Search

ABSTRACTFluoroquinolones are among the drugs most extensively used for the treatment of bacterial infections in human and veterinary medicine. Resistance to quinolones can be chromosome or plasmid mediated. The chromosomal mechanism of resistance is associated with mutations in the DNA gyrase- and topoisomerase IV-encoding genes and mutations in regulatory genes affecting different efflux systems, among others. We studied the role of the acquisition of a mutation in the gyrAgene in the virulence and protein expression of uropathogenic Escherichia coli(UPEC). The HC14366M strain carrying a mutation in the gyrAgene (S83L) was found to lose the capacity to cause cystitis and pyelonephritis mainly due to a decrease in the expression of the fimA, papA, papB, and ompAgenes. The levels of expression of the fimA, papB, and ompAgenes were recovered on complementing the strain with a plasmid containing the gyrAwild-type gene. However, only a slight recovery was observed in the colonization of the bladder in the GyrA complement strain compared to the mutant strain in a murine model of ascending urinary tract infection. In conclusion, a mutation in the gyrAgene of uropathogenic E. colireduced the virulence of the bacteria, likely in association with the effect of DNA supercoiling on the expression of several virulence factors and proteins, thereby decreasing their capacity to cause cystitis and pyelonephritis.

ABSTRACTMicrobial resistance is an increasing health concern and a true danger to human well-being. A worldwide search for new compounds is ongoing, and antimicrobial peptides are promising lead candidates for tomorrow's antibiotics. The decapeptide anoplin (GLLKRIKTLL-NH2) is an especially interesting candidate because of its small size as well as its antimicrobial and nonhemolytic properties. Optimization of the properties of an antimicrobial peptide such as anoplin requires multidimensional searching in a complex chemical space. Typically, such optimization is performed by labor-intensive and costly trial-and-error methods. In this study, we show the benefit of fractional factorial design for identification of the optimal antimicrobial peptide in a combination matrix. We synthesized and analyzed a training set of 12 anoplin analogs, representative of 64 analogs in total. Using MIC, hemolysis, and high-performance liquid chromatography retention time data, we constructed analysis-of-variance models that describe the relationship between these properties and the structural characteristics of the analogs. We show that the mathematical models derived from the training set data can be used to predict the properties of other analogs in the chemical space. Hence, this method provides an efficient means of identification of the optimal peptide in the searched chemical space.

Meningitis is the most serious of invasive infections caused by the Gram-positive bacterium Streptococcus pneumoniae. Vaccines protect only against a limited number of serotypes, and evolving bacterial resistance to antimicrobials impedes treatment. Further insight into the molecular pathogenesis of invasive pneumococcal disease is required in order to enable the development of new or adjunctive treatments and/or pneumococcal vaccines that are efficient across serotypes. We applied genomic array footprinting (GAF) in the search for S. pneumoniae genes that are essential during experimental meningitis. A total of 6,000 independent TIGR4 marinerT7 transposon mutants distributed over four libraries were injected intracisternally into rabbits, and cerebrospinal fluid (CSF) was collected after 3, 9, and 15 h. Microarray analysis of mutant-specific probes from CSF samples and inocula identified 82 and 11 genes mutants of which had become attenuated or enriched, respectively, during infection. The results point to essential roles for capsular polysaccharides, nutrient uptake, and amino acid biosynthesis in bacterial replication during experimental meningitis. The GAF phenotype of a subset of identified targets was followed up by detailed studies of directed mutants in competitive and noncompetitive infection models of experimental rat meningitis. It appeared that adenylosuccinate synthetase, flavodoxin, and LivJ, the substrate binding protein of a branched-chain amino acid ABC transporter, are relevant as targets for future therapy and prevention of pneumococcal meningitis, since their mutants were attenuated in both models of infection as well as in competitive growth in human cerebrospinal fluid in vitro.

