Viability of Selected Microorganisms in Non-Cytotoxic Aseptic Preparations

Abstract Background: Numerous ready-to-use parenteral solutions are aseptically prepared in pharmacy-based aseptic preparation units. Microbiological stability of the preparations is influenced by the cleanroom environment, the complexity of the aseptic process, conditions during administration and the microbiological vulnerability of the products. Object: The aim of the study was to evaluate the ability of four different pathogens related to hospital infections to grow in ready-to-use, non-cytotoxic parenteral products aseptically prepared in hospital pharmacies. Method: In four consecutive series the antimicrobial activity of the following products was tested: caspofungin 35 mg or 70 mg in 250 mL 0.9 % NaCl solution (NS), micafungin 0.5 mg/mL in NS, vancomycin 5 mg/mL in G5/G10, heparin-sodium 1 IE/mL in NS, epinephrine 0.02 mg/mL in G5, norepinephrine 0.01 mg/mL in G5, phenylephrine 0.1 mg/mL, KCl solution 0.8 mmol/mL, trace elements 1:1 in G5/G10, midazolam 1 mg/mL injection solution, tranexamic acid 100 mg/mL injection solution, 50 % glucose solution, SMOFlipid 20 % lipid emulsion, 1 % propofol injection. Nine milliliter aliquots of each test solution were inoculated with 1 mL suspension of selected strains, i. e. S. aureus, P. aeruginosa, E. faecium or C. albicans. Samples of the inoculated solutions were taken in predefined intervals up to 144 h and transferred to tryptic soy agar plates. The plates were incubated at 37 °C and colony forming units (CFUs) counted after 24 h for bacteria and after 72 h in the case of C. albicans. Results: Most of the tested preparations induced no growth inhibition of the tested organisms. The selected strains lost viability in preparations containing vancomycin, phenylephrine or midazolam after a period of a few hours or days. Glucose 50 % w/v solution generated antimicrobial activity against P. aeruginosa and C. albicans immediately after inoculation. In tranexamic acid solutions only P. aeruginosa lost viability after 48 h of inoculation. In the lipid containing emulsions, CFUs increased rapidly. Low pH values and high osmolality are probably the reason for growth inhibition in midazolam and 50 % glucose solutions, respectively. The antimicrobial activity of phenylephrine solutions is caused by the excipient sodium metabisulfite. Conclusion: The lack of antimicrobial properties of ready-to-use, non-cytotoxic solutions should be considered while determining the shelf-life of the products. Ready-to-use preparations should be kept refrigerated whenever possible to inhibit the multiplication of any contaminating organism.


Introduction
Today in most German hospitals various ready-to-use or ready-to-administer parenteral products are prepared in the hospital pharmacy departments. The portfolio of products encompasses cytotoxic as well as non-cytotoxic preparations. Typical non-cytotoxic product types are parenteral nutrition solutions, electrolytes, emergency drugs, analgesics, antibiotics, and antifungals. But the pharmacy based preparation service is never all-encompassing and a lot of preparations are still to be reconstituted in clinical areas. The quality-assured aseptic processing within pharmacies is performed following national (e. g. Apothekenbetriebsordnung [1], ADKA Guideline [2]) and international regulations and guidelines (PIC/S PE 010-3 [3], Ph.Eur.7.7 Monograph Pharmaceutical Preparations [4], Council of Europe Resolution CM/ResAP (2011)1 [5], USP Monograph <797> Pharmaceutical Compounding -Sterile Preparations [6]). By pharmacy-based aseptic processing in well-controlled environments and quality assurance driven procedures the risk of preparation errors and microbial contamination is considerably reduced. Sterility tests of extemporaneously and batchwise prepared products are limited by singularity of each preparation and time restraints. However the results of sterility tests, media fills, and environmental monitoring programs are useful for process validation. Preparation in uncontrolled environments such as clinical areas is associated with a higher potential for microbiological contamination and an increased risk of systemic infection after administration. For the individual product types different factors influencing the risk of microbial contamination and infection are discussed. The specific factors are complexity of the preparation, time between preparation and administration, extended periods of administration, elevated temperatures during administration by infusers and other ambulatory devices and last not least the growth potential of the preparation. Multiple additions of multiple components, mixing of large volumes and growth promoting qualities of the mixed components increase the risk of microbial contamination of the preparations and the risk of infection for the patients. If preparations are contaminated with microorganisms, viability and growth of the specific microorganisms in the specific preparations determine the infection risk for the patient. Previously we investigated the growth of microorganisms in antineoplastic drug preparations and found out that most preparations lack antimicrobial properties and monoclonal antibody preparations do not stimulate the growth of microorganisms [7][8][9][10][11]. From the literature it is known, that lipid emulsions support bacterial and fungal growth and bear an increased risk for iatrogenic infections [12][13][14][15][16][17][18][19][20]. Species-specific growth inhibition is reported for heparin 100 U/mL [21], midazolam [22] and local anaesthetics [23]. Recently the lack of growth inhibition of some vasopressor infusion solutions was published [24]. For a number of parenteral products prepared in our central intravenous additive service information about growth promotion or inhibition is missing. Moreover tests are done under different experimental conditions and extrapolation of the results is hardly possible. Therefore the aim of the study was to evaluate the ability of four different pathogens related to hospital infections (Staphylococcus aureus, Pseudomonas aeruginosa, Enterococcus faecium, Candida albicans) to grow in 17 ready-to-use solutions typically prepared in the aseptic preparation unit of the Pharmacy Department of the University Medical Center Mainz. The chosen test conditions promoted the growth of germs and simulated daily practice.

