Garrett Krushefski is scientific and technical services manager at SGM Biotech.

Garrett Krushefski

	Biological indicators (BIs) are used during cycle development, validation, requalification, and routine monitoring of sterilization processes. Published standards provide appropriate resistance performance characteristics (e.g., D-value) and thus govern the efforts of the manufacturers of biological indicators (1-3). Similarly, BI user guidance documents also reference appropriate selection and use of BIs, and the information provided in such documents will often guide end-user purchase specifications (4). In turn, information about appropriate resistance capabilities of the BIs are sometimes written into regulatory submissions when medical device or pharmaceutical manufacturers seek regulatory clearance for their products.

	As of Dec. 31, 2014, the United States Environmental Protection Agency Clean Air Act will prohibit the sale and use of HCFC-based (hydrochlorofluorocarbon) products in the US, and that will include Oxyfume ethylene oxide (EtO) sterilant blends such as Oxyfume 2000, which consists of 8.6% EtO and 91.4% HCFC-124. (Oxyfume is a registered trademark owned by Honeywell International.) This requirement means that all BI manufacturers will have to move to 100% EtO as the test gas for determining the resistance performance of EtO BIs by the end of 2014. Currently, Oxyfume 2000 is often used by BI manufacturers for assessing BI EtO D-value label claims, so it is in the best interest of the BI community (manufacturers and end users) to assess the potential effects of this change.
	
	There have been previous changes to the EtO gas used for BI testing. A previous version of an HCFC mixture, Oxyfume 2002, which consists of 10% EtO, 63% HCFC-124, and 27% HCFC-22, was eliminated by the Clean Air Act in December of 2009. At the time, many BI manufacturers were using Oxyfume 2002 as the source gas in their resistometers when performing BI EtO D-value resistance assessments. As the elimination of Oxyfume 2002 approached, comparative studies were performed to determine if the switch from Oxyfume 2002 to Oxyfume 2000 would have any impact on measured resistance performance. The results of these studies indicated that the change in gas had no significant effect on the measured resistance of the BIs. As an example, MesaLabs EZTest lot G-162 displayed a D-value of 3.60 minutes when tested in Oxyfume 2002 or 3.58 minutes when tested in Oxyfume 2000. STERIS indicators showed similar results. STERIS VERIFY tested at 4.0 and 3.9 minutes, Spordex strips at 4.2 and 4.7 minutes, and Spordex discs at 3.4 and 3.5 minutes, when exposed in Oxyfume 2002 and Oxyfume 2000, respectively.
	
	


	In anticipation of this mandatory switch from Oxyfume 2000 to 100% EtO for BI testing, additional comparison studies were performed to determine if the switch from Oxyfume 2000 to 100% EtO would have any impact on BI resistance label claims. The test results for this change were markedly different than the change between mixed gasses. Despite programming both resistometers for identical exposure parameters (600 mg/L EtO, 54 °C, 60% relative humidity [RH]), Mesa BIs (both paper strip and self-contained versions) were showing a 26% to 39% reduction in measured D-value when tested in a resistometer using 100% EtO as the source gas. Having obtained these results, Mesa obtained strip and self-contained BIs (SCBIs) from other manufacturers, and when tested in Mesa resistometers, the same trend was observed. Specifically, the D-values when tested with 100% EtO were 29% to 51% lower than the results from the Oxyfume tests. (These data were the subject of a whitepaper [5] previously posted on the MesaLabs website).


	Based on these results and other preliminary tests performed by other BI manufacturers, this issue was brought up for discussion at the Association for Advancement of Medical Instrumentation, AAMI ST/WG 4, Biological indicators working group meeting in Alexandria, Virginia on Oct. 15, 2012. The result of that discussion was the decision to launch a collaborative effort amongst the three BI manufacturers that possessed the ability to perform both Oxyfume 2000 and 100% EtO exposures. Each manufacturer (STERIS Corporation, 3M, and Mesa Laboratories) agreed to exchange BIs to be tested by the other parties and share results. Exchanged BIs included spore disc, spore strip, and self-contained BI configurations.
	
