Chemical Stability of Ceftazidime Compounded in Saline, Glycerin and Dexamethasone‐SP Solutions Stored at −20°C, 4°C and 25°C Over a 60 Day Period

ABSTRACT

Background

Chronic cases of canine otitis externa (OE) often develop infections with Pseudomonas aeruginosa (PA). Given the organism's high level of resistance, veterinary surgeons often turn to compounded solutions. Limited data describing the stability and potency of compounded ceftazidime (CAZ) solutions are available, which may affect clinical outcome.

Hypothesis/Objectives

To evaluate the chemical stability of compounded glycerin (GLY) and dexamethasone sodium phosphate (DEX‐SP) CAZ solutions in three different storage temperatures over a 60‐day period. Based on previous evaluations, CAZ concentrations would decrease with increased temperature and time.

Materials and Methods

Ceftazidime was compounded at 10 mg/mL with 100 mL 0.9% sodium chloride (NA + CAZ), 100 mL glycerin +0.9% sodium chloride (GLY + CAZ) and 100 mL dexamethasone sodium phosphate +0.9% sodium chloride (DEX‐SP + CAZ), stored at −20°C, 4°C and 25°C for 60 days. Mass spectrometry was used to analyse CAZ stability at specific time points (Day[D]0, D7, D14, D28, D60).

Results

Chemical stability of CAZ concentrations was affected by storage time, temperature and diluent. CAZ concentrations decreased over time with increased temperature; frozen CAZ concentrations remained stable over time for all solutions.

Conclusions and Clinical Relevance

Compounded CAZ stability varies by diluent, storage temperature and storage duration. NA + CAZ and DEX‐SP + CAZ solutions are stable for ≤ 28 days refrigerated and retain potency for ≥ 60 days if stored frozen. These solutions offer alternative options for treatment of PA OE.

Short abstract

Background: Chronic cases of canine otitis externa (OE) often develop infections with Pseudomonas aeruginosa (PA). Given the organism's high level of resistance, veterinary surgeons often turn to compounded solutions. Limited data describing the stability and potency of compounded ceftazidime (CAZ) solutions are available which may affect clinical outcome. Hypothesis/Objectives: To evaluate the chemical stability of compounded glycerin (GLY) and dexamethasone sodium phosphate (DEX‐SP) CAZ solutions in three different storage temperatures over a 60 day period. Based on previous evaluations, CAZ concentrations would decrease with increased temperature and time. Conclusions and Clinical Relevance: Compounded CAZ stability varies by diluent, storage temperature and storage duration. NA+CAZ and DEX‐SP+CAZ solutions are stable for ≤ 28 days refrigerated and retain potency for ≥ 60 days if stored frozen. These solutions offer alternative options for treatment of PA OE.

1. Introduction

Recurrent bacterial otitis externa (OE) is a common problem in dogs, with Pseudomonas aeruginosa (PA) infections frequently implicated [1, 2, 3]. The severity of these infections poses a therapeutic challenge given that PA is commonly resistant to a variety of antibiotics [4]. Many practitioners turn to compounded otic preparations containing injectable antibiotics owing to the limited number of efficacious commercially available otic solutions [5, 6]. Topical antibiotic solutions are preferred in most cases because higher antibiotic concentrations can be achieved locally. The main goals of treatment include removing biofilms, clearing infection and reducing inflammation [7, 8, 9]. The options for treatment of PA, however, often are limited to amikacin, marbofloxacin, gentamicin, polymyxin B, ceftazidime and ticarcillin [9].