Material and methods
The viability of the selected microbes was tested by inoculation of 17 non-cytotoxic ready-to-use aseptic preparations belonging to the portfolio of the pharmacy based aseptic preparation service. The experiments were carried out in four consecutive series. In the first series the antimicrobial activity of caspofungin 35 mg or 70 mg in 250 mL 0.9 % NaCl solution (NS), micafungin 50 mg in 100 mL 0.9 % NS, heparin sodium 1 IE/mL in NS and 50 % w/v glucose solution was tested. In the second series adrenaline (epinephrine) 0.02 mg/mL in G5, noradrenaline (norepinephrine) 0.01 mg/mL in G5, trace elements in G5 1:1, and vancomycin 5 mg/mL in G5 were examined. In the third series trace elements 1:1 in G10, midazolam1 mg/mL, lipid emulsion 200 mg/mL and vancomycin 5 mg/mL in G10 were inspected. In the fourth series phenylephrine 0.1 mg/mL, 0.8 mmol/ mL potassium chloride solution, 1 % propofol injection emulsion, and tranexamic acid injection solution 100 mg/mL was tested. Details about the material used and characteristics of the test solutions are given in Table 1.

Preparation of inocula
The microorganisms used in this study were Staphylococcus aureus ATCC strain 6538, Pseudomonas aeruginosa ATCC strain 15442, Enterococcus faecium ATCC strain 6057, and Candida albicans ATCC strain 10231. The strains were cultivated at the Institute of Medical Microbiology and Hygiene of the University Medical Center Mainz, Germany and cultured on agar plates (BD Trypticase™ Soy Agar, BD Medical, Heidelberg, Germany, Lot 5019173, expiration date 12/04/2015) at 37°C for 24 h in the case of bacteria. C. albicans was cultured for 72 h at 37°C. The cultures were collected and suspended in 0.9 % NaCl solution. After that the density of bacterial suspensions was adjusted by using McFarland standards (0.5 for S. aureus and E. faecium, 0.2 for P. aeruginosa and 2.5 for C. albicans) to obtain suspensions of approximately 10 8 CFUs per milliliter. To achieve suspensions with a density of 10 5 CFU/mL, the suspensions were diluted with 0.9 % sodium chloride solution.

Preparation of samples and analysis
The tested product solutions were aseptically prepared in the pharmacy based centralized aseptic preparation unit at the University Medical Center Mainz, Germany in a cleanroom environment following Good Preparation Practice Guidelines. Physical and chemical stability of the preparations is proven for at least 5 days (see Table 1).

