Drug product sterile filtration is the final process step before final filling of drug product into containers such as vials and syringes for clinical use. Here, we show how our drug product filtration system is used with different filter configurations to achieve optimal drug product recovery. Furthermore, we demonstrate the patented system setup to recover a maximized volume of high-value drug product. Also, we show how initial product inline dilution, which can potentially cause product wastage, is reduced after water-wet, pre-use post-sterilization integrity testing (PUPSIT) is performed.
Introduction
Our drug product filtration system is designed with a patented recovery process to maximize product recovery across a range of filter sizes. The high-value drug product recovered enables a greater number of containers to be filled, therefore improving your process efficiency and enabling sufficient drug quantities to meet the market needs for patients.
The system can be operated in single or redundant filter configurations dependent on process requirements. Process flexibility lets you install a wide range of filter sizes from 1 in. capsules (e.g., KA1) to 10 in. capsules (e.g., NP6), see Figure 1 for a typical process setup. Depending on user preference, the system is capable of transitioning from filtration to recovery based on time, product inlet/ outlet weight or detection of air at the product inlet level sensor 2, which can impact the volume of product recovered. After completion of the filtration process, a proprietary and automated recovery method ensures the maximum recovery of your product. This involves several process steps:
- Recovery of drug product volume upstream of sterilizing-grade filter(s) by a pressure decay test to the product recovery bag.
- Recovery of drug product volume from the inner core of the sterilizing-grade filter 2 (the filter nearest to the product outlet, also found at the same position for single filtration configuration) by pumping of air into the filter core from a prefilled air recovery bag.
- Pumping of the recovered drug product into the final filling outlet from the product recovery bag.
The combination of these recovery steps significantly reduces the nonrecoverable hold-up volume in the flow kit, post-filtration. Additionally, the system can reduce possible product loss by reducing the initial product flush volume (potential product dilution) required to replace water after water-wet PUPSIT. This is achieved by applying an additional reverse pumping step and draining water from specific flow kit areas, where, after the PUPSIT, water remains upstream of the filter flow paths.
The system demonstrates several different transition modes from filtration to recovery—the product recovery method and the reduction of initial in-line product dilution after water-wet PUPSIT. We demonstrate this with a redundant filtration system setup utilizing both KA1 (1 in.) or NP6 (10 in.) filter sizes and we used water as the product to be recovered (1 cP).
Materials and methods
The equipment and materials used for this study included:
- Drug product filtration system, CE version
- LevMixer™ drive unit, gen IV
- LevMixer™ 200 L biocontainer, 4 in. powder port, standard design, cubical
- Allegro™ 3D 200 L biocontainer w/ AdvantaPure platinum-cured silicone tubing
The scope of this study was to determine the nonrecoverable hold-up volume within the entire full flow kit. However, any external inlet or outlet connections that were not part of the flow kit were out of scope and excluded from the nonrecoverable hold-up calculations. Therefore, at the end of this process, we completed a full mass balance of the inlets versus outlets to assess how much fluid entered and exited the flow kit and to calculate the hold-up volume within the flow kit.
Drug product filtration system setup
For this study, we performed tests with a redundant filter configuration set-up using KA1 (1”) or NP6 (10”) filters. The hardware and flow kits were installed as described in the operating instructions.
Figure 1 shows the system with the 20 L flush bag, as well as the 200 L LevMixer™ single-use mixer system.
Fig 1. The drug product filtration system process setup. The picture shows the system in a redundant filtration system setup (two Kleenpak™ Nova NP6 filters) equipped with a stainless-steel container, a LevMixer™ drive unit gen IV unit (left), and a 20 L flush bag trolley (right).
Flow kit leak testing
After the installation of the flow kits, we performed an optional predefined leak test (leak test 2), which included the following:
- Leak test of filter 2 flush bag and air recovery bag. This step filled the recovery bag with air.
- Reduction of half of the air from the recovery bag using pump 2. This left the recovery bag partially filled with air to recover the inner core volume of filter 2 during the recovery process later in the batch.
Upon passing the leak test, we removed the filters, weighed (the dry weight to be used when determining the hold-up volumes) and reconnected onto the flow kit.
Filter preparation and integrity testing
We primed the tubing connecting the deionized water to the buffer/WFI inlet for PUPSIT until fluid detection at level sensor 1 (LE001), see Figure 2.