Ceftazidime (CAZ), a third‐generation cephalosporin, has gained increased interest for the treatment of PA otitis, as it has high antimicrobial activity against multidrug‐resistant strains of PA [9, 10, 11]. Ceftazidime has been categorised as a critically important antimicrobial by the World Health Organization and should only be used in exceptional circumstances [12]. However, there are cases where, owing to the lack of availability or efficacy of other antimicrobials, ceftazidime is selected for treatment. The clinical applications of topical CAZ for PA otitis are not well‐documented, which may be a result, in part, of nonstandardisation of compounded preparations. In addition to having poor stability after reconstitution, CAZ is also known to be affected by pH and increased storage temperatures [13, 14, 15, 16]. Owing to the severe ear canal inflammation commonly seen in Pseudomonas otitis infections, topical glucocorticoids such as dexamethasone sodium phosphate (DEX‐SP) are often used for treatment [17]. Additionally, in an effort to improve stability, a glycerinated CAZ solution has been offered by a widely used commercial compounding pharmacy in the USA, yet its stability with CAZ is not published [18]. In the veterinary literature, only one study has evaluated stability and its impact on the efficacy of CAZ when compounded with commercially available ear flushes [19]. To the best of the authors' knowledge, there are no published data on the stability of CAZ when compounded with glucocorticoids or glycerinated solutions.

The objective of this study was to determine the chemical stability of CAZ when compounded with saline, glycerin and injectable DEX‐SP under three storage temperatures over a 60‐day period. Our hypothesis was that CAZ would be the most stable at lower temperatures and that stability would degrade over a 60‐day period.

2. Materials and Methods

2.1. Ceftazidime Solutions

Three compounded solutions were reconstituted for evaluation based on previous clinical recommendations for the treatment of Pseudomonas otitis [4, 9]. Diluents (see Appendix S1) and the volumes used were as follows: NA—0.9% sodium chloride (NaCl) 100 mL (Dechra Veterinary Products); GLY—99.5% glycerin 100 mL (Good Neighbour Pharmacy) + 0.9% NaCl 20 mL; and DEX‐SP—4 mg/mL (50 mL DEX‐SP mixed with 50 mL of 0.9% sodium chloride to make a 2 mg/mL DEX‐SP solution) (VetOne). A commercially available dry‐powdered mixture of 1 g ceftazidime pentahydrate and 118 mg/g sodium carbonate (1 g CAZ) for injection (WG Critical Care) was used to formulate three solutions. The compounded solutions were prepared as follows:

1. NA + CAZ: 0.9% NaCl 100 mL + 1 g CAZ (1 g/100 mL).

2. GLY + CAZ: 99.5% GLY 100 mL + 0.9% NaCl 20 mL + 1 g CAZ (1 g/120 mL).

3. DEX‐SP + CAZ: 4 mg/mL DEX‐SP 100 mL + 1 g CAZ (1 g/100 mL).

In order to minimise bacterial contamination, sterile gloves and syringes were used to transfer and mix solutions. An approximately 1% solution was made by dissolving CAZ in each diluent by gently agitating the solutions by hand for 60 s. The final concentrations of the solution were based upon the volume of each individual diluent (1% for NA + CAZ, 0.83% for GLY + CAZ and 1% for DEX‐SP + CAZ). Based on a previously published protocol, each compounded solution had two duplicate lots made; each lot was separated into 1 mL aliquots and stored in sterile 1.5 mL polypropylene tubes that were then divided and stored in standard freezer boxes (Thermo Fisher Scientific Inc.) at various temperatures (−20°C, 4°C, 25°C) verified by Fisherbrand Traceable Digital Thermometers with Short Sensors (Thermo Fisher Scientific Inc.) within 30 min of preparation [19]. On Day (D)0, D7, D14, D28 and D60, three samples from each lot and storage temperature were randomly selected for evaluation. Each diluent (NA, GLY, DEX‐SP) was stored in individual 1.5 mL polypropylene tubes at each temperature to serve as controls. The controls also were analysed at each time point.

2.2. Chemical Stability

Ceftazidime formulations in the three vehicles (NA, GLY and DEX‐SP) at 10 mg/mL and stored under three conditions (−20°C, 4°C and 25°C) were analysed for CAZ concentrations using liquid chromatography and tandem mass spectrometry (LC–MS/MS) at D0 (baseline), D7, D14, D28 and D60. A Sciex 6500 Q‐TRAP triple quadrupole mass spectrometer (Applied Biosystems Inc.) coupled to an LC‐30ad SIL HPLC liquid chromatography system with an integrated autosampler (Shimadzu Scientific Instruments Inc.) was used for analysis.

