In vitro activity of antibiotics potentially effective against difficult-to-treat strains of Gram-negative rods: retrospective study

Bacterial resistance surveillance is one of the main outputs of microbiological laboratories and its results are important part of antimicrobial stewardship (AMS). In this study, the susceptibility of specific bacteria to selected antimicrobial agents was tested. The susceptibility of 90 unique isolates of pathogens of critical priority obtained from clinically valid samples of ICU patients in 2017-2021 was tested. 50% of these fulfilled difficult-to-treat resistance (DTR) criteria and 50% were susceptible to all antibiotics included in the definition. 10 Enterobacterales strains met DTR criteria, and 2 (20%) were resistant to colistin (COL), 2 (20%) to cefiderocol (FCR), 7 (70%) to imipenem/cilastatin/relebactam (I/R), 3 (30%) to ceftazidime/avibactam (CAT) and 5 (50%) to fosfomycin (FOS). For Enterobacterales we also tested aztreonam/avibactam (AZA) for which there are no breakpoints yet. The highest MIC of AZA observed was 1 mg/l, MIC range in the susceptible cohort was 0.032-0.064 mg/l and in the DTR cohort (incl. class B beta-lactamase producers) it was 0.064-1 mg/l. Two (13.3%) isolates of Pseudomonas aeruginosa (15 DTR strains) were resistant to COL, 1 (6.7%) to FCR, 13 (86.7%) to I/R, 5 (33.3%) to CAT, and 5 (33.3%) to ceftolozane/tazobactam. All isolates of Acinetobacter baumannii with DTR were susceptible to COL and FCR, and at the same time resistant to I/R and ampicillin/sulbactam. New antimicrobial agents are not 100% effective against DTR. Therefore, it is necessary to perform susceptibility testing of these antibiotics, use the data for surveillance (including local surveillance) and conform to AMS standards.

Elimination and penetration of amikacin into urine in patients with decreased glomerular filtration rate

Background

Amikacin monotherapy is recommended for urinary tract infection (UTI) treatment with multi-resistant pathogens. Even though amikacin efficacy in the treatment of UTIs is dependent on its urinary concentration, there are no robust data proving that sufficiently high urinary concentration is reached in patients with reduced glomerular filtration rate (GFR).

Methods

A prospective study to monitor amikacin penetration into urine of 70 patients [40 males, median (interquartile range) age 70 (65-79) years] with different levels of glomerular filtration decline, including patients treated by dialysis, was conducted. The bactericidal efficacy of amikacin in urine samples has been evaluated.

Results

Patients with estimated GFR (eGFR) <30 mL/min had significantly lower median amikacin urinary concentration than patients with eGFR >30 mL/min (89.75 vs 186.0 mg/L, P < .0001; 200.5 vs 830.0 mg/L, P < .0001; and 126.0 vs 408.0 mg/L, P < .0001 for minimal, maximal and minimal together with maximal concentrations, respectively). The amount of amikacin eliminated in the first 10-13 h after dose administration was dependent on eGFR (r2 = 0.6144, P < .0001). The urinary concentration of amikacin in patients treated by dialysis was indirectly proportional to pH of urine. The plasma concentrations of amikacin did not correlate with urinary levels in patients in either of the GFR categories. Microbiological evaluation showed that the critical urinary concentration for efficacy of amikacin during UTI monotherapy in patients treated by dialysis is 100 mg/L. We found that 4 out of 11 patients treated by dialysis did not reach this level during the treatment.

Conclusion

Systemic administration of amikacin monotherapy in patients treated by dialysis is questionable as the concentrations of amikacin in their urine are often below the threshold of effectivity. Amikacin plasma concentrations are not a major determinant of amikacin concentration in urine, therefore pulse dosing is neither necessary nor safe in patients treated by dialysis, and may cause undesirable toxicity.

