A Proof of Principle Study of the Terminal Sterilization of Prefilled Syringes Using A Water Cascade Process

Filling process

For both brand A and B, syringes were filled at the stopper side under GMP conditions (laminar airflow with class C background). For physical stability testing ten syringes were filled with 50 ml 0.9 % saline solution (Baxter, Viaflo) and for sterility testing ten syringes were filled with 50 ml Trypton Soya Broth (TSB, BioTrading). The stopper was placed in the syringe using a needle inserted alongside the stopper for de-aeration. A maximum of 0.5 ml of air inclusion was accepted, based on the amount of air to be expected in the automatic filling process. The needle was removed when the stopper was in place. This procedure was used to mimic an automatic filling process.

Sterilization

All syringes were sterilized using a combined steam and water cascade sterilizer (De Lama DLOV/S-C) with chamber size 1450 mm (diameter) x 1450 mm (length). In this study the water cascade process was used. According to the European Pharmacopoeia (Ph. Eur.), the syringes were sterilized at 121 °C for 15 minutes. The temperature inside the product and inside the chamber is measured at the coolest part of the sterilizer, these parts were established during the general validation process of the sterilizer. Figure 1 depicts the temperature and pressure in the used sterilization process. It consists of four phases (temperature setpoint; pressure setpoint), indicated in Figure 1 with A-D, the warm up period, A (122 °C; 338 kPa), the plateau period, B (122 °C; 330 kPa), forced cooling with support pressure, C (70 °C; 340 kPa) and the cooling down period, D (70 °C; 100 kPa). The sterilization process is validated with a full load of 50 ml and 100 ml injection vials according to the criteria as described in the ISO 17665 [12]. The nest with syringes, fixated in a mesh sterilization tray with the tipcap facing upward, were placed in the middle of the top shelf of the sterilizer. The syringes were sterilized alongside 100 ml vials to simulate a full sterilization load. The two Pt100 temperature sensors (I and II) to monitor temperature of the load were each inserted in a 100 ml vial. As a negative control experiment, two prefilled syringes of each brand were not sterilized.

Figure 1:

Schematic view of sterilization process.

Closure integrity

The closure integrity was determined using the methylene blue dye ingress test (DIT) as described in the Ph. Eur. (9th ed.) and the technical report of the Parenteral Drug Association (PDA) [13]. The syringes filled with saline solution were immersed in a 1 g/L solution of methylene blue with the tipcap facing downward; two not-sterilized syringes with saline solution were included as negative control. The pressure was reduced to 50 kPa atmosphere for 15 minutes, followed by a 15-minute increase in pressure to 200 kPa atmosphere and afterwards the pressure was returned to atmospheric pressure. A pressure container (Walther LDG 20) was used to achieve negative and positive pressure. After immersion the syringes were rinsed with water, dried with paper tissues and visually inspected for the presence of dye. From all immersed syringes a sample was analyzed on a UV spectrophotometer (Shimadzu UV-2401 PC) between 654 and 700 nm. A solution of 0.5 µg/mL methylene blue was used as positive control. The Lower Limit of Quantification (LLOQ) of the UV analysis was 0.8 μg/L. All syringes with a calculated concentration below the LLOQ were considered free of methylene blue.

Stopper movement

Stopper movement was calculated using the equation for linear thermal expansion (eq. (1)) and the combined gas law (eq. (2)). This method is an alternative way to determine stopper movement without the use of intricate physical tests. The sterilizer Pt100 sensors log the temperature (in the two 100 ml vials and chamber) and pressure (in the chamber) every three seconds, this data was used to calculate stopper movement for both syringe brands.

Where ∆L is stopper movement (mm), α is linear thermal expansion of water (0.07^–3 K⁻¹) and ∆T is the difference in temperature (K).

Where P is pressure (kPa), V is volume (mm³), T is temperature (K) and k is a constant.

In the calculations it is assumed that:

1. The gas and fluid behave as independent systems

2. The shape of the stopper equals the shape of the tip of the syringe, therefore the syringe can be seen as a cylinder.

3. The gas behaves as an ideal gas; therefore the ideal gas law applies.

4. The liquid behaves like water; therefore the thermal expansion coefficient of water can be used.

5. The stopper can move freely, therefore the pressure in the syringe equals the pressure in the chamber.

6. The volume changes of the syringe are negligible relative to the volume changes of the gas and liquid.

7. The amount of gas present inside the syringe was estimated to be 0.5 ml.

Sterility testing

The syringes filled with TSB were placed in the incubator at 30–35 °C for 14 days. The syringes were visually inspected for the presence of bacteria by turbidity at 3 and 14 days after sterilization.

