`Fill it up`

The filling line for bottles (between 20 and 100mL) was installed in a Class B area (10,000). Operators are able to access the filling machine from Class B. Technical access is also possible from the outside (unclassified area at the moment). It will become classified as class C after the refurbishment of the area has been completed.

Description of the MDU-system:

• Automatic depalletising and unpacking of the bottles.

Pallets containing sealed packages of bottles are delivered to the depalletiser by forklift. This unit automatically picks up single sealed packages of bottles and moves them to the automatic unpacking machine. There the plastic wrapper is cut open and discarded by downward suction. The bottles are carried to the bottle washer (BOSCH RRU 3043) of the filling line.

• Bottle washer (BOSCH RRU 3043):

Firstly the bottles pass through an ultrasonic bath. They are then washed in a heated waterbath, singled out, picked up by a mechanical grip system and turned by 180°C. The bottles then undergo a cleaning and rinsing process (with a total of four steps, alternating between fresh water and compressed air). After being turned back to the upright position, the bottles are then transferred to the tunnel.

• Laminar flow tunnel (hot air followed by cool air) (BOSCH HQL 5351):

The tunnel is about 10m long and contains three heating segments (for depyrogenation) and five cooling segments. In the heating segments, air is constantly circulated, by being drawn down below the conveyer belt and filtered through HEPA filters after passing through heating units, before it comes into direct contact with the washed and rinsed bottles. A ventilator in the entry segment of the hot air tunnel constantly removes damp air and thus prevents saturation with aqueous vapour. Fresh air flows in crossways from the cooling section of the tunnel. Within the cooling segments of the tunnel, the air is constantly circulated with the aid of a cooler.

• Filling machine (BOSCH FLP4060):

The filling machine uses a time-pressure-filling system (see Fig. 2). The size range of bottles is between 20 and 100mL. The filling rate is 100 units/min for 50mL and 100mL infusion bottles, 150 units/min for all other bottle sizes.

Two large 3000L storage vessels are directly connected to the filling machine (Fig. 1). CIP / SIP (cleaning in place/sterilisation in place) is also part of the machine. There is a docking station for rubber stoppers.

During the filling process the bottles undergo a fully automated inprocess weight control. Each bottle is replaced onto the transportation band at its corresponding place after the weight has been recorded. The movement of the filling needles is synchronised with the bottles on the transporting system. The isolator can be flooded with nitrogen and thus filling can be carried out in an atmosphere >98%N2 atmosphere. The bottles are also prefilled with nitrogen before the actual filling takes place.

• Isolator (INP 1000, Skan):

There are two hard wall isolators. Part one covers the section where the incoming bottles from the tunnel are singled out (7m³) and part two covers the filling unit of the machine (16m³). There are eight double walled glass doors with pneumatic seals for easy access. There is a total of 13 gloves for handling operations within the isolators. The surfaces are decontaminated using H₂O₂ vapour.

The isolator covering the filling equipment can be filled with nitrogen prior to the filling process of O₂ sensitive products. The pressure controls for both isolator units are independent.

There are six different modes for the operation of the isolator:

a – production using conventional LF-Barrier handling,

b – production with isolator using air;

c – production with isolator using N2,

d – decontamination (dehumidifying, conditioning phase with H₂O₂, H₂O₂ decontamination using a SIS 700 system, aeration),

e – initial flooding of the isolator with N2,

f – revision function.

• Autoclave (Webeco):

The size of the autoclave chamber is 9m³. The autoclave is completely automatic and is controlled by the computer governing the whole system. Depending on the chemical characteristics of the product being filled, there are two options: either the bottles undergo terminal sterilisation (heat stable products), or they undergo a washing process ( heat labile products).

• Visual inspection (Eisai AIM W1088) and tightness inspection (Bosch, KLD1042):

There is a cosmetic control of the filled and sealed bottles. They are also checked for particles and the correct filling volume. After this the tightness of the sealed containers is checked by a separate control machine using a high voltage.