	The results in 

 I are presented in a manner that protects the identity of the manufacturer. Of the 15 results, 13 showed a decrease in resistance when tested using 100% EtO as the source gas and two instances showed no change in measured resistance. There were no observations of a higher measured resistance for BIs tested in 100% EtO. Specifically, the D-values in this round of testing are 0% to 39.5% lower when tested in 100% EtO as compared to the Oxyfume results, with an average reduction in measured D-value of 19.5%.

	It is unknown what causes the lowered resistance measurement when 100% EtO is used as the resistometer source gas. The authors speculate that when using an Oxyfume blend gas, the HCFC competes with the EtO molecules for access to critical binding sites on the spores. Such competition would not exist when 100% EtO is the source gas for the resistometer cycles. With HCFC present in the exposure chamber and blocking EtO molecule access to the critical binding sites, the result is fewer alkylation reactions and thus a decreased lethal insult to the spore, despite both chambers providing 600 mg/L EtO, 54 °C, 60% RH conditions.
	
	


	The differences in measured D-value are cause for concern when considering the resistance recommendations that appear in current published standards. 

 shows the recommendations that appear in 

), International Organization for Standardization (ISO), and 

). When use of Oxyfume 2000 becomes prohibited and BI manufacturers switch to 100% EtO, the labeled D-value claims will likely show a pronounced downward shift consistent with the test results reported here, compared to historical values. The authors stress the fact that the BIs are not changing; the spores have not become less resistant to the sterilization process. Rather, the “measuring stick” has changed with the changeover from Oxyfume 2000 to 100% EtO as the source gas used in resistometers. Because the measuring stick is changing, published standards will need to follow suit and adjust the verbiage in the relevant documents.
	 

 Requirements for BI D-value when tested in EtO. 

	The data from this study indicate a decrease in measured value of up to 39.5% solely due to the use of 100% EtO as the supply gas. As such, the authors would recommend a change in published ranges to match. Whereas 2.5 minutes to 5.8 minutes were a suitable range of D-value for BIs tested in an HCFC blend gas, ISO and EP should consider lowering the “not less than 2.5-minutes” specification to “not less than 1.5-minutes” to accommodate the observed percent differences in the test data. 

 provides a lower and upper range of resistance that is typical for EtO BIs. As such, the current 

 citation of 2.5 to 5.8 minutes should be adjusted to perhaps 1.5 to 5.8 minutes (i.e., 2.5 - 40% = 1.5) for BIs that are tested using 100% EtO as the resistometer source gas.
	
	Given that making changes to published standards can takes months, or even years to complete, the industry will likely experience a gap in time where available BIs (tested in 100% EtO) may not have a resistance label claim that meets the minimum values that currently appear in USP, EP, and American National Standards Institute (ANSI)/Association for the Advancement of Medical Instrumentation (AAMI)/ISO documents. Furthermore, end user purchase specifications and/or the information in the end user’s regulatory submissions may also conflict with what is available from BI manufacturers, as the end user’s stated values were based upon BIs that were resistance tested with an EtO/HCFC blend gas, rather than 100% EtO.
	
	


	1. ANSI/AAMI/ISO 11138-1:2006/(R) 2010 (AAMI, Arlington, VA, 2006).
	2. 

 General Chapter <1035>, “Biological Indicators for Sterilization” (US Pharmacopeial Convention, Rockville, MD 2014).
	3. 

 General Chapter - 5.1.2, “Biological Indicators of Sterilization” (Ph. Eur. Commission, Strasbourg, France, 2013).
	4. ANSI/AAMI/ISO 14161:2009 (AAMI, Arlington, VA, 2006).
	5. K. Matzinger, “Oxyfume Blends vs. 100% EtO,” 

http://biologicalindicators.mesalabs.com/wp-content/uploads/sites/31/2014/07/Spore-News-Vol-9-No1.pdf

	* Garrett Krushefski is senior vice-president of operations at Mesa Laboratories Biological Indicator sites in Bozeman, MT and Omaha, NE, 

; Anthony M. Piotrkowski is regulatory manager at STERIS Corporation in Mentor, OH and co-chair of AAMI ST/WG 04 Biological indicators working group; Craig A. Wallace is senior technical specialist at 3M Infection Prevention Division, St. Paul, MN; and Kellie A. Matzinger is manager of Scientific & Technical Services at Mesa Laboratories.
	
	*To whom all correspondence should be addressed.