Intraperitoneally Administered Vancomycin in Patients with Peritoneal Dialysis-Associated Peritonitis: Population Pharmacokinetics and Dosing Implications

Peritonitis is a limiting complication of peritoneal dialysis, which is treated by intraperitoneal administration of antibiotics. Various dosing strategies are recommended for intraperitoneally administered vancomycin, which leads to large differences in intraperitoneal vancomycin exposure. Based on data from therapeutic drug monitoring, we developed the first-ever population pharmacokinetic model for intraperitoneally administered vancomycin to evaluate intraperitoneal and plasma exposure after dosing schedules recommended by the International Society for Peritoneal Dialysis. According to our model, currently recommended dosing schedules lead to possible underdosing of a large proportion of patients. To prevent this, we suggest avoiding intermittent intraperitoneal vancomycin administration, and for the continuous dosing regimen, we suggest a loading dose of 20 mg/kg followed by maintenance doses of 50 mg/L in each dwell to improve the intraperitoneal exposure. Vancomycin plasma level measurement on the fifth day of treatment with subsequent dose adjustment would prevent it from reaching toxic levels in the few patients who are susceptible to overdose.

[Can a difficult-to-treat resistance approach improve cooperation between microbiologists and clinicians?]

BACKGROUND

We introduce a relatively new difficult-to-treat resistance (DTR) category to specialists. It significantly influences the predictability of morbidity and mortality in patients with invasive infections caused by DTR strains. Therefore, surveillance of DTR is an important tool of antimicrobial stewardship (AMS) and widely contributes to cooperation between microbiologists and clinicians. We also report the prevalence of strains meeting the criteria for the category in a teaching hospital.

METHODS

This retrospective cohort single center study included invasive isolates of gram-negative rods from patients hospitalized in the General University Hospital in Prague in 2009-2013 and in 2017-2021.

RESULTS

From a total of 1 732 (920 and 812, respectively) unique strains of gram-negative rods isolated from blood cultures in both periods, 6.6 % were carbapenem-resistant in 2009-2013 and 6.0 % in 2017-2021; 3.7 % were identified as DTR in both periods. Most of the DTR strains were A. baumannii (23.1 % and 45.0 %, respectively) and P. aeruginosa (22.2 % and 15.3 %, respectively). We identified no carbapenem-resistant E. coli and therefore no DTR E. coli.

CONCLUSION

Infections caused by bacterial strains with a DTR phenotype are grave complications and are tricky to manage. The prevalence of severe infections caused by these strains was relatively low in the studied facility. Antibiotics with anti-DTR effects should be considered the last resort, so it is very important to comply with AMS rules and examine the susceptibility of these agents.

Vancomycin-releasing cross-linked collagen sponges as wound dressings

The study presents a novel vancomycin-releasing collagen wound dressing derived from Cyprinus carpio collagen type I cross-linked with carbodiimide which retarded the degradation rate and increased the stability of the sponge. Following lyophilization, the dressings were subjected to gamma sterilization. The structure was evaluated via scanning electron microscopy images, micro-computed tomography, and infrared spectrometry. The structural stability and vancomycin release properties were evaluated in phosphate buffered saline. Microbiological testing and a rat model of a wound infected with methicillin-resistant Staphylococcus aureus (MRSA) were then employed to test the efficacy of the treatment of the infected wound. Following an initial mass loss due to the release of vancomycin, the sponges remained stable. After 7 days of exposure in phosphate buffered saline (37°C), 60% of the material remained with a preserved collagen secondary structure together with a high degree of open porosity (over 80%). The analysis of the release of vancomycin revealed homogeneous distribution of the antibiotic both across and between the sponges. The release of vancomycin was retarded as proved by in vitro testing and further confirmed by the animal model from which measurable concentrations were observed in blood samples 24 hours after the subcutaneous implantation of the sponge, which was more than observed following intraperitoneal administration. The sponge was also highly effective in terms of reducing the number of colony-forming units in biopsies extracted from the infected wounds 4 days following the inoculation of the wounds with the MRSA solution. The presented sponges have ideal properties to serve as wound dressing for prevention of surgical site infection or treatment of already infected wounds.

Evaluation of collagen/hydroxyapatite electrospun layers loaded with vancomycin, gentamicin and their combination: Comparison of release kinetics, antimicrobial activity and cytocompatibility

