were observed on PDM medium for both teicoplanin and vancomycin. Thirty-seven percent of strains failed to grow on MH medium, but supplementation of MH medium with horse blood significantly reduced the zone diameter for group B strains both for teicoplanin and vancomycin. Poor predictability of MIC from the zone diameter was found especially for strains with MICs < or = 1 microgram/ml. The medium type hardly affected the results of regression analysis. In contrast, the medium type markedly affected the results of error-rate bounded analysis. No errors were recorded with the SS disc on MHB medium for either teicoplanin or vancomycin, but no strains with MICs of vancomycin within the intermediate group could be correctly classified on DBA medium.

Screening method for beta-lactamase substrate profiles.

A disc diffusion method, based on the idea of Klundert, for screening of substrate profiles of beta-lactamases was developed in order to perform epidemiological studies. The method was tested against 30 different reference beta-lactamases and 59 clinical isolates known to produce TEM-1, SHV-1 and BRO-1. The reproducibility and discriminating power of the disc diffusion method made it possible to differentiate between eight types of substrate profiles for the 30 reference beta-lactamases and to differentiate between TEM-1, SHV-1 and BRO-1 from clinical isolates. In combination with analytical isoelectric focusing the disc diffusion method gives a reliable identification of beta-lactamases.

In vitro activity of teicoplanin and vancomycin against gram-positive bacteria from human clinical and veterinary sources.

The minimum inhibitory concentration (MIC) of teicoplanin and vancomycin was determined by the agar dilution method for 186 Gram-positive bacteria from human clinical and veterinary sources. Teicoplanin MIC values were less than or equal to 4 micrograms/ml for 94% of staphylococci (group A, n = 52) and less than or equal to 2 micrograms/ml for all streptococci, enterococci, aerococci and pediococci (group B, n = 75). Seventy-eight percent of Gram-positive rods, Rhodococcus and Leuconostoc spp. (group C, n = 59) were inhibited by 4 micrograms/ml. Teicoplanin resistance (MIC greater than or equal to 16 micrograms/ml) was demonstrated for all Nocardia strains and for some strains of Lactobacillus, E. rhusiopathiae, Leuconostoc, and S. haemolyticus. Cross-resistance between teicoplanin and vancomycin was observed for all Nocardia strains and for some strains of Lactobacillus, E. rhusiopathiae, and Leuconostoc. Three methicillin-resistant S. haemolyticus strains were either resistant or intermediately susceptible to teicoplanin and susceptible to vancomycin. Eight strains (motile enterococci four, E. rhusiopathiae three and Leuconostoc sp. one) were susceptible to teicoplanin and resistant to vancomycin. Teicoplanin disc diffusion on Danish Blood Agar with NeoSensitabs (Rosco), PDM AB Biodisc and locally prepared discs revealed a wide range of zone diameters in groups B and C. The relation between MIC values and zone diameters for teicoplanin was analysed by the error-rate bounded method. Zone size interpretive criteria as suggested by the manufacturers (greater than or equal to 15 mm) produced 2.7% (95% confidence limits 0.9-6.2%) and 1.6% (95% confidence limits 0.3-4.6%) very major errors for NeoSensitabs and PDM-disc, respectively. Using a zone size breakpoint for susceptibility of greater than or equal to 25 mm for NeoSensitabs and greater than or equal to 20 mm for PDM-disc, the proportions of very major errors were 0.5% (95% confidence limits 0.0-3.0%) at the expense of 5.9% (95% confidence limits 3.0-10.3%) indeterminate strains that belonged to E. rhusiopathiae, Leuconostoc, Lactobacillus and S. haemolyticus. However, using these zone size breakpoints five major errors (beta-haemolytic streptococci, group B three, S. aureus one, Leuconostoc sp. one) were observed for NeoSensitabs and two major errors (beta-haemolytic streptococcus, group B one, Leuconostoc sp. one) were observed for PDM-disc. Susceptibility testing against teicoplanin among these taxa should therefore include a determination of MIC.

The antibacterial activity of a siderophore. 2. The influence of deferoxamine alone and combined with ascorbic acid on the activity of antibiotics against Staphylococcus aureus.