	Submitted: March 31, 2014. Accepted April 17, 2014.

 When referring to this article, please cite it as G. Krushefsk, "Effects of 100% Ethylene Oxide Test Gas on the Resistance of Ethylene Oxide Biological Indicators," 

Articles

In anticipation of this mandatory switch from Oxyfume 2000 to 100% EtO for BI testing, comparison studies were performed to determine if the switch from Oxyfume 2000 to 100% EtO would have any impact on BI resistance label claims.

Effects of 100% Ethylene Oxide Test Gas on the Resistance of Ethylene Oxide Biological Indicators

There is no universally recognized method to establish an acceptable minimum incubation time (MIT) of less than seven days for biological indicators (BIs) used to monitor the effectiveness of sterilization processes. A method to qualify a reduced incubation time (RIT) of less than seven days was published by FDA in the Center for Device and Radiological Health (CDRH) 

Guide for the Validation of Biological Indicator Incubation Time

 in 1986 (1). The international community, however, has been reluctant to accept this method due to the lack of published results to support the chosen test requirements and acceptance criteria. In 2007, the ISO/TC198 Working Group 4 (WG4), Biological Indicators, initiated a New Work Item Proposal (NWIP) to develop an ISO Technical Specification (TS) that would contain a method for validation of an MIT for BIs. The position taken by ISO WG4 was that any proposed method should be supported by results published in a peer-reviewed technical journal. At the request of the ISO WG4, the Association for the Advancement of Medical Instrumentation (AAMI) BI Working Group 4 formed an ad hoc committee in 2008 to develop a plan for such research and the publication of the results. To date, two articles have been published that contain results that support a rationale for a method to determine an MIT for BIs (2, 3).

This article is a result of the ad hoc committee’s efforts and proposes a method to establish an MIT for BIs. The method is based on an analysis of the grow-out times for more than 10,000 BIs exposed to a variety of sterilization processes (2, 3). The BIs used in these studies were manufactured in conformance to ISO 11138-1 and specific subsections, as appropriate, and released for commercial use (4). The manufacturer’s incubation instructions were followed as appropriate.




The studies included several BI configurations, including both self-contained BIs and conventional spore strips. The spore carriers included both paper and stainless steel. The two most common species of bacterial endospores used in the manufacture of BIs, 

, were tested. The growth media for the studies were controlled and included media ampuls for self-contained BIs and tubed media for spore strips as previously described (2, 3).

The sterilization processes evaluated were moist heat at 121, 132, 134, and 135 °C, ethylene oxide (EO) gas (Oxyfume 2000), hydrogen peroxide (H

) gas. These processes were used for BI testing because they represent the most commonly used sterilization processes monitored with BIs and also exhibit different lethal mechanisms that include H

O-mediated energy transfer, alkylation, and oxidation.

Separate groups of BIs were evaluated under the following conditions:  

Exposed to the various sterilization processes in a manner that resulted in 30 to 80% of the BIs testing nonsterile (positive for growth). This quantal zone result, in which a fraction of the BIs are sterile and a fraction are nonsterile, meets the requirement of the FDA CDRH protocol 

Exposed using the times calculated from the Calculated Survival Time (CST) formula published in ISO 11138-1, ISO 14161, and the 
  

All exposed BIs were incubated until a nonsterile result was observed or for a maximum of seven days. A nonsterile result was based on evidence of microbial growth such as turbidity and/or a change in color of the pH indicator in the growth medium.   
  
  Key findings from these grow-out time studies that are relevant to identification of a results-based MIT method are listed as follows: 

The ranges of grow-out times observed approximate a normal distribution irrespective of the type of sterilization process and associated lethal mechanism(s) or the spore species on the BI.

An inverse relationship exists between the number of surviving spores on an exposed BI and the observed grow-out time. BIs that have a greater number of surviving spores have a shorter grow-out time than those that have fewer surviving spores.

Prolonged grow-out times were found for a small percentage (~1.1%) of the BIs tested. Such prolonged grow-out times were observed with all of the sterilization processes (moist heat, EO, H2O2 vapor, and ClO2 gas) and only occurred when the sterilization exposure gave a quantal-zone result.

BIs demonstrating prolonged grow-out times were not observed when exposed to a sterilization exposure at the CST. BIs exposed in this manner had faster and more consistent grow-out times when compared to BIs exposed under conditions that gave quantal-zone results. 
  