ABSTRACTMeningitis is the most serious of invasive infections caused by the Gram-positive bacterium Streptococcus pneumoniae. Vaccines protect only against a limited number of serotypes, and evolving bacterial resistance to antimicrobials impedes treatment. Further insight into the molecular pathogenesis of invasive pneumococcal disease is required in order to enable the development of new or adjunctive treatments and/or pneumococcal vaccines that are efficient across serotypes. We applied genomic array footprinting (GAF) in the search for S. pneumoniaegenes that are essential during experimental meningitis. A total of 6,000 independent TIGR4 marinerT7transposon mutants distributed over four libraries were injected intracisternally into rabbits, and cerebrospinal fluid (CSF) was collected after 3, 9, and 15 h. Microarray analysis of mutant-specific probes from CSF samples and inocula identified 82 and 11 genes mutants of which had become attenuated or enriched, respectively, during infection. The results point to essential roles for capsular polysaccharides, nutrient uptake, and amino acid biosynthesis in bacterial replication during experimental meningitis. The GAF phenotype of a subset of identified targets was followed up by detailed studies of directed mutants in competitive and noncompetitive infection models of experimental rat meningitis. It appeared that adenylosuccinate synthetase, flavodoxin, and LivJ, the substrate binding protein of a branched-chain amino acid ABC transporter, are relevant as targets for future therapy and prevention of pneumococcal meningitis, since their mutants were attenuated in both models of infection as well as in competitive growth in human cerebrospinal fluid in vitro.

Modern urine-meters are provided with the urine collection bag directly attached to the measuring device, thus providing a closed collection system. The systems vary, however, in respect to how the measuring devices are emptied into the urine collection bag. Using an in vitro ascending contamination model we compared three systems: the Unometer (Unomedical), the Bard Urinemeter and the Curity Urinemeter (Kendall), which all differ in the methods of emptying. The urine collection bags were inoculated withPseudomonas aeruginosa, and subsequently sampled daily at six different sites for up to seven days in order to determine the degree and timing of ascending contamination of the systems. Unometer was significantly less contaminated at the urine-meter level at day seven than both Bard (0/8 vs. 4/4, P=0.002) and Curity (0/8 vs. 8/8, P=0.0002), respectively, as well as significantly less contaminated in the infusion-catheter above the urine-meter at day seven than Bard (0/8 vs. 4/4, P= 0.002), but not so as compared to Curity (0/8 vs. 1/8, P =1,0). The Bard and Curity systems empty the devices by tilting thereby probably contaminating the devices, while the Unometer is emptied via a non-return valve in the bottom of the device.

ABSTRACTResistance to antibiotics used for the treatment of urinary tract infections (UTIs) is increasing worldwide. The impact of in vitro resistance on clinical outcome in UTIs requires further study, since most studies of both humans and animals have evaluated only the efficacy of antibiotics toward bacteria susceptible in vitro. We were interested in evaluating the relationship between the in vitro antibacterial effect and the in vivo efficacy after antibiotic treatment. We simulated a natural ascending UTI by use of the ascending UTI mouse model and used Escherichia colistrains with various susceptibilities to amdinocillin (mecillinam) and sulfamethizole. Mice were treated for 3 days with antibiotic doses approximating human urinary tract concentrations after a standard oral dose. For a susceptible strain (MIC, 0.5 μg/ml) and a resistant strain (MIC, 128 μg/ml), respectively, there were significant reductions in bacterial counts in the urine, bladder, and kidneys after treatment with amdinocillin, whereas for a strain for which the MIC was 16 μg/ml, there was a significant reduction in bacterial counts in the kidneys only (P< 0.05). Treatment with sulfamethizole resulted in a significant reduction in bacterial counts in all samples from a susceptible strain (MIC, 128 μg/ml) and a resistant strain (MIC, 512 μg/ml). Infection with a sulIIgene-positive strain (MIC, >2,048 μg/ml) could not be treated with sulfamethizole, as no effect could be demonstrated in the urine, bladder, or kidneys. For amdinocillin, there was no clear-cut relationship between the in vitro susceptibility and the in vivo outcome, while for sulfamethizole, we found a relationship between the MIC for the strain and the effect in the urinary tract.