Results
Most of the tested aseptic preparations affected the growth of the test organisms in the same manner as the control solutions (water for injection, NS, G5, G10). The number of CFUs remained unchanged (E. faecium, C. albicans), decreased (S. aureus) or increased (P. aeruginosa) over a period of five days (see Tables 2-5).     These results reflect the species-specific capability of the microorganisms to survive and grow in nutrient-deficient solutions. However, the tested microorganisms lost viability in preparations containing vancomycin, phenylephrine, and midazolam after a period of a few hours or few days (see Figures 1 and 2). C. albicans rapidly lost viability in caspofungin and micafungin containing test solutions (see Figure 3). Thereby the proven antimicrobial activity of vancomycin and the echinocandins, which are used as an antibiotic and antifungals, respectively, confirmed the validity of the experimental design and the results.
Species-specific antibacterial activity was observed in tranexamic acid solutions. Only P. aeruginosa lost viability 48 h post inoculation, while the other strains tested remained viable. Glucose 50 % w/v solution also generated species-specific antibacterial activity against P. aeruginosa. Antifungal activity of glucose 50 % against C. albicans got already obvious 1 hour post inoculation. Caspofungin and micafungin exhibited strong antifungal activity as expected. Noteworthy, caspofungin containing test solutions in both concentrations exhibited antibacterial activity against S. aureus and E. faecium. Viability of P. aeruginosa was not affected (see Figure 3). Moreover, micafungin did not inhibit bacterial growth. In some of the test solutions C. albicans tends to grow (compare Table 5).
The lipid containing formulation of 1 % propofol emulsion and the 20 % SMOFlipid emulsion served as nutritive media for all selected microorganisms and the number of CFUs increased rapidly (see Figure 4).
Trace elements mixed with G5 or G10 stimulated the growth of P. aeruginosa and C. albicans, but had no influence on the growth of the tested Grampositive bacteria (see Tables 2-5).

Discussion
Viability and growth of microorganisms in pharmaceutical preparations is directed by extrinsic (e. g. temperature, oxygen) and intrinsic factors (e. g. type and concentration of ingredients). It is well known that refrigeration retards the growth of microorganisms. The experiments were performed to increase our knowledge about the intrinsic factors of ready-to-use parenteral preparations compounded in the pharmacy department except the drugs used in anticancer therapy. The portfolio of products encompasses different indications, e. g. Table 5: Viability of C. albicans in non-cytotoxic drug preparations and control solutions. parenteral nutrition, catecholamines, antibiotics, antifungals, anaesthesia and the use in different patient groups (paediatric patients, intensive care patients, operating theatre patients). These preparations contain different active substances and excipients and have different physicochemical characteristics. The experimental conditions of the intrinsic factors were chosen according to clinical practice. All preparations contained liquid water and some of them contained nutrient sources, minerals or trace elements. The extrinsic factors were the same throughout the study. As a reasonable compromise of typical temperature conditions in clinical practice and optimum growth temperature for pathogenic bacteria (37°C) the experiments were carried out at room temperature (22°C). The inoculum size was kept unchanged and simulated low level contamination. The experiments were performed in duplicate for each readyto-use preparation at each time interval. Each sample was tested in three degrees of dilution (in total 6 experiments). This allows the detection of the influence of the CFU concentration and of process errors. Because these are biological experiments, not the average but a representative value was chosen to be presented as result.
The growth inhibition detected in some preparations is not related to the active substance, but to the physicochemical parameters such as pH and osmolality. Bacterial growth is inhibited by low pH, while the optimum pH for the growth of most fungi is pH 5. Therefore, low pH values are most probably the reason for growth inhibition in midazolam injection solutions [20,22]. The pH values of diluted vancomycin infusion solutions amount to pH 3-5. That might be the reason for the growth inhibition of P. aeruginosa and C. albicans recognized in our experiments.
The phenylephrine injection preparation (pH 5) contains citric acid and sodium metabisulfite. The comparatively high concentration of sodium metabisulfite (2 mg/ mL) reduces the redox potential of the phenylephrine preparations, what may explain the observed antimicrobial activity. This assumption is confirmed by the fact that inhibition of microbial growth did not occur in the adrenaline and noradrenaline containing preparations containing small amounts of sodium metabisulfite (0.072 mg/mL). Similar findings were reported by Bostan et al. [24]. Among the tested catecholamine preparations only dobutamine preparations showed antimicrobial activity because of high sodium metabisulfite concentrations compared to adrenaline and noradrenaline preparations [24]. Notably, the stimulation of bacterial growth by catecholamines [25,26] is eliminated by the antioxidative excipients in the medicinal products. High osmolality is also known to inhibit microbial growth what explains the antimicrobial of glucose 50 % preparations [13,27]. Antifungal activity of glucose 50 % against C. albicans got already obvious 1 hour post inoculation and is commonly known. To our knowledge there is no information available about the antimicrobial activity of tranexamic acid containing preparations. The fact that growth inhibition was only given for P. aeruginosa suggests that the activity is substance specific.
According to the studies of Rosett et al. the antimicrobial activity of heparin is a result of the reduction of divalent cations from the growth media [21]. The lack of antimicrobial activity of the heparin 1 IE/mL containing preparations is to be explained by the low heparin concentration and the experimental conditions used. Obviously the low amount of heparin in our preparations was insufficient to bind the cations essential for bacterial growth. During in vivo experiments also high concentrations of heparin sodium lacked antibacterial activity against S. aureus [28]. The same explanation is applicable for the missing antimicrobial activity of the 0.8 molar potassium chloride preparations. Different concentrations and experimental designs lead to inconsistent results [29,30].
Fat emulsions like 1 % propofol injection and 20 % lipid emulsion are known to serve as growth medium for a number of microorganisms and as origin of bloodstream infections [31,32]. The lack of antimicrobial activity is related to the high value of pH (7-8.5) and the high lipid content serving as nutrient source. Fat emulsion serves as an excellent non-nitrogen energy source for a number of microorganisms, including bacteria and fungi. When microbial growth in five different commercially available lipid emulsions was tested no difference in growth patterns due to the nature of the oil or its concentration was observed [33]. SMOFlipid 20 % consists of fish oil, soybean oil, medium chain triglycerides, and olive oil and supported the growth of the selected bacteria and yeast in a similar manner. In both settings the number of CFUs rapidly increased to ≥ 10 6 over a period of less than 24 h after inoculation. Pure fat emulsions and lipid containing total parenteral nutrition solutions   are the most vulnerable preparations and should always be prepared under strict aseptic conditions and should not be stored or infused more than 12 h after preparation [33,27]. The species-specific growth promoting activity of the trace elements mixed with glucose solutions is not yet reported in the literature. But plausibility is given as P. aeruginosa and C. albicans grow in nutrient deficient solutions and microorganisms require also trace elements for their growth [34].