One of the issues negatively affecting the widespread acceptance of the FDA CDRH protocol is the lack of test results to support the method itself and its acceptance criteria. The FDA CDRH protocol requires the performance of three sterilization exposures, each of which use 100 BIs and must result in 30 to 80% of the BIs testing nonsterile. Critics view the 30 to 80% values and the additional requirement for greater than 97% correlation of the RIT results to the seven-day incubation time results as arbitrary. The international community has consistently raised the issue of why 30 to 80% survival is appropriate and why a 100% correlation to the seven-day incubation time results is not required. No rationale or supporting test results have been presented to justify the use of the 30 to 80% nonsterile and the greater than 97% correlation values.

The exposure of microorganisms in general, and spores in particular, to a sterilization process results in a range of effects characteristic of the sterilization process itself and the extent of treatment. As the extent of treatment progresses, a BI has fewer and fewer surviving spores until, at some treatment level, there are no surviving spores. In the quantal zone, some of the BIs have only one surviving spore and others have a low number of survivors.

For exposed BIs that have only one surviving spore, one of two conditions can exist: the spore has not been damaged in a manner that significantly affects either germination time or post-germination growth rate, or damage has occurred that significantly affects one or both of these characteristics. When a sterilization exposure gives a quantal-zone result of 50% of the BIs testing nonsterile, the Poisson distribution predicts approximately 35 BIs will have only one surviving spore and the remaining 15 will have more than one surviving spore (2). From the results of the previously reported grow-out time studies, it is clear that most of the BIs with only one surviving spore, and, most likely, all the BIs that have more than one surviving spore, exhibit grow-out times within a relatively narrow range. Also, a very low number of BIs would be expected to exhibit a prolonged grow-out time; prolonged times were observed for 2/458 (0.44%) and 3/309 (0.97%) BIs for moist heat and EO sterilization processes, respectively.  

BIs that have prolonged grow-out times are more likely to occur when the exposure conditions result in a lower percentage being nonsterile, for example 35% versus 75%. For a 35% nonsterile result, the average number of viable spores per exposed BI is 0.43 versus 1.39 when 75% of the BIs test nonsterile. As the percent of nonsterile BIs decreases, proportionally more of the nonsterile BIs will have only a single viable spore. For a 35% nonsterile result, approximately seven of the 35 nonsterile BIs have two or more surviving spores; for a 75% nonsterile result, approximately 40 of the 75 nonsterile BIs have two or more surviving spores (2).  

To develop a viable MIT method, the use of a sterilization exposure that results in a consistent low number of surviving spores is an important prerequisite. Treatment of BIs in a resistometer at an exposure time specified by the CST calculation will result in a more consistent and reproducible method to determine an MIT as opposed to producing 30 to 80% nonsterile BIs. 

The CST, as indicated in Equation 1, is an exposure time that takes into account both the spore population level and the D

 value and is intended to result in 100% of the BIs testing nonsterile (4-6). A BI subjected to a CST exposure would be expected to have approximately 100 surviving spores; the results of the previous grow-out time studies showed that when BIs were exposed to the CST, no prolonged grow-out times were observed in 1297 CST exposures.

value) x (log10 population level - 2)  	

 value are in minutes. The extent of treatment for a CST exposure is marginally less than that used to obtain the 30 to 80% nonsterile result required by the FDA CDRH protocol. BIs treated in a CST sterilization exposure will challenge the incubation medium and conditions with ~100 spores that have received a significant lethal insult (~99.99 % of the spores inactivated due to the lethal action(s) of the sterilizing agent). This level of microorganisms is similar to that used in the microbiological-medium growth promotion testing and bacteriostasis/fungistasis testing where inoculations are performed with low levels (≤ 100) of test microorganisms (7).


Based on the results of the testing summarized in the two previously published studies and the subsequent analysis of the data, the following method for determination of an MIT for BIs is proposed:

Obtain a minimum of 100 BIs from each of three lots, which are produced from different spore crops. 

Calculate the survival time for each BI lot using Equation 1.

Expose each lot of BIs at the CST in separate resistometer runs. 

Use recovery methods specified by the BI manufacturer (or specified growth medium) and defined environmental conditions. 

