Study on the Preparation Method of Quality-Assured In-Hospital Drug Formulation for Children—A Multi-Institutional Collaborative Study

The quality-assured preparation of crushed and diluted preparations for children is a challenge. In this study, a multicenter study was conducted to validate the preparation method for the quality assurance of baclofen powder, clonidine powder, and hydrocortisone powder prepared from tablets according to a previously established method. In-hospital preparations were prepared at five medical facilities under different crushing and mixing conditions. After storage in closed bottles, in-use bottles, and laminated paper for 120 days, ingredients stability, drug elution, and content uniformity after packaging were evaluated. All three ingredients were maintained at between 90% and 110% of their initial content for 120 days under packaging conditions of 25 ± 2 °C and 60 ± 5% relative humidity, with no change in dissolution in all formulations made at all five facilities. The content uniformity was also acceptable. The established method may contribute to quality-assured pediatric dosage form modification.


Introduction
In pediatric medical practice in Japan, tablets are often crushed, or capsules decapsulated in the pharmaceutical department of medical facilities when oral formulations are not flexible for dose adjustment or when tablets cannot be administered as supplied without modification [1][2][3]. With the increasing demand for drug information, information on the stability of tablets after crushing is sometimes provided by the pharmaceutical companies that manufacture and sell the tablets. On the other hand, in actual medical practice, tablets are crushed or decapsulated prior to dispensing as an in-hospital preparation, rather than being crushed when necessary. In pediatric patients, where the number of crushed and decapsulated tablets is high in relation to the total number of oral prescriptions, in-hospital preparation as a ready-to-use product is essential to improve operational efficiency [1]. Many in-hospital pre-products often lack quality assurance, such as the assurance of long-term stability and ingredient dissolution. The National Center for Child Health and Development (NCCHD) has been studying the standardization of preparation methods to ensure the quality of in-hospital preparation for pediatric drug therapy, and has examined preparation methods for baclofen, hydrocortisone, and clonidine [4][5][6]. Although the quality of the preparations prepared at the Department of Pharmacy of NCCHD was confirmed to be assured, the quality of products prepared in different environments and by different methods at other facilities is unclear. The purpose of this study was to confirm the robustness and validity of the standard protocol for in-hospital formulations by evaluating the quality of premixed products prepared at medical facilities other than NCCHD using the published preparation methods or under different preparation conditions. Additionally, in particular, the uniformity of drug content after packaging is required because powder medicines in Japan are generally packaged in single doses by automatic packaging machines and provided to patients in sachets. Although there have been no detailed physical properties studies on the post-packaging uniformity of powders, it is generally reported that in the pharmaceutical manufacturing process, the particle size of a powder affects the uniformity of the composition of the main ingredients and excipients when adhering to equipment and filling the equipment [7][8][9][10]. Therefore, different crushing and mixing conditions may affect uniformity after dispensing and during weighing. We attempted to conduct an additional study on this Japan-specific issue of uniformity after packaging as a particularly important research item because of its impact on clinical efficacy.

Preparation Method
To prepare the powder formulations, Gabalon ® 10 mg 100 tablets (Alfresa Corporation, Osaka, Japan) was used for baclofen powder, and Catapres ® 0.075 mg 100 tablets (Medical Parkland K.K., Tokyo, Japan) was used for clonidine hydrochloride. Hydrocortisone was supplied by Cortril Tablets 10 mg 100 tablets (Pfizer Japan Inc., Tokyo, Japan). After crushing and sieving, the product was diluted to 10 mg/g, 0.2 mg/g, and 20 mg/g, respectively, using crystalline lactose (EFC Lactose Whey, Viatris Inc., Tokyo, Japan) for dilution. The crushing and mixing conditions of each preparation at each of the five institutions are shown in Table 1. An automatic tablet mill was used for tablet crushing at all five medical institutions. Two facilities each used the same crushing conditions, either 14,500 rpm or 6000 rpm. The remaining facility crushed at 16,320 rpm. The crushed tablets were sieved through a sieve with a 300 µm (one facility) or 500 µm (four facilities) aperture. Manual mixing with a pestle and mortar was performed at two medical facilities and mixing with an automatic mixer was used at three facilities.