The aim of this study was to develop a biodegradable nanostructured electrospun layer based on collagen (COL), hydroxyapatite nanoparticles (HA), vancomycin hydrochloride (V), gentamicin sulphate (G) and their combination (VG) for the treatment of prosthetic joint infections and the prevention of infection during the joint replacement procedure. COL/HA layers containing different amounts of HA (0, 5 and 15 wt%) were tested for the in vitro release kinetics of antibiotics, antimicrobial activity against MRSA, gentamicin-resistant Staphylococcus epidermidis and Enterococcus faecalis isolates and cytocompatibility using SAOS-2 bone-like cells. The results revealed that the COL/HA layers released high concentrations of vancomycin and gentamicin for 21 days and performed effectively against the tested clinically-relevant bacterial isolates. The presence of HA in the collagen layers was found not to affect the release kinetics of the vancomycin from the layers loaded only with vancomycin or its combination with gentamicin. Conversely, the presence of HA slowed down the release of gentamicin from the COL/HA layers loaded with gentamicin and its combination with vancomycin. The combination of both antibiotics exerted a positive effect on the prolongation of the conversion of vancomycin into its degradation products. All the layers tested with different antibiotics exhibited potential antibacterial activity with respect to both the tested staphylococci isolates and enterococci. The complemental effect of vancomycin was determined against both gentamicin-resistant Staphylococcus epidermidis and Enterococcus faecalis in contrast to the application of gentamicin as a single agent. This combination was also found to be more effective against MRSA than is vancomycin as a single agent. Importantly, this combination of vancomycin and gentamicin in the COL/HA layers exhibited sufficient cytocompatibility to SAOS-2, which was independent of the HA content. Conversely, only gentamicin caused the death of SAOS-2 independently of HA content and only vancomycin stimulated SAOS-2 behaviour with an increased concentration of HA in the COL/HA layers. In conclusion, COL/HA layers with 15 wt% of HA impregnated with vancomycin or with a combination of vancomycin and gentamicin offer a promising treatment approach and the potential to prevent infection during the joint replacement procedures.

A novel gentamicin-releasing wound dressing prepared from freshwater fish Cyprinus carpio collagen cross-linked with carbodiimide

Our study presents a novel collagen wound dressing prepared from freshwater fish skin ( Cyprinus Carpio) collagen type I. Half of the sponges were cross-linked with carbodiimide. The cross-linked and non-cross-linked collagen sponges were subsequently impregnated with gentamicin and lyophilized thus allowing for the attainment of the appropriate gentamicin content without the removal thereof during the cross-linking stage. The structure was evaluated via micro-CT and infrared spectrometry and the structural stability and gentamicin release properties were evaluated in phosphate buffer solution. The sponges were further tested via a rat model of an infected wound with Pseudomonas aeruginosa inoculation and compared with a reference commercial product. The sponges thus prepared provided a degree of open porosity that was comparable to or higher than that of the reference commercial product. Spectrometry analysis revealed that the cross-linked collagen sponge and reference commercial product sponge preserved their secondary collagen structure after 168 h while early accelerated degradation was observed with respect to the non-cross-linked collagen sponge. Gentamicin was released rapidly from all the sponges. Compared to those animals with gentamicin-containing sponges or gentamicin administered intramuscularly, the animals with the cross-linked collagen sponge without gentamicin exhibited marked clinical and laboratory infection signs. Both the administration routes (intramuscular and via gentamicin-containing sponges) provided similar gentamicin plasma levels. The resulting highly homogeneous product which was characterized by excellent structural and clinical properties proved effective in terms of the treatment of a surgical wound infection in a rat model. We demonstrated that all the gentamicin was released from the sponge and was absorbed in the systemic circulation. This is the first time that Cyprinus Carpio collagen has been used in the preparation of wound dressings. Thus, gentamicin-containing sponges provide a promising tool for the treatment and prevention of surgical site infections.

Draft genome sequences of three clinical isolates of teicoplanin-resistant Staphylococcus epidermidis from patients without prior exposure to glycopeptide antibiotics

OBJECTIVES

The aim of this study was to analyse the DNA sequences of three teicoplanin-resistant Staphylococcus epidermidis isolates collected from patients not previously treated with glycopeptide antibiotics.

METHODS

The minimum inhibitory concentrations (MICs) of 12 antibiotics, including teicoplanin and vancomycin, were determined by the broth microdilution method. Genomic DNA was isolated, was sequenced by HiSeqX paired-end sequencing and was assembled into draft genome sequences using MyPro pipeline.

RESULTS

Analysis of the draft genome sequences demonstrated that the teicoplanin-resistant S. epidermidis isolates belonged to multilocus sequence typing (MLST) sequence types ST5 and ST87 and encoded multiple antimicrobial resistance genes, including the methicillin resistance gene mecA.

CONCLUSIONS

This report highlights the risk of dissemination of S. epidermidis strains resistant to a wide range of clinically important antibiotics.