For all seven runs (Table 2), we primed the filters (filter 1 and filter 2) at 0.1 L/min irrespective of filter size. We flushed KA1 (1 in.) filters at 0.1 L/min for 5 min and NP6 (10 in.) filters at 1 L/min for 5 min before performing PUPSIT. All filters passed PUPSIT using the forward-flow test procedure.
Fig 2. The drug product filtration system with existing level sensors 1 (LE001) for buffer/WFI inlet, 2 (LE002) for product inlet and 3 (LE003) for product outlet and redundant filters at positions 1 and 2.
Assessing in-line dilution after water-wet PUPSIT
We assessed the draining capabilities and its effect on the required volume of product to fully prime the flow kits after water-wet PUPSIT and before filtration. The system reduces potential inline dilution effects of the drug product after the performance of a water-wet PUPSIT and before drug product filtration, the remaining water in the tubing can be reduced by the two following steps (Fig 3):
- Reverse pumping the upstream flow path of filter 1 back to the buffer/WFI inlet.
- Draining the flow path incline between filter 1 and filter 2 into flush bag 1.
In this study, we compared the required product volume to fully prime the KA1 and NP6 redundant filter flow kits after water-wet PUPSIT with and without the reverse pumping and draining steps.
Fig 3. The drug product filtration system flow path to reduce remaining water in the flow kit upstream of filter 1 and filter 2 after water-wet PUPSIT. The picture shows the system with the flow path (left, in yellow) when the pump is working in the reverse direction to reduce water in the upstream flow path of filter 1. The remaining water is pumped back to the buffer/WFI inlet, and the remaining water in the flow path area between filter 1 and filter 2 (middle, in orange) is reduced by draining to flush bag 1.
Assessing transitions from filtration to recovery
To start the filtration step, we primed the tubing connecting the product inlet tote from the LevMixer™ single-use mixing system until fluid detection at level sensor 2 (LE002), see Figure 2.
We reprimed filters 1 and 2 and flushed, as described above; however, we used the product inlet instead of the buffer/WFI inlet. All filtration runs in this study were performed at a constant speed. Once the filtration transition was achieved, and with the outlet tubing out of scope for this study, we disconnected and drained into the product outlet tote.
The breakdown for each test run is detailed in Table 2 below. The choice of transition from filtration to recovery depends on your preference. Two were tested to assess the effect on nonrecoverable hold-up volumes:
- Transition 1: After a predefined filtration time (Table 2, run nos. 1–6).
- Transition 2: Complete filtration of the product batch, that is, detection of air at level sensor 2, LE002 (Table 4, run no. 7).
Table 2. Overview of performed redundant filtration runs. showing the filtration transition type, filter type, feed, and pump control used for each test
| Run no. | Filtration transition | Filter type used for filter1 and filter 2 | Feed (product inlet) | Pump control |
| 1–2 | Time (10 min) | KA1DFL | Deionized water | Constant speed (10%) |
| 3 | Time (10 min) | KA1EKV | Deionized water | Constant speed (10%) |
| 4–6 | Time (10 min) | NP6DFL | Deionized water | Constant speed (10%) |
| 7 | LE002 (loss of fluid detection at level sensor 2) | NP6DFL | Deionized water | Constant speed (10%) |
Product recovery
During the recovery process, the product upstream of filter 1 and filter 2 was drained via a pressure decay test at 1 bar (14.5 psi, 0.1 MPa) and collected in the recovery bag (Fig 4). Then we pumped the air from the recovery bag (prefilled at leak test 2) into the inner core of filter 2, displacing the product within the inner core and that product was collected in the recovery bag. This was followed by manually clamping the recovery bag at the inlet and moving it from a horizontal position to a vertical position to facilitate draining.
Once this was complete, we pumped all recovered product (product upstream of filter 1, filter 2 and product from the inner core of filter 2) into the product outlet tote, with the pump stopping 15 s after air was detected at the level sensor 3 (LE003) to ensure maximum product recovery.
We measured the weight of the recovery bag between all of the automated recovery steps. Upon completion of the recovery process, filter 1 and filter 2 were removed and weighed (wet weight used when determining flow kit hold-up).
Fig 4. The full recovery process of the drug product filtration system (A–D). The orange lines show product flow path, while yellow lines show air pumping flow path. (A) Recovery of the product upstream of filter 1, (B) Recovery of the product upstream of filter 2, (C) Recovery of product from the inner core of filter 2, (D) recovered product pumped from the recovery bag to the product out tote.