2.3. Standard and Quality Control Preparations

The CAZ powder and internal standard cefotaxime (CEFO) stocks (10 mm in 1 mL DMSO) were obtained from Med Chem Express (HY‐A0088A CEFO lot 83966; HY‐B0593 CAZ lot 13829) as lyophilised powder and stored according to manufacturer recommendations. Stock solutions of CAZ were made in NA, GLY and DEX‐SP at 6.9, 3.5 and 3.4 mg/mL, respectively. From stocks, a working solution of 1.0 mg/mL was made for each formulation in 25/75 water/acetonitrile diluent (for NA and DEX‐SP) or water (for GLY). A five‐point calibration curve was prepared in diluent for each formulation using CAZ at 10‐fold concentrations: 50, 20, 10, 5 and 2.5 μg/mL. 50 μL of each 10× standard dilution of CAZ plus 450 μL of diluent containing 2 μg/mL CEFO as an internal standard (IS) was prepared in a 96 well deepwell plate for use as a standard calibration curve with final concentrations of 5000, 2000, 1000, 500 and 250 ng/mL. Twelve samples were prepared at six replicates each for low (500 ng/mL) and high (2500 ng/mL) quality‐control samples in the same manner. The deepwell plate was sealed, shaken at 750 rpm on a plate shaker for 20 min at 25°C (room temperature [RT]) and protected from light.

2.4. Sample Processing

Investigative samples were submitted as individual 1.5 mL polypropylene tubes containing 1.0 mL from each designated formulation. Six replicates from each formulation from each storage temperature and sample time point were analysed. Each submitted sample was made up to a 10 mg/mL solution in its respective vehicle. This formulation was serially diluted 10‐fold three times (1000×) in diluent (25% lab grade deionised water +75% acetonitrile) with a fourth 10× final dilution containing internal standard at 2 μg/mL in diluent.

2.5. Liquid Chromatography (LC) & Mass Spectrometry Conditions and Analysis

The methodology was based on previous studies with a slight modification to accommodate different instrumentation [19, 20]. A 10 μL aliquot of each diluted sample was injected into a high‐performance liquid chromatography (HPLC) unit onto a Waters HILIC, 4.6 × 50, 5 μm column at a flow rate of 800 μL/min and eluted over 4 min. The column oven was manually set to 30°C. The LC gradient conditions utilised an aqueous mobile phase of 0.1% formic acid in water (Buffer A) and an organic mobile phase of 0.1% formic acid in acetonitrile (Buffer B). Scheduled run time over 4 min started with 15% Buffer B, 85% Buffer A for 0.1 min; at 2.0 min, 95% Buffer B, 5% Buffer A, held for 0.5 min; at 3.0 min, 15% Buffer B, 85% Buffer A, held for 1 min. Instrument parameters included resolution set to unit; Curtain Gas 35; Collision Gas Medium; Ion Spray Voltage 5500; Source Temperature 300°C; Ion Source Gases set at 20. Mass‐to‐charge ratios for CAZ and CEFO (IS) were 547–467.9 and 456–277, respectively. Compound‐dependent parameters for both CAZ and IS were determined using the optimisation algorithm included in the sciex analyst software package. A diverter valve moved extraneous eluent to waste from 0 to 0.5 min and from 1.6 to 4 min. Peak elutions for CAZ and CEFO were 1.21 and 0.929 min, respectively. Data acquisition was performed using sciex analyst software v1.7.3. Quantitation analysis of CAZ was performed using a linear fit to calibration with a weighted least square (1/x ²) regression using the five standards.

The upper and lower limits of quantitation were 5000 and 250 ng/mL, respectively. Low (500 ng/mL) and high (2500 ng/mL) quality control samples were performed in six replicates for each formulation, and 98% of quality controls passed acceptance criteria of > 85% accuracy.

