Nafcillin sodium was purchased from Sigma Co., USA. Normal saline (NS, 0.9% NaCl in water) and 5% dextrose (D₅W, injectable 5% dextrose water) were purchased from Choongwae Pharmaceutical Co. Ltd., South Korea. Disposable silicon balloon infusion device (Accufuser^®) was obtained from Woo Young Medical Co., Ltd, South Korea. Potassium phosphate monobasic was purchased from Sigma Co., USA, and all the chemicals for HPLC analysis were all HPLC-grade and were prepared immediately before use.

To prepare the test samples, the appropriate amounts of nafcillin sodium were added to a portion of the infusion solutions and were brought to a final volume of 100 mL with NS and D₅W. The test solutions were packed in sterile Accufuser^® system for testing. All manipulations were aseptically performed under laminar airflow in a biological safety cabinet. The nominal nafcillin sodium concentration for the sample used in this testing was 2.5 mg/mL. The Accufuser^® systems were each filled with 100 mL of drug solution. Triplicate test solutions were prepared and stored in RT and CT conditions. After the Accufuser^® systems were filled, an extension-tubing set was attached and primed. About 3 mL of drug solutions was collected in a 4 mL glass vial with a Teflon-lined screw cap and stored in a freezer at −70˚C until the analyzing process. One sample (approximately 3 mL) was obtained from the distal end of the tubing after 48 h, 1, 2, 4, and 6 weeks at RT and CT and stored in the manner described for the time-zero samples (Table 1).

The physical stability of the nafcillin sodium solutions were assessed by visual examination and HPLC analysis. Visual examinations were performed in normal diffuse fluorescent room light with unaided eye and high-intensity monodirectional light. The pH of the solutions was measured by a stainless electrode pH meter (Thermo Scientific Co., MA, USA). The drug concentrations were determined using a stability-indicating HPLC assay method based on several references [4-7]. The HPLC system [4, 7] consisted of an isocratic solvent delivery pump (Model 515, Waters Scientific Co. USA) which pumped a mixture (v/v, 65/35, pH 3.0) of acetonitrile (ACN, Sigma Co. USA) in 0.05M potassium phosphate (Sigma Co. USA)

Table 1. Study designs for the stability testing of nafcillin sodium solutions in Accufuser^® system.

^*O: processed sample, RT: room temperature, CT: cold temperature, NS: normal saline, D₅W: 5% dextrose water.

as a mobile phase through a Capsell Pak C₁₈ UG120 (4.6 × 150 mm, 5 μm, Shiseido Co., Japan) column at 1200 µL/min. The ratio of ACN to 0.05 M potassium phosphate (65/35) was held constant during the chromatographic run. The samples of 3.0 µL were injected into the HPLC system using an autosampler (Nanospace SI-2, Shiseido Co., Japan). The effluent from the column was monitored with a variable wavelength ultraviolet detector (Nanospace SI-1, Shiseido Co., Japan) at 220 nm. The integration of the chromatograms was performed by dsCHROM^® software (Do-Nam Instrumental Co., Seoul, Korea). The method was validated for linearity, precision (inter-day and intra-day), accuracy and selectivity [7]. The experiment was repeated three times on the same day and additionally on two consecutive days to determine interand intra-day precisions. Assays of control solutions from nafcillin sodium solutions (2.5 mg/mL) were undertaken to calculate the intra-day and inter-day variations using external standard method. Linearity was evaluated by serial dilutions of nafcillin sodium solutions with NS and D₅W for analysis. The stability of nafcillin sodium infusion solutions was determined in disposable Accufuser^® infusion device during 6 weeks of storage under RT and CT conditions. Periodically, samples were evaluated for appearances, visible particles, pH and chromatographic analysis. We analyzed the concentration of nafcillin sodium in two solutions at 0, 48 h, 1, 2, 4, and 6 weeks after the preparation of solutions by HPLCUV method. On each day, 1.0 mL of samples of nafcillin sodium solution with a nominal concentration were drawn from Accufuser^® infusion device for chromatographic analysis and 3.0 µL were directly injected into HPLCUV system. Three aliquots of each solution were processed. Statistical analysis was performed using one-way ANOVA with the level of significance set at 0.05 (PCS, v. 4.0, Springer-Verlag, New York, USA).

The linearity of calibration curves of nafcillin sodium solution was established which showed good linearity over the range concentration of 0.5 ~ 2.8 mg/mL (r² =

0.9951, Figure 2). Table 2 lists the relative standard deviation (R.S.D.) data obtained from the analysis of the samples on the same day (n = 5) and on consecutive days (n = 5). The R.S.D values were <1.26% and <2.35% for intra-day and inter-day results, respectively, meaning that the method was sufficiently precise.

There were no significant changes in physical appearances and clarities of the solutions in CT condition. The color of the samples was transparent with no color discoloration. Particles were not detected. However, the color of NS solutions became opaque after 10 days of study in RT condition. After 2 weeks, solutions changed into yellowish color and particles were detected. On the 4^th week, the filter of the silicone balloon infuser was partially clogged, and the injected samples did not flow out with the normal flow rate of 2.0 mL/h. The flow rate was less than the half of the normal flow rate. At the 6^th week, the filter of the silicone balloon infuser was completely clogged and the injected samples did not flow out at all. Similarly, the color of D₅W solutions in RT condition became opaque after 2 weeks of study and changed into yellowish color at the 3^rd week. At the 4^th week, particles were detected and at the 6^th week, the filter of the silicone balloon infuser was completely clogged and the injected samples did not flow out at all. The pH of D₅W and NS solutions slightly decreased from 6.84 to 6.06

Table 2. Intra-day and Inter-day precision studies (n = 5).

and from 6.51 to 5.87 in the CT condition, but significantly (p < 0.05) decreased from 6.85 to 4.49 and from 6.57 to 4.11 in the RT condition during 6 weeks (Table 3).

The typical HPLC chromatogram of nafcillin sodium (2.5 mg/mL) is shown in Figure 3. The retention time for nafcillin sodium was about 10.2 min. The initial concentration and the percentage of the remaining concentration which were observed at analytic time during 6 weeks for each nafcillin sodium solutions and storage conditions are listed in Table 4. The concentration of nafcillin sodium decreased significantly with time and reached 31.97% in D₅W and 27.66% in NS after 4 weeks under RT condition. However, 97.30% of initial concentration remained in D₅W and 92.66% remained in NS under CT condition.

Table 3. The change of pH (Mean ± SD, %) of nafcillin sodium solutions according to the time of storage.

¹⁾RT: room temperature, ²⁾CT: cold temperature, ³⁾D₅W: 5% dextrose, ⁴⁾NS: normal saline, ^*p < 0.05 vs 0 h.

Table 4. The retention of the amount (%) of nafcillin sodium in solutions (NS and D₅W) according to each storage temperature (RT and CT) and time (0 ~ 6 weeks).