Antibiotic resistance of urinary tract pathogens has increased worldwide. Our aim was to provide information regarding resistance patterns of Escherichia coli in urinary tract infections (UTIs) and E. coli bacteraemia in Denmark. The overall resistance ranged from: ampicillin 20-47%, mecillinam 0-7%, trimethoprim 10-28%, sulfamethizole 22-47% and nitrofurantoin 0-3%. In strains with sulfamethizole MICs > 2048 mg/L, 97% carried sulI, sulII or both genes, with sulII being the most common. Among the sulI gene-positive strains, 96% were intI 1 gene positive.

Aerococcus urinae is a newcomer in clinical and microbiological practice, causing urinary tract infections, bacteraemia/septicaemia and/or endocarditis. This study presents for the first time an evaluation of the activity of a representative panel of antibiotics against a large number of A. urinae isolates. The in vitro susceptibilities (MICs) of 56 isolates of A. urinae to 14 antibiotics were determined by agar dilution. In general, A. urinae isolates showed little inter-isolate variability, and had low MICs of penicillin, amoxicillin, piperacillin, cefepime, vancomycin and rifampicin. High-level aminoglycoside resistance was not found for any of the isolates. Moderate to good activity was seen with quinolones, erythromycin and tetracycline. Isolates from two patients with endocarditis were studied with time-kill curves for penicillin, gentamicin and vancomycin. Penicillin and vancomycin alone exhibited slow or no bactericidal activity against the two strains. When combining either penicillin or vancomycin with gentamicin, rapid bactericidal activity was obtained for both strains with both combinations. The treatment options for A. urinae seem to include penicillins for less severe cases. In severe cases, i.e. endocarditis, the time-kill investigations suggest a beneficial effect of combination with gentamicin. In the penicillin-allergic patient vancomycin in combination with gentamicin represents the most obvious alternative.

Three methods were compared for the susceptibility testing of yeast isolates to fluconazole and amphotericin B: two fagar diffusion methods (Etest and a tablet diffusion test) and the National Committee for Clinical Laboratory Standards (NCCLS) broth microdilution method. Given as MIC(50)s (range), fluconazole endpoints were: for the 24 h broth microdilution test, 0.25 mg/L (0.06-32 mg/L); for the Etest, 0.38 mg/L (0.064-24 mg/L); and for the NCCLS broth microdilution test, 2 mg/L (0.06->or=64 mg/L). With breakpoints of <3 mg/L for susceptible and >16 mg/L for resistant, the Etest and the 24 h microdilution test classified the isolates in agreement with the classification obtained by the NCCLS method. Results obtained by Etest were in closer NCCLS method than those obtained with the tablet test. Amphotericin B endpoints were lower for the 24 h microdilution and Etests than MICs obtained by the NCCLS broth microdilution method. Reproducibility was high for all tests; however, disadvantages of both diffusion tests were microcolonies in the inhibition zone and dependence on stringent standardization of inoculum.

Criteria for defining urinary tract infection have changed as the importance of low count bacteriuria in patients with symptoms of infection has been recognized. Little has been published, however, regarding the ability of routine laboratory methods to meet this new challenge. We compared 1 μl and 10 μl calibrated plastic loops (Nunc, Denmark), to a commonly used dipslide method (Uricult, Orion, Finland) using a 0.1 ml pipette as the gold standard. Four typical uropathogens, E. coli, P. aeruginosa, S. saprophyticus and E. faecalis, were mixed in pooled human urine in numbers representing low counts, i.e. 10², 10³, and 10⁴, together with 10⁵ colony-forming units (cfu)/ml, and sampling with the four methods was performed 20 times for each with each bacterium and each dilution. Colony counts showed that in many cases the 1 μl loop did not deposit any bacteria on the agar plate when colony counts were lower than 10⁴ cfu/ml. The 10 μl loop constantly deposited 1.5 times higher numbers of bacteria than predicted, which correlated with the results of weighing the amount of water sampled, i.e. 15.24 (1.55) μg (mean (SD)). The dipslide counts correlated with the 10 μl loop although the agar mounts on the slides absorbed water equal to about 200 μl. There were no significant differences in the counts of the four different bacteria. Variation coefficients increased with decreasing volumes sampled, but the 10 μl loop was found sufficiently exact to detect counts from 10³ to 10⁴, since only 10-fold multiples are usually reported for urine cultures.