2.6. pH Determination

The pH of the three compounded solutions and individual constituents (NA, GLY, DEX‐SP) was determined using a Model 215 pH meter (Denver Instrument Co.) and hydration papers (Micro Essential Laboratory). For the compounded solutions, three aliquots were combined from each lot and storage temperature for pH analysis.

2.7. Statistical Methods

A D'Agostino–Pearson analysis was performed to evaluate the normality of the data. The coefficient of variation (CV) for mass spectrometry‐based measurements was calculated from replicates for each concentration at each storage temperature and time point; CV < 15% was deemed acceptable analytical precision.

In order to simplify comparisons between the solutions, the concentrations at each storage temperature and time point were converted to the percentage of CAZ recovered (PeCR) using the following equation:

$$\text{PeCR}=\frac{\text{Recovered}\mathrm{CAZ}\left(\mathrm{mg}/\mathrm{mL}\right)}{\text{Initial}\mathrm{CAZ}\text{concentration}\left(\mathrm{mg}/\mathrm{mL}\right)}\times 100$$

The PeCR conversions were necessary for analyses because the starting concentrations of CAZ varied between the three formulations. A two‐way ANOVA with Tukey's correction for multiple comparisons was used to compare the PeCR of the three compounded solutions at each temperature and time point. p < 0.0001 was considered statistically significant. Statistical analyses were performed using commercially available software (Prism 10; Graphpad Software LLC).

3. Results

3.1. Chemical Stability

The CV of CAZ measurements ranged from 1.9% to 13.9%. The diluent, storage temperature and storage time all individually had significant effects on the PeCRs (Table 1; Figure 1; and Stability data in Appendix S1). Frozen samples remained stable or had a greater PeCR after D0, while all refrigerated and RT samples, regardless of diluent, showed degradation over time. For all diluents, RT samples showed the greatest degradation of CAZ in comparison to refrigerated and frozen samples. Between frozen samples, there was no significant difference in PeCR for NA + CAZ and DEX‐SP + CAZ until D60 (p > 0.0001). By D7, GLY + CAZ had significantly lower PeCRs at all storage temperatures compared to NA + CAZ and DEX‐SP + CAZ (p < 0.0001). Likewise, the PeCRs for DEX‐SP + CAZ were significantly lower than NA + CAZ for RT and refrigerated samples, while no significant difference was seen between frozen samples except at D60.

TABLE 1.

Mean percentage of ceftazidime recovered (PeCR) at 25°C, 4°C and −20°C over time, compared to Day (D) 0. Data presented as mean ± standard deviation.

Solution	Storage temperature	D0	D7	D14	D28	D60
NA + CAZ	25°C	102.6 ± 0.14	84.3 ± 0.03	65.1 ± 0.06	43.1 ± 0.12	13.4 ± 0.13
	4°C	100.7 ± 0.11	106.5 ± 0.02	103.2 ± 0.06	94.3 ± 0.05	82.7 ± 0.06
	20°C	101.8 ± 0.03	111.1 ± 0.02	103.9 ± 0.1	107.9 ± 0.05	114.9 ± 0.07
GLY + CAZ	25°C	100.1 ± 0.02	13.0 ± 0.05	2.7 ± 0.05	a	a
	4°C	100.9 ± 0.03	67.2 ± 0.03	48.4 ± 0.03	21.0 ± 0.07	4.9 ± 0.08
	20°C	101.8 ± 0.05	101.0 ± 0.04	97.9 ± 0.04	98.0 ± 0.06	93.9 ± 0.11
DEX‐SP + CAZ	25°C	101.6 ± 0.09	73.5 ± 0.05	65.6 ± 0.07	38.4 ± 0.09	16.2 ± 0.07
	4°C	100.2 ± 0.08	107.7 ± 0.03	102.8 ± 0.04	92.9 ± 0.13	67.2 ± 0.21
	20°C	100.3 ± 0.12	109.9 ± 0.09	112.9 ± 0.08	114.2 ± 0.06	106.3 ± 0.09

Note: NA + CAZ, 0.9% NaCl 100 mL + 1 g ceftazidime compounded solution (CAZ); GLY + CAZ, 99.5% glycerin 100 mL + 0.9% NaCl 20 mL + 1 g CAZ compounded solution; DEX‐SP + CAZ, 4 mg/mL dexamethasone sodium phosphate (50 mL DEX‐SP/50 mL saline) + 1 g CAZ compounded solution. Red text indicates PeCRs that are outside the USP standard range for compounded medications.