Conclusion
As most of the tested parenteral preparations did not generate antimicrobial activity, preparation should be done under strict aseptic conditions in order to avoid any microbial contamination. Furthermore the insufficient antimicrobial properties of ready-to-use solutions should be considered while determining the shelf-life of the products. Lipid containing preparations should be kept refrigerated whenever possible to inhibit the multiplication of any contaminating organism.
Judith Thiesen Department of Pharmacy, University Medical Center Mainz, Johannes Gutenberg-University, Langenbeckstraße 1, 55131 Mainz, Germany Judith Thiesen is working at the Pharmacy Department of the University Medical Center of Johannes Gutenberg-University Hospital in Mainz since 1997. In 2001 she completed her doctoral thesis entitled: Evidence-based optimization of parenteral drug application for oncological patients: incompatibilities-reducing infusion schemes, stability of ready-to-use parenteral solutions of camptothecin-derivatives and taxanes. Her special interests and research projects include aseptic drug preparation, quality control, total quality management as well as physicochemical and microbiological stability of parenteral drug solutions.

Irene Krämer
Department of Pharmacy, University Medical Center Mainz, Johannes Gutenberg-University, Langenbeckstraße 1, 55131 Mainz, Germany Irene Krämer is currently Director of the Pharmacy Department, University Medical Center, Johannes Gutenberg-University Hospital, Mainz and is also a Professor for clinical pharmacy at the Pharmacy School of Johannes Gutenberg-University. She completed her postdoctoral thesis in Pharmaceutical Technology entitled: Development, quality assurance, and optimization of ready-to-use parenteral solutions in the integrated cancer care concept. Her special interests include oncology pharmacy, infectious diseases, and aseptic drug preparation. She is doing research projects in the field of physicochemical and microbiological stability of cytotoxic drugs, compatibility of admixtures of nebulizer solutions, and monitoring of medication compliance.