The polysaccharide fucoidin is a selectin blocker that inhibits leukocyte recruitment into the cerebrospinal fluid (CSF) during experimental pneumococcal meningitis. In the present study, the effect of fucoidin treatment on the release of the proinflammatory cytokines tumor necrosis factor alpha (TNF-alpha), interleukin-1 (IL-1), and IL-8 into the CSF was investigated. Rabbits (n = 7) were treated intravenously with 10 mg of fucoidin/kg of body weight every second hour starting 4 h after intracisternal inoculation of approximately 10(6) CFU of Streptococcus pneumoniae type 3 (untreated control group, n = 7). CSF samples were obtained every second hour during a 16-h study period. Treatment with fucoidin caused a consistent and significant decrease in CSF IL-1 levels (in picograms per milliliter) between 12 and 16 h (0 versus 170, 0 versus 526, and 60 versus 1,467, respectively; P < 0.02). A less consistent decrease in CSF TNF-alpha levels was observed in the fucoidin-treated group, but with no significant difference between the two groups (P > 0.05). In contrast, there was no attenuation in CSF IL-8 levels. Indeed, there was a significant increase in CSF IL-8 levels (in picograms per milliliter) in the fucoidin-treated group at 10 and 12 h (921 versus 574 and 1,397 versus 569, respectively; P < 0.09). In conclusion, our results suggest that blood-derived leukocytes mainly are responsible for the release of IL-1 and to some degree TNF-alpha into the CSF during pneumococcal meningitis, whereas IL-8 may be produced by local cells within the brain.

A possible immunomodulatory role of granulocyte colony-stimulating factor (G-CSF) was investigated in an experimental pneumococcal meningitis model in rabbits. Animals were pretreated with G-CSF (10 micrograms/kg subcutaneously twice a day) starting 48 h before in vivo and ex vivo experiments, causing a five- to six-fold increase in the peripheral leukocyte level. Meningitis was induced by intracisternal inoculation of approximately 4 x 10(5) CFU of Streptococcus pneumoniae type 3. Neutrophil pleocytosis and interleukin-8 (IL-8) levels were significantly attenuated in G-CSF-pretreated animals compared to untreated animals (P < 0.05). Furthermore, G-CSF pretreatment significantly delayed alterations in cerebrospinal fluid (CSF) tumor necrosis factor alpha and IL-1beta levels, as well as protein and glucose levels (P < 0.05). No difference in CSF bacterial concentrations was found, whereas the blood bacterial concentration was significantly decreased in G-CSF-pretreated animals (P < 0.05). Ex vivo chemotaxis of neutrophils isolated from G-CSF-pretreated animals was significantly decreased compared to that of neutrophils from untreated animals (P < 0.05). In conclusion, G-CSF pretreatment attenuates meningeal inflammation and enhances systemic bacterial killing. Further preclinical studies are required to investigate whether this may affect the clinical course of meningitis and thus whether G-CSF treatment may have a beneficial role in pneumococcal meningitis.