^a Below the limit of quantitation (2500 ng/mL) of the mass spectrometry assay.

FIGURE 1.

Percentage of ceftazidime recovered (PeCR) over time at three different storage temperatures. 25°C represents room temperature, 4°C is refrigerated temperature and −20°C is frozen temperature. All temperatures are with respect to time. NA + CAZ, 0.9% NaCl 100 mL + 1 g ceftazidime compounded solution (CAZ); GLY + CAZ, 99.5% glycerin 100 mL + 0.9% NaCl 20 mL + 1 g CAZ compounded solution; DEX‐SP + CAZ, 4 mg/mL dexamethasone sodium phosphate (50 mL Dex‐SP/50 mL saline) + 1 g CAZ compounded solution.

3.2. pH Determination

The initial pHs were 5.4 for NA, 4.5 for GLY and 8.1 for DEX‐SP. When the diluents were compounded with CAZ, a slight increase in the pH was noted for NA + CAZ (5.5), and a decrease in pH was noted for GLY + CAZ (4.0) and DEX + CAZ (5.5) (see pH values in Appendix S1). There was no change in pH for any of the compounded solutions at any time point or storage temperature. All compounded solutions remained acidic throughout the duration of the study.

4. Discussion

Successful treatment of Pseudomonas otitis is often complex and involves multimodal therapy such as biofilm removal, anti‐inflammatory therapy, treatment of the underlying cause and appropriate antimicrobial therapy concurrently [1, 9]. Given this complexity, it is important to ensure stability of these products when formulating extralabel otic solutions. The use of compounded products is controversial in both human and veterinary medicine owing to the absence of studies confirming stability and efficacy. In‐house compounded otic medicated solutions often are used; however, commercially available antibiotic therapy for Pseudomonas otitis is limited. A previous study evaluated the stability of CAZ compounded with NA and showed no significant change at D28 between refrigerated and frozen samples [19]. Based on those findings, we chose to extend the duration of storage to 60 days because Pseudomonas otitis infection often takes > 30 days to resolve [17, 19, 21, 22, 23]. A 1% CAZ solution contains 1 g CAZ/100 mL and could be used for 50 days if instilled twice daily for a single ear approximating 0.5 mL per application [24].

In this study, the chemical stability of compounded CAZ solutions varied based on the diluent selected, storage temperature and storage time. Increasing temperature and time had the most significant effects on each compounded solution. It is important to note that for compounded medications prepared in the US, there is currently limited oversight, with few enforced regulations or published guidelines in place. By United States Pharmacopoeia (USP) definition, the chemical stability of a compounded product should typically contain 90%–110% of the intended concentration of its primary active ingredient (CAZ) from D0 to be considered suitable for use [25]. By D7 in this study, all 25°C solutions fell below this standard, with GLY + CAZ being the most significantly affected. With regard to 4°C solutions, NA + CAZ and DEX‐SP + CAZ maintained chemical stability up to D28, while GLY + CAZ lost chemical stability at or before D7. All −20°C solutions remained within or just slightly above this recommended range. Specifically, there were four concentrations within the frozen samples that exceeded the 110% USP recommended concentration, although this was not associated with time. All frozen concentrations remained within the accepted US Federal Drug Administration (FDA) range for valid chromatographic assays which allows for ±15% variance when compared with original sample concentration [26]. Aside from sample variance, the reasons for the increase in concentrations are unknown; increased concentrations have been observed in previous stability studies. Concentrations of amikacin stored in plastic liquid bottles increased on D56, and this change was thought to be related to evaporation [27]. A previous study assessing the stability of CAZ with prolonged storage at −20°C found that some solutions reached ≤ 123% of the original concentration [28]. It is possible that ‘freeze‐concentration’ can increase solute concentrations when ice crystals form in solution, although we did not directly observe any remaining crystals after thawing [29, 30].

