Use of Overkill Half-Cycle Qualification Data to Support Reduction of Exposure Time in Validated Ethylene Oxide Sterilization Cycles

A cross-industry collaboration team was formed during the 2019 Kilmer Conference to address the growing demand to reduce ethylene oxide (EO) consumption. Shortly after the Kilmer Conference, the Food and Drug Administration (FDA) further challenged the industry to innovate on reduction of EO use and emissions. This FDA challenge gave further emphasis to the need to develop an approach that could be aligned on by the collaboration team with the intent to further share the approach with the industry.

The majority of EO cycles used for medical device sterilization have been validated based on the overkill half-cycle approach in accordance with ANSI/AAMI/ISO 11135:2014.¹  However, this approach provides a very conservative overestimate of the sterility assurance level (SAL), often providing a sterilization cycle that is longer than required to meet a 10⁻⁶ SAL.

The overkill half-cycle approach requires the use of a microbiological challenge that typically is more resistant than the product or component bioburden. In addition, the starting population is well above the normal product or component bioburden level. The microbiological challenge ordinarily used is the Bacillus atrophaeus spore, which is the standard industry challenge for EO sterilization.² 

Generally, the microbiological challenge organism is inoculated in the most challenging location to sterilize within the product, yielding what is referred to as a process challenge device (PCD). During qualification studies, PCDs are spread throughout the sterilization chamber load to include the most difficult-to-sterilize chamber location. The overkill half-cycle approach is validated by executing a series of sterilization processes to demonstrate inactivation of the PCD (i.e., where no growth of the PCD is observed following exposure). This process is referred to as the microbiological performance qualification (MPQ). The exposure time that delivers these conditions is then at least doubled to determine the production sterilization exposure time. The half cycle usually gives > 6-log population reduction and doubling the half-cycle time will provide > 12-log reduction; therefore, the method is called the overkill half-cycle approach.

The conservative overestimate of the SAL is due to the requirement to have no growth of the PCDs during the half-cycle exposure. The sterilization chamber volume that is being qualified dictates the number of PCDs that are required (typically a minimum of 10 PCDs) that exhibit no growth of the PCDs in the half cycle and that will provide a minimum SAL of 10⁻¹ to 10⁻² (i.e., less than one in 10 or less than one in 100, respectively). For example, if a minimum of 35 PCDs were selected due to the usable chamber volume and zero positives were observed following half-cycle exposure, then a SAL of approximately 10^−1.5 is demonstrated using a conservative D-value with the assumption of one positive PCD.

11135 Annex B, section B.1.2.b. provides another qualification approach for calculation of the EO exposure time for a routine production cycle. This qualification method, called the cycle calculation approach, uses data collected over a series of sublethal cycles to establish the PCD decimal reduction value (D-value; i.e., a resistance value for sterilization effectiveness). Sublethal cycles are exposure conditions where some of the PCDs are expected to be positive for growth while others are expected to be negative for growth. Using the PCD positives, an estimated PCD D-value can be determined. This estimated PCD D-value is then utilized to determine the routine production cycle parameters, providing the process parameters that will deliver the desired SAL.

The analysis presented in this article uses both the data obtained from the overkill half-cycle approach and the foundations of the cycle calculation approach to determine new EO exposure time parameters. Data obtained from the overkill half-cycle qualification allows for a conservative estimate of the PCD D-value. The D-value is considered conservative as one must assume one positive-growth when zero positives were observed from the PCDs following exposure to the overkill half-cycle.

Following load conditioning (e.g., humidification), cycle lethality occurs upon injection of the sterilant into the chamber and continues through the entire process. An equation established by Mosley et al.³  can be used to calculate the equivalent exposure time (U) for an EO sterilization process, taking into consideration the lethality achieved during gas injection through exposure and the gas evacuation phases. Data obtained in the overkill half-cycle MPQ and calculation of equivalent exposure time will allow a more accurate understanding of the delivered cycle lethality.

Using the equivalent exposure calculation U from the overkill half-cycle MPQ data and data from routine production cycles, the EO exposure time parameters can be determined to more accurately define the EO exposure time required to deliver the desired SAL. Given that medical device companies already have this data, a more accurate EO exposure cycle time can be a determined and documented. The documented analysis for the more accurate EO exposure time can be used to support regulatory submission.

Before beginning the evaluation of cycle data, the following prerequisites exist to ensure proper analysis and associated conclusions:

1. The cycles under evaluation must have been validated in accordance with the overkill half-cycle method per requirements in 11135.

2. The relationship between the product bioburden and the PCD must be understood. Relative resistance and bioburden quantity must be less than that of the PCD as previously defined.

3. The bioburden program for the product families associated with the sterilization cycle must be stable, with historical data to demonstrate maintenance of bioburden quantity and resistance.

4. The cycle stability during the gas injection and postexposure washes must be demonstrated. The analysis includes the time from the initiation of EO gas injection, through any inert gas injection (if used) prior to the start of the EO dwell phase, the exposure time, and the postexposure washes. It should be noted that variability in the time and rate of injection and wash phases will have to be considered to establish worst-case conditions.