Summary Fifteen strains ofEnterococcus faecalis, all clinical blood culture isolates from patients with endocarditis, were studied by kill-kinetic experiments using penicillin G, ampicillin and amoxicillin alone and in combination with tobramycin. The median minimal inhibitory concentrations (MIC), were penicillin 4 mg/l, ampicillin 2 mg/l, amoxicillin 2 mg/l and tobramycin 32 mg/l. Equipotent doses of the antibiotics (1/2 × MIC, 1 × MIC and 4 × MIC) were used in the kill-kinetic studies. Synergism was studied using a combination of 1/2 × MIC of the beta-lactam antibiotic and 8 mg/l of tobramycin. The bactericidal activity did not exceed 10+3 cfu/ml at 5 hours for any single compound. After 5 h all three beta-lactam antibiotics in combination with tobramycin resulted in synergism, i.e. more than one hundredfold reduction of colony forming units (cfu) as compared to the most active single agent. Amoxicillin had a significantly higher bactericidal potential than ampicillin or penicillin both alone and in combination with tobramycin. The clinical significance of these findings warrants further studiesin vivo.

The correlation between pharmacokinetic parameters and the in-vivo effect of antibiotics in relation to bacterial growth phases was evaluated using the mouse peritonitis model with a penicillin-resistant pneumococcus. Different 8 h dosing regimens were applied, with different total doses and initiated at different times during the bacterial growth phase. The effect was measured as the decline in bacterial counts in the peritoneal cavity. The pharmacokinetic parameters showed major changes during the phases of growth, as the serum elimination of penicillin decreased during the infection. The same effect of dosing regimens was observed in the exponential and stationary phases. In two regimens where T(>MIC) (the time the serum concentration exceeded the MIC) was 50% of the treatment period, a significantly better effect was achieved with a 2 hourly regimen than with a regimen with treatments every 20 min. The T(>MIC) of each dose was shown to be a critical parameter for achieving an effect in all growth phases. The maximum effect of penicillin, a 5-6 x log10 decline in bacterial counts in the peritoneum of the mice, was achieved when T(>MIC) was >50% of the treatment time or longer than approximately 40 min of each dose. The 50% effective dose for protection after a single injection, ED50, was measured in the different phases of the infection and found to increase with the duration of the pneumococcal infection, while mice treated 24 h after challenge were beyond therapeutic range. The correlation between the effect of penicillin and pharmacokinetic parameters appears to follow the same rules during the different in-vivo growth phases of pneumococci.

No standardized method of susceptibility testing for Helicobacter pylori is currently available, so before a large agar dilution study comprising 230 H. pylori strains belonging to more than 80 genetically different groups was initiated, we performed a relatively small preliminary study to determine the influences of medium, inoculum density, and incubation time. Seven media were investigated and were primarily evaluated on the basis of their abilities to support growth both semiquantitatively and qualitatively; Iso-Sensitest agar supplemented with 10% horse blood was found to be well suited for the purpose; this was closely followed by Mueller-Hinton agar with 10% horse blood, Mueller-Hinton with 10% sheep blood, and finally, 7% lysed horse blood agar. Investigations of two inoculum densities and two incubation times resulted in recommendations for the use of 10(9) CFU/ml (10[6] CFU/spot) as the inoculum and 72 h as the incubation time. A modest inoculum effect was noted for amoxicillin and metronidazole. By the methodology derived from our preliminary study, the susceptibilities of 230 H. pylori strains to six antibiotics were subsequently determined. The results were generally in accord with those of others, and apart from metronidazole, the MIC of which for approximately 25% of the strains tested was >8 microg/ml, resistance was low in Denmark. The situation might, however, quickly change when and if the number of indications for antibiotic therapy for H. pylori infections increase. Consequently, susceptibility testing of all H. pylori strains is recommended in order to survey the development of resistance, and in our hands the described methodology was relatively easy to perform and the results were easy to read.