As seen in the earlier study by Hoff et al. [19], we found that CAZ was the least stable at 25°C across all diluents and time points. Overall, as storage temperatures decreased, stability was maintained across all compounded solutions. For NA + CAZ and DEX‐SP + CAZ solutions, there was a significant difference between PeCRs at D28 and D60 between refrigerated and frozen samples, while there was no significant difference at D7 or D14. GLY + CAZ was the least stable solution, showing significant degradation of CAZ at 25°C and 4°C by D7.

A pH outside the 4.5–6.5 range has been reported in previous studies to accelerate the degradation of CAZ [13, 14, 15, 16]. In the present study, pH did not fall outside this range. This indicates that other factors, including storage duration and storage temperature, might influence the stability of CAZ.

An important component of Pseudomonas otitis treatment is anti‐inflammatory therapy. Evaluating the stability of CAZ with DEX‐SP is important because these cases are often severely painful and benefit from glucocorticoid therapy to reduce inflammation within the ear canal [17]. A previous paper hypothesised that because DEX‐SP has a pH of 7.0–8.0, which is above the optimal range for CAZ, combining the two solutions together may cause the solution to lose stability [19]. The DEX‐SP + CAZ solutions at all time points and temperatures remained at a constant pH of 5.5, which is within the normal range for CAZ. Our results show that DEX‐SP + CAZ stability was maintained for ≤ 28 days at 4°C and retained potency for ≤ 60 days at −20°C. This suggests that the inclusion of dexamethasone to CAZ would be likely not to affect stability and could be a valuable adjunctive therapy in the treatment of Pseudomonas otitis.

This study had some limitations. First, because the solutions were made up into two separate lots and individually placed into 1 mL polypropylene tubes in hospital rather than in a manufacturing setting, there was likely to have been minor variation between the samples. This study also was conducted in vitro so we cannot state how effective the compounded solutions would be in vivo.

Future studies could explore other facets of stability, including microbiological, therapeutic and toxicological factors. Additionally, the clinical value of topical glucocorticoids in the treatment of OE remains commonly used in veterinary medicine, yet is largely unexplored in controlled clinical studies.

In conclusion, this study showed that CAZ when compounded with 0.9% NA or DEX‐SP 4 mg/mL is chemically stable for 28 days when kept refrigerated and retains potency for ≥ 60 days if kept at a minimum of −20°C. Glycerinated CAZ solutions are not recommended for use because they did not maintain stability at room or refrigerated temperatures at any time point. Storage at RT for any of these compounded solutions is not recommended given that they all degraded after D0.

Author Contributions

Conceptualization: M.S., J.B., S.H., and D.G.; Methodology: M.S., K.B., J.B., S.H., and J.B.D.; Investigation: M.S. and W.T.; Writing – Original Draft: M.S., J.B., and S.H.; Writing Review and Editing: M.S., K.B., J.B., and S.H.; Funding Acquisition: M.S., J.B., and D.G.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1. Stability data: Chemical stability of CAZ in the solutions (NA, GLY and DEX‐SP) was determined using liquid chromatography and tandem mass spectrometry. Three samples were randomly selected from each solution and temperature (for both lots 1 and 2) for analysis at Day (D)0, D7, D14, D28 and D60. Pink indicates 25°C samples, green indicates 4°C samples and blue indicates −20°C samples for analysis.

Batch numbers for diluents used for compounding study solutions.

pH values: pH values for each compounded solution at each temperature and time point. The pH of freshly constituted CAZ solutions ranged from 5 to 8 per the products’package insert label.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.