Stability Tests
For drug stability, the samples were obtained from each storage condition at days 0, 30, 60, 90, and 120, diluted according to previously published methods [4][5][6] and analyzed using an HPLC system Ultimate 3000 (Thermo Fisher Scientific, Tokyo, Japan) containing an autosampler, a column oven, and a diode array detector ( Table 2). The concentration changes were calculated as (measured concentration/initial concentration) × 100 (%), and within 10% of the initial concentration were considered acceptable changes [12]. Table 2. Assay for each target drug.

Assay Validation
Linearity was determined by triplicate and the concentration ranges were set as follows: 2.0 to 20.0 µg/mL for baclofen, 0.01 to 1.0 µg/mL for baclofen impurities, 2.0 to 20.0 µg/mL for clonidine hydrochloride, 0.01 to 1.0 µg/mL for 2,6-DCA, 0.5 to 100 µg/mL for hydrocortisone, and 0.05 to 10 µg/mL for 10 hydrocortisone impurities; six concentration levels were set for all drugs. Validation was performed to ensure that sample concentrations were within the linear analyte response range. Calibration standard concentrations were back-calculated to verify the suitability of the standard calibration curves. Standard calibration curves and correlation coefficients were obtained using a weighted linear regression method (weighted coefficient: 1/x 2 ). The calibration curve was considered acceptable for measurement with the LC-DAD system if the correlation coefficient was 0.99 or higher. The accuracy was then evaluated in terms of reproducibility (within day) and intermediate accuracy (between day). Data were presented as mean values and relative standard errors. Reproducibility was evaluated by six repeated injections of a standard solution spiked at 10 µg/mL. Intermediate precision was assessed with six repeat injections on three different days. Accuracy was calculated as (measured concentration/nominal concentration) × 100 (%). Statistical analysis was performed using one-way analysis of variance and limits of detection and quantification calculated using ICH guidelines. Each calibration curve was stored at −80 • C, as these samples were also used for quality control (QC). The standard deviations of the response (Sy) of the calibration curve and the slope (S) of the calibration curve were used to determine the lower limit of detection (LOD). The standard deviation of the response was determined from the standard deviation of the y-intercept of the regression line. The lower limit of quantitation (LLOQ) was obtained by multiplying Sy/S by 10.

Drug Dissolution Tests
Dissolution tests were performed under the same conditions as previously reported [4][5][6] according to ICH Japanese Pharmacopoeia 17.6.10. (paddle method; NTR-6400AC; Toyama Sangyo, Tokyo, Japan) using 900 mL of dissolution solution stirred at 37 • C ± 0.5 • C and 50 rpm. Mean dissolution rates were compared to those of day 0 formulation.

Uniformity of Drug Product Content Test
Since the crushing conditions at each medical institution were different, the particle size of the prepared formulation would be different, which may affect the uniformity of drug content after the packaging of the powder. To confirm the uniformity of the content of the drug product, we used an automatic dispersing machine (Crestage-Pro2, Takazono, Tokyo, Japan) to divide each package into 30 packages, each containing 0.5 g. Ten of these packages were then taken and the content of the ingredients was determined. The determination value was then calculated according to the following formula, and if the acceptance value did not exceed 15.0 the product was considered to be acceptable [16].

Results
The results of the ingredient stability tests, changes in the dissolution properties of each formulation, and the uniformity of drug contents after packaging made at five medical facilities with different crushing and mixing conditions were shown.

Stability Study
The ingredient stability of the in-hospital preparations of the study drugs, baclofen, clonidine hydrochloride, and hydrocortisone, was examined under each of the three storage conditions (closed bottle, in-use bottle, and packaging) after being prepared at the five medical facilities and stored for 120 days. The ingredients content was within specification, ranging from 90.0% to 110.0% of the initial concentration (Table 3) [12].

Impurities Studies
Baclofen impurities A and B [13] and 2,6-DCA (Clonidine impurity-C), an impurity for clonidine hydrochloride [14], were not detected in the samples after any of the storage conditions or the storage period. For hydrocortisone, small amounts of impurities B (cortisone) and G (hydrocortisone-21-aldehyde) were found in all preparations prepared at the five medical facilities, regardless of storage conditions, in samples taken on day 120 as previously reported [6], but relative concentrations were less than 0.05% of hydrocortisone. No other impurities were detected.

Dissolution Tests
Dissolution profiles of all components were comparable to those of day 0 immediately after preparation under all storage conditions for the formulations prepared at all five medical facilities, and no changes were observed with long-term storage (see Supplementary Materials, Figures S1-S3).