The activities of glycopeptides against pneumococci were studied in vitro and in vivo. The MICs of two glycopeptides, vancomycin and teicoplanin, in different media against 10 strains of pneumococci with different susceptibilities to penicillin were determined. The MICs of teicoplanin were four times lower than those of vancomycin in Mueller-Hinton media supplemented with 5% blood, but the MICs were similar in mouse and human sera supplemented with 5% blood. The serum protein binding levels in mouse and human sera were 90% for teicoplanin in both and 25 and 35%, respectively, for vancomycin. The MICs for vancomycin and teicoplanin were only correlated in human serum (P < 0.001). The single doses giving protection to 50% of the animals in the mouse peritonitis model after a lethal challenge of pneumococci, the ED50s, were similar for vancomycin and teicoplanin, between 0.1 and 1 mg/kg of body weight for all 10 strains. The log ED50s were significantly correlated only to the log MICs of teicoplanin determined for mouse serum with 5% blood (P = 0.01) and to the log MICs of vancomycin determined by the E test (P = 0.03). Among the pharmacokinetic parameters analyzed at the ED50s, the most constant parameter was the time for which the drug concentration exceeded the MIC (T(>MIC)) when each drug was considered separately; however, when both drugs were considered together, the maximum concentration of drug in serum (Cmax) varied the least. This indicates that both these parameters are of importance for predicting the effect of the drugs. We conclude that the effect of glycopeptides was not influenced by the penicillin resistance of the pneumococci, either in vitro or in vivo, and that the superior activity of teicoplanin over that of vancomycin in vitro was abolished in vivo, an effect which probably was due to the high serum protein binding of teicoplanin. Both the pharmacokinetic parameters T(>MIC) and Cmax are important predicting the effect of glycopeptides, but the pharmacodynamics of glycopeptides are still not completely elucidated.

A new mouse model of foreign body infection has been developed. Intraperitoneal placement of a silicone catheter followed by injection of 10(8) Staphylococcus aureus organisms resulted in a reproducible, localized foreign body infection. The infection persisted as an intra-abdominal abscess surrounding the catheter for at least 30 days. Treatment with up to nine doses of methicillin or gentamicin or both was started 3 days after infection. The treatment showed a significant effect (P < 0.05), measured as reduction of bacteria on the foreign body, for all three regimens with a reduction of up to 2 log units, but no synergism was observed. The result of the treatment was poor, despite the facts that the local concentrations of methicillin were greater than the MIC for at least 72 h and that nine peak concentrations of gentamicin of > 13 micrograms/ml were obtained. The poor result of the treatment was not caused by development of antibiotic resistance or influenced by protein concentration, pH, or local presence in the pus of inhibitors of antibiotics. Both antibiotics showed good effects in time-kill studies in vitro on bacteria on catheters taken out of infected mice and catheters infected in vitro. During treatment, the proportion of intracellular bacteria increased in all treated mice to 60 to 75% compared with 20 to 30% in nontreated mice (P < 0.05). This indicates that intracellular survival of staphylococci may influence the outcome of the treatment in foreign body infections.

Summary The pharmacokinetics of dicloxacillin and flucloxacillin were studied in 12 healthy volunteers after oral administration. The participants received a single dose of either dicloxacillin (0.5 g, 0.75 g or 1.0 g) or flucloxacillin (0.75 g) in a cross-over fashion. Antibiotic concentrations were determined in serum and urine by bioassay and followed for 8 and 24 h, respectively. The three dicloxacillin dosages showed no significant differences for the serum elimination half-lives (t1/2ß, median: 72 min). Comparing 0.75 g flucloxacillin with the same dose of dicloxacillin, no significant differences between the values of Cmax, t1/2ß and AUC were found. Protein binding as determined by ultrafiltration in pooled serum was 94.7–96.2% for flucloxacillin and 96.4–97.2% for dicloxacillin. The serum bactericidal titers were similar for the two drugs. In conclusion, dicloxacillin and flucloxacillin showed similar pharmacokinetic behavior after 0.75 g doses in human volunteers.

The clinical manifestations of and epidemiological data from 11 patients infected withVibrio vulnificus admitted to Danish hospitals during the unusually warm summer of 1994 are reported. All patients contracted the disease after exposure to seawater; however, none had consumed seafood. Four patients developed bacteremia, one of whom subsequently died; nine patients, including the four with bacteremia, exhibited skin manifestations. Four patients contracted the disease while fishing; in at least one case the patient had handled eels. AllVibrio vulnificus strains were highly susceptible to 11 antimicrobial agents tested. Plasmid analysis revealed that 8 of 11 strains carried plasmids. Ribotyping using the enzymeHindIII on the 11 strains showed five different types, two of which comprised four strains each. The present study provides the first clinical and epidemiological data about a series of humanVibrio vulnificus infections from a temperate zone.