The in vitro activity of deferoxamine (DFO) combined with cephalothin, gentamicin, cefotaxime, vancomycin, and fusidic acid, in the presence or absence of the reductant ascorbic acid (AA) was investigated against Staphylococcus aureus by a macrobroth dilution technique and killing curve kinetics. DFO and in particular DFO + AA lowered the MICs of cephalothin, gentamicin, cefotaxime, and fusidic acid for most of the strains and in some instances also the MICs of vancomycin. To characterize the interaction between DFO or DFO + AA and antimicrobials we applied the growth constants of logarithmic growth phase. Generally DFO acted synergistically with cephalothin, gentamicin, vancomycin, and fusidic acid, particularly in the presence of AA, and in some cases synergy was demonstrated with cefotaxime, too.

Evaluation of Stuart's transport medium containing penicillinase. An in vitro study. Brief report.

An in vitro study was performed to evaluate the value of penicillinase in Stuart's transport medium. Swabs were taken from serum broth cultures of Staphylococcus aureus, Escherichia coli and Proteus rettgeri, incubated with cefuroxime. Following 24 hours' storage in the transport medium, the recovery of P. rettgeri was significantly higher from Stuart's transport media containing penicillinase in a concentration of 100,000 units/ml, than from swabs stored in Stuart's medium without penicillinase. Although susceptible to cefuroxime, we did not find the same effect on S. aureus and E. coli, presumably because these strains had higher MIC values than the P. rettgeri strain tested.

Experimental pneumococcus infection in mice: comparative in vitro and in vivo effect of cefuroxime, cefotaxime and ceftriaxone.

In a mouse model using intraperitoneal inoculation of Streptococcus pneumoniae type 3, the 50% effective dose, ED50, after single doses one hour post-inoculation was considerably lower for ceftriaxone (CRO) than for cefuroxime (CXM) and cefotaxime (CTX), in spite of the same minimal inhibitory concentration, MIC, of 0.02 mcg/ml against the pneumococcus for all 3 drugs. The bactericidal activity as measured by time-kill curves was similar for the 3 drugs, as was the post-antibiotic effect in vitro. Protein binding in mouse serum was considerably higher for CRO (87%) than for both CTX (35%) and CXM (15%), respectively. Of pharmacokinetic parameters investigated on doses equal to the ED50s, the time the serum antibiotic concentration remained above the MIC (delta T(MIC)) was the factor that varied the least among 3 drugs. Therefore, the superior in vivo effect for CRO is not due to higher intrinsic activity against the pathogen but to the long serum-elimination half-life resulting in an extended delta T(MIC), probably related to the high serum protein binding.

The pneumococcus and the mouse-protection test: correlation of in vitro and in vivo activity for beta-lactam antibiotics, vancomycin, erythromycin and gentamicin.

The mouse-protection test with intraperitoneal inoculation of Streptococcus pneumoniae type 3 was employed to compare the effect in vivo and to correlate with activity in vitro of ampicillin, piperacillin, methicillin, cefuroxime, erythromycin, vancomycin and gentamicin. The MICs for these drugs were lowest for the beta-lactam antibiotics, highest for vancomycin and gentamicin. Relative to the MIC, gentamicin and vancomycin showed the highest bactericidal rates against the pneumococcus. Vancomycin was the most effective in vivo as measured by the 50% effective dose (ED50) after single doses 1 h post-inoculation. Serum vancomycin concentrations measured after doses equal to the ED50 were below the minimal concentration measurable by our bioassay (i.e. less than 3.7 micrograms/ml). For the other drugs, peak concentrations in serum were higher than those produced even by high doses in man. Among pharmacokinetic parameters studied at dosages equal to the ED50s, the period during which the serum concentration exceeded the MIC (delta T(MIC] was the factor which varied the least for the beta-lactam antibiotics (range, 2-5 h). For gentamicin and probably also for vancomycin the delta T(MIC) was below 20 min, while it was considerably longer (i.e. 19 h) for erythromycin, although the bactericidal activity of this drug in vitro was comparable to that of the beta-lactam antibiotics.

Experimental pneumococcus infection in mice: correlation of bactericidal activity in vitro with the effect in vivo for gentamicin, netilmicin and tobramycin.

An experimental model in mice, incorporating the intraperitoneal inoculation of a Streptococcus pneumoniae type 3, was used to evaluate the effect in vivo after single-dose administration of the three aminoglycosides, gentamicin, tobramycin and netilmicin, and to correlate this effect with their in vitro activity against the pathogen, in particular the bactericidal rate. The minimal inhibitory concentrations (MIC's), which were equal to the minimal bactericidal concentrations (MBC's), were 12.5 micrograms/ml for netilmicin, and 25 micrograms/ml for the two other aminoglycosides, respectively. All three antibiotics showed excellent bactericidal activities even at concentrations 1/4 times the MIC's, but the bactericidal rate was clearly lower for tobramycin than for the two other aminoglycosides. The effect in vivo measured as the 50% effective dose (ED50) closely reflected the relative bactericidal activities of the drugs. Of the pharmacokinetic parameters investigated on dosages equal to the ED50's for the three drugs, the best to correlate with the bactericidal rates in vitro were the peak serum concentrations.