• Automatic stacking systems (Rotzinger):

There are three of these automatic stacking systems: one stacking bottles after the filling and capping process. After the cooling-down section (following the sterilisation or washing process of the bottles) there is a second one to de-stack the bottles and to feed them into the visual inspection machines. A third one stacks the bottles after the inspection and tightness controls are completed.

• Automated guided vehicles (AGV) and main computer system (MLD):

A main computer controls the whole production process (filling, sterilisation, visual inspection) for the MDU line. This means that the state of each individual machine is supervised and known to the computer. In addition, all of the components used for each batch are recorded. At the end of each completed batch, a printout is provided. This lists all of the components used, the sterilisation batches, and lists the number of bottles counted at each one of the Rotzinger systems.

The bottles stacked by the automatic stacking systems are guided along the production process by two AGVs. They are also controlled by the main computer. The two AGVs cover different parts of the whole filling line.

In order to connect the stoppers to the isolator without direct contact of the staff with the sterilised stoppers, a new technical approach was taken. Large process containers are used to wash, siliconise and autoclave the rubber stoppers. At the end of this process, the container valves are closed and the container remains under pressure until it is aseptically connected with the stopper feeding tube of the filling machine. The staff therefore do not come into direct contact with the stoppers. The batch number of the stoppers is automatically scanned and passed to the centralised computer of the system. The capacity of each of these SMEJA process containers is 150L. Since the total weight including the stoppers is about 750kg, a HANAG crane system was installed to move the container up to the feeding tube of the filling machine.

Benefits of using an isolator filling line

Employing this innovative technology reduces the contact between staff and product to a minimum. Therefore, a higher sterility assurance level for the filling of aseptically produced multidose units is achieved. In addition, the handling of parts in contact with the product can be minimised.

One example of the improvement in sterility assurance is the direct connection of storage vessels of the filtrate with the filling line. Each of these components are CIP/SIP-ed independently. The piping from the vessels to the filling line are cleaned and sterilised together with the filling machine. The piping can also be cleaned separately.

During the filling of oxygen sensitive products by flooding the isolator with nitrogen, the protection of the operator during the fillling process is guaranteed. Although the recommendations at present justify setting up the isolator in a Class D controlled area, this particular isolator was installed in a class B area.

The reasons are as follows: Firstly, in a Class B area the filling machine can be operated as a conventional LF-barrier system. This allows staff to familiarise themselves with the equipment well before it is run as an isolator. The LF-barrier system means that it is possible to open one of the doors of the isolator during the production process should this be necessary in order to carry out repair work which can otherwise not be carried out. This process was validated by completing media fills. Secondly, this mode of operation is used for terminally sterilised products only.

Initially the OQ (operational qualification) checks the overall ergonomic properties of the isolator. This is carried out to ensure that all areas of the isolator can be reached with ease during production. Secondly the H₂O₂ decontamination cycle is tested. Thirdly, while the isolator is in production mode using a nitrogen atmosphere, the residual O2 concentration is measured.

The following steps were tested during OQ:

All HEPA/ULPA filters of the two isolators (total of 14 HEPA filters) were tested for any leakage using the DEHS (Di-ethyl-hexyl-sebazate) test. This test is carried out at regular intervals. To guarantee a virtually particle free atmosphere the rate must be ±0.01% for all EU14 filters. The certificates of the filters from the manufacturer were also checked. All filters were clearly identified and checked for any mechanical strains and/or damages.

The air velocity of a vertically directed constant air flow for a cleanroom class A has to be = 0.45m/s with an acceptable deviation of +/- 20%. Ten different points per filter unit were used and the measurements were taken 300mm below the filter supplying the air to the isolator. Laminar air flow was proven and documented by video. While in the decontamination mode, the air velocity inside the isolator is reduced to 0.2ms-1 in order to increase the H₂O₂ concentration within the isolator.