Uniformity of Drug Content Test
The formulations prepared at each medical institution were divided into 30 packets, and the component content of 10 of these packets was quantified and a determination value was calculated. The content uniformity values for each ingredient after preparation at the five institutions were all less than 15 (Table 4).

Discussion
The results of this study suggest that the standard protocols discussed in this study will enable the provision of quality-assured in-hospital preparations of baclofen, clonidine hydrochloride, and hydrocortisone to pediatric patients.
The in-hospital preparations were prepared at five medical facilities under different crushing and mixing conditions, and all three ingredients remained stable up to 120 days, similar to the preparations made in the Department of Pharmacy at the NCCHD. The generation of impurities was also similar. No change in dissolution was observed after 120 days of storage. In our previous study of the quality evaluation of in-hospital formulations, the uniformity of formulation content after dispensing in a dispensing machine was not sufficiently evaluated. In this study, we conducted a formulation content uniformity test in accordance with the regulations stipulated in the Japanese Pharmacopoeia, and were able to confirm that the formulation met the criteria for content uniformity after packaging. Uniformity was maintained after packaging, regardless of the crushing method.
In Japan, tablet crushers for crushing tablets and mixers or mortars for mixing crushed powder are commonly available for changing the dosage form for pediatric use, and it is considered possible to prepare in-hospital formulations with assured quality by using the method described in this study. Although the storage conditions were extremely limited, to 25 • C ± 5 • C and 60% ± 5%RH in a light-shielded environment, this is considered feasible in a clinical setting where pediatric care is available.
All three formulations studied in this study are frequently crushed and administered to the pediatric population. Since there are no appropriate pediatric formulations in the world, many studies have been conducted on compounding these three drugs [14,[16][17][18][19][20][21][22][23][24][25], and there have been reports of errors in preparation and incidents involving patients during the compounding process [26][27][28][29][30]. Therefore, in addition to assuring quality, there is an urgent need to ensure safety through the standardization of formulation preparation. Some other countries have established protocols for standardized preparation methods and have normalized in-hospital formulations [31][32][33][34]. The development of dose-adjustable formulations from pharmaceutical companies is highly warranted. In addition to the three drugs examined in this study, there are many other drugs that have been compounded in the pediatric setting, and it is believed that drug therapy is currently being administered without an assurance of quality (or based only on limited information provided by pharmaceutical companies). It is difficult to request pharmaceutical companies to develop all drugs. Similar studies for other drugs may be needed to continue improving the environment for drug use in pediatric pharmacotherapy.
In addition, this study did not confirm changes in the crystal system using powder X-ray diffraction and differential scanning calorimetry. At least, no significant changes in dissolution were observed in the preparation method and after long-term storage, suggesting that there were no physical changes that could affect the dissolution. Limitations of this study include the following: (1) The stability, homogeneity, and dissolution of the drug product were only verified, and it is not clear whether bioequivalence is ensured.
(2) The quality may not be guaranteed when prepared with different crushing and mixing equipment and with different concentrations and raw materials (e.g., generic drugs). (3) The safety and efficacy of the drug product should be investigated separately. Efforts should continue to be made to ensure the quality of compounding in pediatric drug therapy, which is currently being conducted at each medical facility. Baclofen, clonidine, and hydrocortisone, which were selected as model drugs in this study, are crushed in many pediatric facilities in Japan, and excipients such as lactose are added and provided to patients. Similarly, they are frequently crushed, compounded in the pharmacy, divided by a dispensing machine as a long-term drug, and administered. Although only three drugs were examined in this study, there are many drugs whose quality should be ensured. Although quality standardization efforts are underway outside of Japan, we believe that the quality evaluation of more drugs will continue to be essential.

Conclusions
The results of this study suggest that compliance with the procedures established in the previous study will enable the supply of uniform and quality-assured in-hospital formulations.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/children10071190/s1, Figure S1. Dissolution profiles of baclofen hospital formulation in water stored under three conditions (closed bottles (A), bottles in use (B), and laminated paper (C)). Open circles indicate dissolution rate at day 0; closed circles indicate dissolution rate at day 120. Figure S2. Dissolution profiles of clonidine hospital formulations stored under three conditions in water ((A), (B), and (C) are the same as in Figure S1). Figure S3. Dissolution profiles of hydrocortisone hospital formulation in water stored under three conditions ((A), (B), and (C) are the same as in Figure S1).