The purpose of the study was to investigate the correlation of in vitro activity with the in vivo effect and the pharmacokinetics of penicillin in the treatment of infections with pneumococci with various susceptibilities to penicillin. We used 10 pneumococcal strains for which penicillin MICs ranged from 0.016 to 8 micrograms/ml. Time-kill curve experiments were performed with all strains. We found that the effect of penicillin in vitro is concentration independent, with a maximum effect at two to four times the MIC for penicillin-susceptible as well as penicillin-resistant pneumococci. The mouse peritonitis model with an inoculum of approximately 10(6) CFU, to which mucin was added, resulted in a reproducible lethal infection with the pneumococci. The 50% effective dose was determined for each strain, and we found a highly significant correlation between the log MIC and the log 50% effective dose of penicillin against these strains (P < 0.01). Furthermore, it was shown that the most important pharmacokinetic parameter determining the effect of penicillin in vivo was the time that the concentration of penicillin in serum was greater than the MIC.

Summary The clinical and bacteriological efficacy as well as the tolerance of mecillinam, a newß-lactam antibiotic, administered parenterally in a dose of 40 mg/kg body weight, was investigated in 21 hospitalized patients with urinary tract infections or septicaemia. Success, defined as eradication of infecting organisms two to five days after treatment, was found in eight of 16 patients with urinary tract infections. Persistence of the original pathogen after treatment was seen in four patients, all with complicated urinary tract infection. Reinfection was seen in two patients, while the results were unevaluable in two cases. The five patients with septicaemia were all cured of their infection, this result being attributed solely to mecillinam in two cases, while a combined action of mecillinam and another antibiotic produced a cure in the other three cases.Escherichia coli resistant to mecillinam were isolated from two patients with persistence of bacteriuria following mecillinam treatment. Impairment of renal function may have been a contributory factor to the poor treatment response in patients with complicated urinary tract infection. Inasmuch as no important side-effects were recorded and mecillinam appears safe in patients with impaired renal and liver function, a higher dose may be indicated in these more complicated cases.

&lt;it>Objectives&lt;/it>: It has been suggested recently that intracellular bacteria surviving antibiotic treatment might serve as a reservoir for recurrent infection. The purpose of this study was to directly examine the location of &lt;it>Escherichia coli&lt;/it> bacteria in the mouse bladder after treatment with mecillinam. &lt;it>Methods&lt;/it>: The bladders were studied by use of colony counts, &lt;it>in situ&lt;/it> hybridization and electron microscopy. &lt;it>Results&lt;/it>: The bacterial counts in the bladder remained ∼103–4 cfu/bladder even after mecillinam treatment had finished, and re-growth in the urine was observed. In the bladder epithelium from treated mice, bacteria cells were occasionally seen, presumably representing intracellularly located bacteria. &lt;it>Conclusions&lt;/it>: This is the first &lt;it>in vivo&lt;/it> study indicating that during mecillinam treatment &lt;it>E. coli&lt;/it> cells can penetrate the mouse bladder epithelium and persist.