The temperature distribution within the isolator was determined. Firstly, it is of importance to mark all cold spots which can lead to condensation of the steam/hydrogen peroxide mixture. Secondly, it is for defining the worst case position for the decontamination process. Bio-indicators are placed in the cold spots. The points chosen for these measurements were those where either an interruption of the air flow was possible or where the total mass of the metal is large. In addition, two measuring sensors with no contact to the surface have been installed as references for the air temperature.

The air flow properties were checked by carrying out a smoke test. The whole chamber of the isolator covering the filling machine was tested in production mode and also in critical transitional conditions (e.g. beginning and end of filling process). The tests proved that there was low turbulance displacing flow in the aseptic central area and no turbulant air flow in the isolator. These smoke studies were documented on video.

The differential pressure was checked to verify the setting point. It was also carried out to check that regulation of any pressure differences during standard working conditions and transitional conditions takes place. Between the two isolators and the class B cleanroom the pressure difference is 15 Pa +/- 5Pa with all doors closed during all the production modes. It also includes the flooding modes (H₂O₂ and N₂) of the isolator. The differential pressure is not regulated by the isolator while in the revision mode or the decontamination mode. Under transitional conditions the pressure difference has to be = 1Pa. Unidirectional flow and differential pressure prevent cross-contamination within the isolator.

To ensure that there is no H₂O₂ leakage during the decontamination cycle, the leakage test of the isolator is carried out. The leakage rate has to be less than 10% of the internal volume of the isolator per hour. The leakage rate was determined by measuring the time needed for an equilisation of the pressure within the isolator while the doors were closed.

There are two parts depending on the chemical characteristics of the products.

• heat stabile products which can undergo terminal sterilisation - using conventional LF-barrier technology

• heat labile products – using aseptic processing and isolator techology

For the operating staff to become familiar with the equipment the multi unit dose filling line has been validated for conventional LF-barrier use as a first step

The requirements for conventional LF-barrier production: During conventional LF barrier production the doors of the isolator are closed and the door seals activated. Since the isolator is placed in a class B cleanroom, the operating personnel are fully gowned according to Class A/B requirements. The surfaces of the isolator are disinfected with 70% ethanol. The routine handling is carried out by using the gloves on the isolator doors. Minor repairs (e.g. straightening of the needles) are performed by opening one of the isolator doors, if it is not possible to use gloves for access and repair.

There was an initial microbiological validation of the process for conventional LF barrier system routine production. Three media fills were done with the smallest and also with the largest bottle size (i.e. 20mL injection bottles and 100ml infusion bottles). These were carried out according to the SOPs for media fills. The routine handling operations which are simulated include the following:

• fallen over bottles which are either straightened up by opening the appropiate door of the isolator or by using the handling gloves

• realigning the rubber stoppers in the feeeding tube using those forceps placed inside the isolator.

After the initial microbiological validation of the filling line, the process is necessary for using the isolator as conventional barrier technology for terminally sterilised products.

This is followed by product validation of the heat stable products. Product bracketing was used to cover all of these products with the minimal number of batches neccessary. The most concentrated product was used for the validation of the compounding and the least concentrated product was used for the validation of the filling process. The cleaning validation was also part of the product compounding validation step.

Since this part has been completed, media fills for the initial microbiological validation of the aseptic filling process with isolator-technology will follow.

Three media fills will be carried out with the smallest bottle size in use (20mL). The machine is running at different speeds and different routine handling operations which can be done using the gloves will be carried out. This ensures that worst case conditions are simulated during media fills. After the initial validation this aseptic process will be revalidated every six months with one media fill. Once these initial media fills are completed, the product validation covering the compounding and filling and cleaning validation for the corresponding heat labile products will follow.

With this new filling machine we can chose conventional LF-barrier technology or isolator technology according to the requirements of the different products.