Evaluation of four different EO sterilization cycles from four different medical device companies was performed for calculating equivalent exposure time. The approach used the Mosely et al. equation for calculating equivalent exposure time.

U_i

equivalent exposure time for a given time interval

t_T

time interval (t) at temperature (T)

z

z-value (30°C)

T_i

temperature

T_ref

reference temperature

C_i

EO concentration

C_ref

reference EO concentration

In Table 1, the reference EO concentration (C_ref) was derived from the overkill half-cycle MPQ data. The EO concentration calculations include total partial pressure of EO injected into the chamber and use the average chamber temperature from the MPQ cycle documentation. The reference gas concentration was calculated using the following formula:

n

44,000 mg/mole (molecular weight of EO)

P_EO

partial pressure of EO in chamber after injection

R

62.361 mmHg-L/gm-mole-Kelvin

T_ref

reference temperature (converted to Kelvin)

The reference temperature (T_ref) was also based on the average chamber temperature from the MPQ documentation. The rationale for using these data as the reference values for the formula is that the MPQ runs are conducted at lower temperatures and partial pressure of EO when compared to the routine production cycles. Using the MPQ data to set the reference values adds to the conservative nature of the evaluation.

The EO concentration (C_i) and temperature (T_i) come from the run record data. C_i represents the calculated EO concentration for that time stamp entry. T_i represents the average chamber temperature for that same time stamp entry. Each time stamp entry on the run record correlates to a new C_i and T_i used in the equation.

D-values were calculated for each cycle using the MPQ data. As previously mentioned, all PCDs were negative for growth in the MPQ cycles, therefore a single positive PCD was assumed in order to calculate an estimated D-value using the Stumbo-Murphy-Cochran Procedure⁴  with the following formula:

D_T

D-value

U_i

total equivalent exposure

N₀

initial population

n_i

number of units tested

r_i

number of units sterile

Using the D-value (D_T) and the equivalent exposure time (U) obtained from the previous calculations, a spore log reduction (SLR) was determined using the following formula:

Of the three overkill half-cycle MPQ documentation, the MPQ half cycle with the longest equivalent exposure time was used for the analysis in Table 1. This was selected as the longer equivalent exposure time results in a longer estimated D-value (i.e., minutes). A longer estimated D-value equates to a conservative estimate of the PCD resistance, and therefore provides a conservative estimate of the SAL based on exposure time.

Data from a typical routine production cycle was used in the establishment of the routine process SAL. SLR value was used along with the microbiological challenge population (N₀) to calculate the production cycle SAL.

Data in Table 1 represents the time from EO charge through the end of the process.

It should be noted that Cycle A has a longer equivalent exposure time than the actual run time for the production cycle. This happens as the production cycle uses a higher partial pressure of EO than that used during the MPQ cycle. Since the EO reference concentration is derived from the MPQ cycles, the higher EO concentration in the production cycle equates a longer equivalent exposure time for each time stamp during EO dwell. When accumulated throughout the cycle, a longer equivalent exposure time than the actual run time is achieved. For the other three cycles, only slight increases in EO partial pressure is seen during production cycles.

Based on this analysis and a required SAL of 10⁻⁶, Cycles A, C, and D have between 5.6 to 6.8 logarithms (logs) of additional inactivation. Therefore, these three cycles are candidates for utilizing the proposed methodology to reduce exposure times while still achieving the desired SAL. This information could reduce cycle exposure times by up to 40%. Table 2 provides potential reduction in EO dwell time settings for each cycle.

There are multiple approaches to reducing EO utilization for the sterilization of medical devices. The 11135 standard provides other approaches to qualify the sterilization cycle in addition to the overkill half-cycle and cycle calculation methods described earlier. One of these approaches—referred to as the bioburden approach—uses the product bioburden instead of a resistant microbiological challenge to demonstrate process lethality and would allow for use of shorter exposure times or reduced EO concentration. Similarly, use of a microbiological challenge that is more consistent with the production bioburden (i.e., quantity and resistance) would also allow for cycle development with decreased exposure time and/or EO concentration. This second method is referred to as the bioburden/biological indicator method.

Qualification activities using the overkill half-cycle approach or calculation approach can take existing cycles and convert them to cycles using lower EO concentrations. In some instances, this may result in longer sterilization cycles but reduce EO utilization. The analysis presented in this article uses data that is already available and could allow for immediate exposure time reduction while working through qualifications with other methods.

Using the proposed approach, in lieu of solely relying on the half-cycle overkill approach, to calculate new exposure times can provide potentially significant overall reduced routine production cycle time. Reduction in exposure time could also reduce overall product absorption of EO, thereby reducing the amount of EO that needs to be removed from the product during the post EO dwell remainder of the process. In addition, the reduced cycle exposure time could allow for additional evacuations to remove additional EO residuals from the product or provide for more EO processing capacity. This will help the medical device industry to ensure that hospitals, healthcare providers, and patients have access to medical devices that are safely and effectively sterilized.

The authors thank Byron Lambert (Abbott), Christophe Deneux, and Tony Faucette (BD) for their contributions to the development of this article. The team also thanks AdvaMed for hosting this cross-industry collaboration under their research and development auspices.