by Paul E. Bledsoe and Scot Wepfer, R.Ph., FIACP

A unit is put to the test to determine if it can provide repeatable, reliable, uniform temperature distribution

Sterile injectable products, which are unavailable from commercial manufacturers, are commonly prepared by compounding pharmacies. Depending on whether the product is a solution or a suspension, the preparation of these products involves two techniques. Solutions can be passed through a 0.22-micron sterile filter before the product is filled into a vial or syringe. Suspensions cannot be filtered and must be sterilized before filling into the final package form.

Autoclaves are used for sterilization of the suspensions and the stoppers that are used to seal the vials. In order for the materials to be sterilized, they must reach 121°C and hold that temperature for 15 minutes.

Typically, the autoclave used in compounding pharmacies is a tabletop model that comes with predetermined cycles for sterilization. The liquid cycle, used for sterilization of suspensions, is the 15-minute period after the internal temperature of the chamber has reached the required 121°C temperature.

A study was recently conducted to determine if the actual performance of the autoclave would match the preset cycles, providing the proper sterilization, by reaching the 121°C and holding the temperature for the required time.

When the autoclave was tested empty or without a load, the results indicate that the autoclave performance matched the autoclave printout data.

However, a load consisting of 2000 ml of water in a 3000-ml beaker was placed into the autoclave chamber to simulate the actual product. The measurements showed that the actual temperature of the water and the temperature indicated on the autoclave printout varied.

According to the autoclave printout, the water in the beaker did not reach 121°C until the sterilization cycle was more than one-half completed, meaning there was less than the required 15-minute exposure time for sterilization. This would indicate that the liquid contained in the beaker was not sterilized during the preset liquid cycle of the autoclave.

Different materials and quantities produced conflicting results during the testing. The stir bar in the liquid did not reach the required temperature for two minutes after the solution reached temperature, meaning any contamination on the bar would not have been sterilized.

Previous experience with several similar autoclaves indicated that individual units perform differently and a single unit cannot be used to determine sterilization cycles for a particular brand or type of unit. Each unit should be tested to determine the correct sterilization cycle for a specific load.

Testing based on loads
This testing must be based on the load characteristics. Different loads require different cycles. Pharmaceutical companies base their autoclave testing on a predetermined, worst-case scenario. The most difficult load to sterilize is used to determine the cycle time that is necessary to assure the materials contained in the autoclave are sterile at the end of the cycle.

All autoclaves used for the preparation of sterile injectable products that are not sterile filtered through a double 0.22-micron filter should be subjected to this type of testing. Pharmacists should check the autoclaves used for sterilization to assure that the autoclave has the flexibility to adjust the cycle length to compensate for the “heat up.”

When a load is present in the autoclave, under no circumstances should the autoclave liquid sterilization cycle be used as an indication that the load is sterile. The cycle does not consider the time required for the “heat up” of the materials in the chamber in the 15-minute sterilization period. Cycle times should be developed for individual autoclaves with specific loads. This study should be used only as a guide and in no case should this information be considered the correct time required for sterilization cycles for liquids or other materials placed in the autoclave.

Temperature data analysis
Acceptance of the autoclave performance is based on analysis of the temperature data, recorded for the nine runs to determine if the autoclave being testing can provide repeatable, reliable uniform temperature distribution.

Before and after testing with the 12 Type T thermocouples, each thermocouple was calibrated against the Kaye IRTD serial # B1019, which was calibrated October 24, 2001. Based on the temperature data recorded over the two-day period of November 5 and 6, 2001, the following observations were made:

• The three empty chamber sterilization cycles ran with a sterilization set point of 122°C for 15 minutes and demonstrated uniformity of temperature distribution through out the chamber. This was verified by distributing 12 thermocouples throughout the 12- x 18-inch chamber. Number one (1) thermocouple was place next to the autoclave control sensor to verify control accuracy. The maximum spread between the 12 thermocouples during the cycles was 0.14°C with the control sensor reading 122°C and the Number one (1) thermocouple average reading 122.61°C.
• The two (2) liquid cycle sterilization runs with a sterilization set point of 121°C for 15 minutes, which is the cycle designated as the liquid cycle on the autoclave control panel, demonstrated the inability to bring a liquid load of 400 ml of water in a 500-ml flask up to sterilization set point until nine (9) minutes into the exposure time, as expressed by the autoclave control system. The flask also contained a 3/4-inch-diameter Delrin rod used to emulate a stir bar. Three (3) thermocouples were placed into the flask. Number one (1) thermocouple was attached to the Delrin rod at the bottom and the other two at higher levels in the liquid. The Delrin rod thermocouple lagged the other two by two minutes in reaching the set point of 121°C. The preset liquid cycle on this autoclave provided only six minutes of exposure at 121°C.
• The two (2) liquid cycle sterilization runs with a sterilization set point of 122°C for 30 minutes, which is an optional cycle, that can be set by the operator on the autoclave control panel demonstrated the inability to bring a liquid load of 2000 ml of water in a 3000-ml beaker up to sterilization set point until 15 minutes into the exposure time as expressed by the autoclave control system. The flask in the first run contained only water. The second run also contained a 3/4-inch-diameter Delrin rod used to emulate a stir bar. Three (3) thermocouples were placed into the flask. Number one (1) thermocouple was close to the bottom and the other two at higher levels in the liquid for the first run. Number one (1) thermocouple was attached to the Delrin rod at the bottom in the second run. The Delrin rod thermocouple lagged the other two by four minutes in reaching the set point of 122°C. The liquid temperature lagged the control temperature by 15 minutes without the Delrin rod and 16 minutes with the Delrin rod. The optional liquid cycle of 30 minutes at 122°C on this autoclave only provided 14 to 15 minutes of exposure at 121.1°C or above.
• Two additional sterilization cycles with a load of eight (8) glove o-rings and three thermocouples imbedded inside the bundle in a triangular configuration did show a one-minute delay in reaching the temperature indicated by the control system. The assumption at this point is that any load will cause the load temperature to lag the control temperature by some margin dependent on the load.

In conclusion
Based on the temperature-mapping studies, the autoclave is capable of providing sterilization temperatures. Adequate time for the cycle needs to be determined to allow sufficient exposure to complete the sterilization cycle.

Sufficient exposure time can be determined by a test method, similar to one described in this article. The lag in “heat up” time required by the load can be compensated for by increasing the cycle time. At no time should the preset cycles be relied upon as representing the time necessary to bring an empty chamber up to temperature and hold that temperature for a 15-minute period. Above all, the operator should not rely on the printed documentation generated by the autoclave control system as a verification of sterilization exposure time.

Paul Bledsoe is a retired engineer from Eli Lilly and Company. His engineering career includes more than 20 years working with the design and validation of autoclaves. Scot Wepfer owns the Compounding Shoppe in Birmingham, Alabama. He is a practicing compounding pharmacist and a fellow of the International Academy of Compounding Pharmacists.