Other physical tests

To investigate the mechanical influence of the sterilization process several measurements on the physical stability were performed including weight, diameter and physical appearance. Weight was measured on a Mettler Toledo XS105 scale (d=0.1 mg) before sterilization and afterwards when the syringes were completely dry. The diameter of five syringes was measured at three positions (Figure 2) before and after sterilization using an analog Vernier caliper (Nippon Sokutei). The position of measurement was marked to ensure that this was the same before and after sterilization. After sterilization the syringes were visually inspected regarding clarity and color according to Ph. Eur. monographs 2.2.1 and 2.2.2. The syringes, tipcap and stopper were visually inspected to rule out physical deformation.

Figure 2:

Positions of diameter measurement.

Statistical analysis

The skewness and kurtosis of the data were analyzed to determine normal distribution. Diameters before and after sterilization were statistically compared using the paired t-test or the non-parametric Wilcoxon Rank test, where appropriate. All statistical tests were conducted using IBM SPSS Statistics version 23.0.

Closure integrity

The results for closure integrity are shown in Table 1. The positive control showed maximal absorbance at a wavelength of 664 nm. No absorption at 664 nm was found in all twelve syringes for both brands, that is, including the two negative controls. On visual inspection of the syringes after dye ingress testing the methylene blue had penetrated between the ribs of the stopper in one sterilized syringe and in one negative control of brand A. However, the syringes that showed dye between the ribs of the stopper showed no UV absorption within the saline solution. Taking into account the LLOQ, this shows that the concentration methylene blue inside the syringes was less than 0.8 µg/L.

Stopper movement

The results for stopper movement are depicted in Figure 3. Figure 3(C) and 3(D) show the volume change caused by the liquid and gas separately and Figure 3(E) provides these results combined. In Figure 3(F) the total volume change is converted to total stopper movement in millimeters for both brands. The total stopper movement for brand A using the data from Pt100 sensor I and II is, respectively 1.29 and 1.27 mm. The total stopper movement for brand B using the data from Pt100 sensor I and II was 1.42 and 1.40 mm, respectively.

Figure 3:

Results calculation stopper movement. Figure 3(A) depicts the pressure during the process. Figure 3(B) depicts the temperature during the process, where curve a and b depict the results from the two pt100 sensors and curve c depicts the temperature in the chamber. Figure 3(C-E) depict the liquid, gas and combined volume change calculated using the two pt100 sensors (curve a and b). Figure 3(F) curve a and b depict plunger movement for brand A and curve d and e depict plunger movement for brand B calculated with the results from the two pt100 sensors.

Sterility testing

No microbiological growth was observed in the syringes for both brand A and B following incubation at 30–35 °C for 14 days.

Other physical tests

The results for weight, diameter and physical appearance are shown in Table 2. Weight and diameter for both brands appeared not normally distributed, therefore the Wilcoxon rank test was performed. There was a very minimal, although significant difference in median (range) weight before (80.54 g (80.06–81.50)) and after (80.55 g (80.07–81.51)) sterilization for syringe brand A (Z=2.070, p<0.05). No weight difference was found for syringe brand B (79.05 g (78.78–79.31)) and after (79.05 g (78.78–79.31) (Z=1.414, p=0.157). There was no significant difference in diameter at the three points of measurement before and after sterilization for syringe brand A and B. In addition, no physical deformation of the syringe, tipcap and stopper was observed after sterilization. Furthermore, no change in color and clarity had occurred due to sterilization.

Conclusion

The results of this study demonstrate the proof of concept that the physical and microbiological stability of the cyclic olefin (co)polymer syringes is guaranteed during sterilization using a water cascade sterilizer. These results do not rule out the necessity of further chemical stability experiments with individual drug products in prefilled syringes and the validation of the process with a full load of syringes. These tests are necessary to fully validate the process of water cascade sterilization as a suitable and valid method for the sterilization of prefilled syringes.