Results
The study shows the capabilities of the system to recover additional volume of high-value drug product through the patented recovery process. Also, it shows how the initial product inline dilution is reduced when water-water-wet PUPSIT is performed.
Enhanced product recovery volume
Typically, the recovery process after filtration involves only a pressure decay test on the upstream side of the sterilizing grade filters. The system’s proprietary recovery process shows enhanced drug product recovery, which incorporates not only a pressure decay test on the upstream side of both redundant filters, but also an additional pumping step of air from the recovery bag into the inner core of filter 2 to further recover drug product from the inner filter core.
For the tested KA1 (1 in.) redundant filter configuration, the recovered average product volume by utilizing the patented recovery process increased from 182 ± 5 mL to 281 ±18 mL (~ 54% increase).
For the tested NP6 (10 in.) redundant filter configuration, the recovered average product volume increased from 1351 ± 36 mL to 1517 ± 34 mL (~ 12% increase). The total volume and a comparison between the recovered volume upstream of filter 1 and filter 2 versus the recovered volume from the inner core of filter 2 is shown in Figure 5.
Fig 5. Comparison of the total average volume of product recovered using the DPFS recovery process. A breakdown of the volume recovered from the pressure decay test of the upstream side of filter 1 and filter 2 (blue) and additional recovery from the inner core of filter 2 (orange). Data shows the average value for each filter configuration, n = 3.
The additional drug product recovered by the patented recovery process from the inner core of filter 2 directly improves process efficiency. This results in a greater quantity of containers (e.g., vials or syringes) that can be filled (Table 3). An estimated additional 49 × 2R vials and 83 × 2R vials can be filled for KA1 and NP6 configurations, respectively.
Table 3 provides an estimated breakdown of extra vials by size that can be filled, highlighting the additional doses available for patients through using the system.
Table 3. Quantity of additional vials for typical vial sizes that can be filled as a result of the system’s recovery process with either KA1 or NP6 filter configurations
| Vial size | Vial volume (mL) | Quantity of additional filled vials by patented product recovery process | |
| KA1* | NP6† | ||
| 2R | 2 | 49 | 83 |
| 4R | 4 | 25 | 42 |
| 6R | 6 | 17 | 27 |
| 8R | 8 | 13 | 21 |
| 10R | 10 | 10 | 16 |
*KA1 filter configurations with additional 99 mL recovered.
† NP6 filter configurations with additional 166 mL recovered.
Nonrecoverable hold-up volume in the flow kit after filtration process and patented recovery process
We determined the nonrecoverable hold-up volume in six redundant filtration flow kits to be less than 620 mL, excluding filter hold-up volume and irrespective of batch volume.
The average nonrecoverable hold-up volume of the flow kits was determined to be 432 ±120 mL. This shows the system's ability to minimize product loss across a range of flow-kit configurations.
Fig 6. Determined nonrecoverable hold-up volume in six single-use flow kits after redundant filtration and applied recovery process based on filtration transition on a fixed time (10 min) with the system. Filter hold-up volumes are excluded in the determination.
Reduction of potential product dilution after water-wet PUPSIT before filtration
We performed a study on KA1 (1 in.) and NP6 (10 in.) redundant filter flow kit configurations (Fig 7) to show the reduction of potential product dilution after water-wet PUPSIT
Using the automated reverse pumping and draining steps, we reduced the average determined residual water volume in the KA1 redundant flow kit (after water-wet PUPSIT) from 375 mL (+/- 38 mL) to 163 mL (+/- 11 mL). Therefore, the potential inline dilution of product can be reduced with a dilution factor of 1.4× versus 2.7×. Similarly, when looking at the NP6 redundant filter flow kit, the residual water volume was reduced from 1070 mL (+/- 126 mL) to 776 mL (+/- 108 mL). This would result in the potential inline product dilution to be reduced with a dilution factor of 1.4× instead of 1.7×.
Fig 7. Comparison of the average residual water volume present after water-wet PUPSIT and reverse pumping and draining step of the drug product filtration system. Bars reflect the average value ± standard deviation, n = 3.
Conclusions
- The drug product filtration system is optimized for maximal product recovery by applying a fully automated recovery process, regardless of batch volume and filter configuration.
- Transitioning from filtration to the recovery phase based on the detection of air at the product inlet level sensor, instead of a fixed time, additionally increases the volume of product recovered.
- Additional steps such as reverse pumping and draining aid in reducing product losses due to possible inline dilution effects (when water-wet PUPSIT is performed).
CY51403