Microbiological validation of a fully automated cytotoxic preparation system compared to manual preparation

In the pharmacy department of the University Medical Center Mainz ready-to-use cytotoxic preparations are prepared fully automated with a robotic system (APOTECAchemo™) or manually at cytotoxic workbenches. According to the Good Preparation Guidelines validation of the aseptic procedures should be performed by environmental monitoring and simulating the process with nutrient media. We simulated the aseptic preparation of patient individual injection solutions by using double concentrated tryptic soy broth as growth medium, water for injection and plastic syringes as primary packaging material. Media fills were either prepared automated (500 units) in the robot or manually (500 units) in cytotoxic workbenches in the same cleanroom over a period of 18 working days. The test solutions were incubated at room temperature (22°C) over 4 weeks. Products were visually inspected for turbidity after a 2 week and 4 week period. Following incubation, growth promotion tests were performed with S. epidermidis. During the media fill procedures, passive air monitoring was performed with settle plates and surface monitoring with contact plates on predefined locations as well as fingerprints. The plates got incubated for 5-7 days at room temperature, followed by 2-3 days at 30-35 °C and the colony forming units (cfu) counted after both periods. The robot was cleaned and disinfected according to the established standard operating procedure on 2 working days prior to the media-fill session, while on 6 other working days only 6 critical components were sanitized at the end of the media fill sessions. Every day UV irradiation was operated for 4 hours after finishing work.

None of the 1000 media fill products prepared in the two different settings showed turbidity after the incubation period thereby indicating no contamination with microorganisms. The reliability of the nutrient medium and the process was demonstrated by positive growth promotion tests with S. epidermidis. During automated preparation the recommended limits < 1 cfu per settle/contact plate set for cleanroom Grade A zones were not succeeded in the carousel and working area, but in the loading area of the robot. There was no difference in the microbial contamination rate depending on the extent of cleaning and disinfection of the robot.

Extensive media fill tests simulating manual and automated preparation of ready-to-use cytotoxic injection solutions revealed the same level of sterility for both procedures. The results of supplemental environmental controls confirmed that the aseptic procedures are well controlled.