Experimental infection with Streptococcus pneumoniae in mice: correlation of in vitro activity and pharmacokinetic parameters with in vivo effect for 14 cephalosporins.

A mouse model using intraperitoneal inoculation of Streptococcus pneumoniae type 3 was used to compare in vitro and in vivo effects of 14 cephalosporins, selected to encompass a wide range of minimal inhibitory concentrations (MICs) against the organism. Antibiotics were subcutaneously administered as single doses 1 hr after inoculation of pneumococci, and the effect was measured as the 50% effective dose (ED50). The correlation between log ED50 and log MIC was highly significant (r = .87, P less than .001). Pharmacokinetic properties of the cephalosporins were estimated after a fixed dose of 5 mg per mouse (167 mg/kg) for all drugs. The only correlation that was significant was between log ED50 and the time the serum concentration remained above the MIC for each drug (r = -.90, P less than .001). Ceftriaxone was the most-effective cephalosporin in vivo because of a combination of high in vitro activity and prolonged serum elimination half-life.

Susceptibility testing with disc diffusion for new cephalosporins: pre-diffusion revisited.

The purpose of the present study was to investigate the effect of 20-hour pre-diffusion, i.e. placing the antibiotic-containing discs on the agar for 20 hours prior to inoculation, as compared to direct diffusion, i.e. placing the discs on the agar immediately after inoculation, for the newer 3. generation cephalosporins as represented by ceftazidime and ceftriaxone. Regression lines (zone sizes vs. inhibitory concentrations, as measured by plate-dilution) were constructed for three groups of bacteria chosen because of their differences in growth characteristics on agar: E. coli (n = 50), Enterobacter sp. (n = 35), and streptococci (n = 51). The results for both cephalosporins were: 20-hour pre-diffusion produced larger zones than direct diffusion, regression studies for 20-hour pre-diffusion as compared to direct diffusion resulted in greater variation in zone sizes, numerically lower slopes, lower residual variances and higher correlation coefficients, and regression lines were significantly different for the 3 groups of bacteria with direct diffusion but not so with 20-hour pre-diffusion. Considering the interpretation of zone sizes with disc diffusion for the cephalosporins tested, 20-hour pre-diffusion was superior to direct diffusion.

The pneumococcus and the mouse protection test: inoculum, dosage and timing.

Intraperitoneal inoculation in mice of a S. pneumoniae type 3 in beef broth resulted in immediate growth in vivo as evidenced by bacterial counts in peritoneal washings and in blood. Treatment with penicillin 1 hour after inoculation reduced the bacterial counts in vivo; however, different doses of penicillin-G showed a similar effect as measured by bacterial counts, in spite of differences in their effect upon survival of the mice. Therefore, the effect of antibiotics in vivo in this model was better correlated with death/survival of the animals. For comparative purposes the ED50, i.e. the 50% effective dose, should be determined. The effect of cephalosporins, i.e. cefuroxime and cefotaxime, in this model highly depended upon timing of the antibiotic administration as related to inoculation.

Antibacterial activity in vitro and regression studies for ceftazidime and ceftriaxone.

The antibacterial activity in vitro of ceftazidime and ceftriaxone was investigated against 575 recent clinical isolates. Both cephalosporins displayed excellent activity against most of the pathogens tested, in particular Gram-negative bacteria, including Pseudomonas aeruginosa. Apart from the Pseudomonas group, Acinetobacter calcoaceticus and Campylobacter jejuni ceftriaxone was slightly to moderately more active than ceftazidime overall. Ceftriaxone was moderately active against Streptococcus faecalis. Regression lines for the two antibiotics were almost identical. Corresponding to differences in susceptibility, the zone sizes differed for the two drugs with respect to certain bacterial groups, e.g. Pseudomonas sp. and enterococci. Therefore, the two cephalosporins cannot substitute for each other in disc susceptibility testing. Breakpoints for disc tests around 8-16 micrograms/ml, as suggested in the literature, appear too high considering the beneficial pharmacokinetic properties of the two drugs.