Incubate until all BIs test nonsterile or until seven days incubation time has been completed. (Note: If all BIs do not test nonsterile after seven days of incubation, reconfirm D

Record the time for all BIs to test nonsterile.

The MIT is the longest time found for a nonsterile result for the three BI lots. 
  

For determination of the CST, it is recommended that the D

 value be calculated using the Holcomb-Spearman-Karber method (8). This method uses all of the fraction negative results from the series of sterilization exposures to calculate the “mean time to sterility” and associated D

 value calculated in this manner is more likely to be the most accurate and provide the best D

 value estimate from a statistical perspective.
  



In developing a new method for MIT determination, it is important to consider the evolution of the practices for monitoring and control of sterilization processes during the past 25 or more years. For example, early moist-heat sterilizers had only a chamber pressure gauge and a thermometer in the drain line to monitor the process. Active electromechanical control of the process did not exist, and, consequently, the results of BI testing were critical to determining process efficacy.

Today, we operate under the umbrella of process validation. Validation requires comprehensive documentation of the physical/chemical aspects of each sterilization cycle. The documentation must be sufficient to show that the process was executed as expected and that the process specification was met in its entirety. Modern sterilizers, in contrast to early vessels, have sophisticated monitoring and control systems that provide detailed run records for each sterilization process.

In current practice, results of BI testing are subordinate to the recorded physical/chemical results; a sterilization run is nonconforming if the physical/chemical requirements are not met even though all BIs are inactivated. The use of BIs in the current context of sterilization process-monitoring control must be matched to our current understanding of BI outgrowth.

Extensive grow-out time testing for BIs exposed to four different sterilization processes under controlled conditions has given a very clear picture of the kinetics of spore germination and outgrowth for both unexposed BIs and BIs exposed to differing extents of sterilant treatment.  Unexposed BIs grow out the most rapidly and exhibit the least variation in grow-out time. As would be expected, BIs exposed to a sterilization process exhibit longer grow-out times on average and also show more variability in grow-out times. The longer average grow-out time is clearly related to the presence of a lower number of viable spores after sterilant exposure and also likely due to some damage to the spore metabolic and/or genetic apparatus.

BIs exposed to a sterilant that results in a quantal-zone outcome demonstrate more variability than unexposed BIs but also show a low frequency of BIs with a prolonged grow-out time. The MIT method proposed in this paper using the CST sterilization exposure takes into account this increased variability but properly avoids the occurrence and inappropriate influence of “outlier” results. In contrast, an unforeseen consequence of the 30-80% survival window of the FDA CDRH protocol is that the RIT is driven by the outlier results of large percentages of BIs having only a single spore. This effect can be seen in the fact that a shorter incubation time can be achieved if the average nonsterile results are closer to 80% versus 30%, which is a result attributed to the difference in the fraction of nonsterile BIs having only a single surviving spore. This consequence is clearly undesirable for a test method and demonstrates a lack of robustness and reproducibility. 

The adoption of the proposed MIT method would not be predicted to reduce the effectiveness of BIs for monitoring and control of sterilization processes but rather provide a means for reproducible determination of BI grow-out time.

, “Understanding Biological Indicator Grow-Out Times,” www.pharmtech.com/BIgrowout, Jan. 2, 2010, accessed May 6, 2013.  

Sterilization of Health Care Products--Biological Indicators--Part 1: General Requirements 

Sterilization of Health Care Products--Biological Indicators--Guidance for the Selection, Use and Interpretation of Results

 (US Pharmacopeial Convention, Rockville, MD 2013), p. 2659-2662.  

 General Chapter <61>, “Microbiological Examination of Nonsterile Products: Microbial Enumeration Tests” (US Pharmacopeial Convention, Rockville, MD 2013).  

I.J. Pflug, R.G. Holcomb, and M.M. Gomez, “Thermal Destruction of Microorganisms,” in 

Disinfection, Sterilization, and Preservation

, S. Block, Ed. (Lippincott, Williams, & Wilkins, Philadelphia, PA, 2001), pp. 79-129. 

The industry lacks an accepted method for establishing a minimum incubation time (MIT) of less than seven days for biological indicators (BIs). The authors propose an MIT method that provides a means for reproducible determination of BI grow-out time.