&lt;it>Objective&lt;/it>: To evaluate cefoxitin 5 and 10 μg discs for detection of methicillin resistance in staphylococci. &lt;it>Methods&lt;/it>: Six hundred and forty-one &lt;it>Staphylococcus aureus&lt;/it> (261 &lt;it>mecA-&lt;/it>negative and 380 &lt;it>mecA&lt;/it>-positive) and 344 coagulase-negative staphylococci (CoNS) (132 &lt;it>mecA-&lt;/it>negative and 212 &lt;it>mecA&lt;/it>-positive) were investigated. The CoNS represented nine species, &lt;it>Staphylococcus epidermidis&lt;/it> being the most frequent (&lt;it>n&lt;/it>=231). All isolates were tested using semi-confluent growth on Iso-Sensitest agar (ISA), and Mueller–Hinton agar (MH) using a 5 and a 10 μg cefoxitin disc and overnight incubation in ambient air at 35–37°C. &lt;it>Results&lt;/it>: For &lt;it>S. aureus&lt;/it>, both cefoxitin discs performed with high accuracy on both media. The sensitivity and specificity for the following proposed interpretive zone diameters were: ISA 5 μg, R &lt; 14 mm (99.5% and 98.1%); ISA 10 μg, R &lt; 22 mm (99.5% and 98.1%); MH 5 μg, R &lt; 12 mm (99.7% and 98.1%); and MH 10 μg, R &lt; 18 mm (99.5% and 98.9%), respectively. All four variants were superior to oxacillin using the former SRGA methodology. In CoNS, a substantial overlap was seen for all variants. However, by avoiding primary interpretation in the overlapping interval, highly accurate results could be obtained for 81%, 80%, 91% and 97% of the isolates, respectively. &lt;it>Conclusion&lt;/it>: For &lt;it>S. aureus&lt;/it>, cefoxitin 5 and 10 μg discs performed with high accuracy on both ISA and MH using semi-confluent growth and standard incubation conditions. With the introduction of a defined interval in which primary interpretation should be avoided, the method could also be used for CoNS.

&lt;it>Objectives&lt;/it>: To evaluate the performance of a cefoxitin 30 µg disc on Iso-Sensitest agar, using a semi-confluent inoculum and overnight incubation at 35–36°C, for detection of methicillin-resistant &lt;it>Staphylococcus aureus&lt;/it> (MRSA). &lt;it>Methods&lt;/it>: A total of 457 &lt;it>S. aureus&lt;/it>, including 190 MRSA of several defined PFGE types and a number of low-level resistant isolates, were tested with a cefoxitin 30 µg disc on Iso-Sensitest agar, using a semi-confluent inoculum and overnight incubation at 35–36°C. This method was compared with the standard SRGA (Swedish Reference Group for Antibiotics) method (oxacillin 1 µg disc on Iso-Sensitest agar supplemented with 5&percnt; defibrinated horse blood, confluent growth and 24 h incubation in ambient air at 30°C). &lt;it>Results&lt;/it>: The cefoxitin method was excellent, with a sensitivity of 100&percnt; and a specificity of 99&percnt; using an interpretative zone diameter of S ≥ 29 mm and R &lt; 29 mm. Its performance was much better than the SRGA method, which with this collection of difficult strains had a sensitivity of only 78&percnt; using the current breakpoint of S ≥ 12 mm. &lt;it>Conclusion&lt;/it>: We suggest that the cefoxitin method should replace that currently recommended by the SRGA for the detection of MRSA, and that it would fit well into BSAC methodology.

Of the 15,170 cases of bacteremia due to Staphylococcus aureus that occurred in Denmark between 1959 and 1988, we review 525 cases of acute hematogenous osteomyelitis and 185 cases of septic arthritis that developed subsequent to the bacteremia and 134 cases of contiguous osteomyelitis in which the bacteremia developed secondarily. The pattern of acute infections of bones and joints has changed over the three decades studied. The frequency of secondary bone or joint infections due to S. aureus bacteremia has changed, as have the phage-type pattern and antibiotic resistance of the infective strains. The prevalence of hospital-acquired cases has increased and the age distribution of patients has changed, as is reflected in an increasing number of older patients. The localization of hematogenous osteomyelitis has shifted, and the vertebral column is now the most common site of infection. The rate of chronic cases of osteomyelitis that occur following acute hematogenous osteomyelitis has been reduced from 34% to 6%. The mortality associated with S. aureus bacteremic infections of bones or joints is low compared to that associated with other cases of S. aureus bacteremia.