Understanding Biological Indicator Grow-Out Times—Part II

June 2, 2013
Pharmaceutical Technology

Volume 37, Issue 6

The authors report on the range and distribution of grow-out times for biological indicators exposed to sublethal sterilization processes.

Biological indicators (BIs) are used to monitor the efficacy of sterilization processes for medical products. BIs contain high numbers of bacterial spores (generally 104 to 106 that are highly resistant to the sterilization process for which they are designed. This paper is Part II of a series reporting on the range and distribution of grow-out times for BIs exposed to sublethal sterilization processes. The authors describe studies to confirm and extend their original findings that the grow-out times for a set of nonsterile BIs approximates a normal distribution and that one or more BIs from such a set would occasionally exhibit delayed outgrowth.

This paper is Part II of a series reporting on the range and distribution of grow-out times for biological indicators (BIs) exposed to sublethal sterilization processes. Part I described grow-out times for self-contained BIs with Geobacillus stearothermophilus spores on paper carriers that were exposed to moist-heat sterilization processes at 121 °C (1). The studies described in this paper were performed to confirm and extend the authors' original findings that the grow-out times for a set of nonsterile BIs approximates a normal distribution, and occasionally, one or more BIs from such a set would exhibit a delayed outgrowth. Delayed nonsterile BIs were almost exclusively found in sets of BIs where the average number of surviving colony-forming units (CFU) was estimated at less than 1.0/BI unit.

The previous studies exposed BIs to moist-heat sterilization processes that gave results that met the FDA protocol for establishing a reduced incubation time (RIT) (2). Sets of exposed BIs that have 30 to 80 nonsterile out of the 100 tested have a range estimate of 0 to 5 CFU per BI (1). In the studies reported in this article, the FDA RIT protocol was again followed, but testing was also performed using exposures based on the calculated survival time that is referenced in the US Pharmacopeia (USP) and Association for the Advancement of Medical Instrumentation (AAMI)/International Organization for Standardization (ISO) documents (3–5). The calculated survival exposure time results in BIs with approximately 100 surviving CFU, a much higher average value when compared to the results found after exposure of BIs to the conditions required to meet the FDA RIT protocol.

For the studies reported in this article, the moist-heat sterilization processes tested were expanded to include exposures at 132 °C, 134 °C, and 135 °C. BIs exposed to hydrogen peroxide (H2O2) vapor were self-contained with 106 spores of G. stearothermophilus inoculated on stainless-steel discs. BIs exposed to ethylene oxide (EO) gas and chlorine dioxide (ClO2) gas sterilization processes were also tested. For the EO and CIO2 processes, BIs with Bacillus atrophaeus spores were used; both paper spore strips and self-contained BIs with paper carriers were tested.

Materials and methods

BIs. A list of the BIs used in the studies is shown in Table I, which lists the spore species, spore crops, sterilization process, BI configuration, spore carrier, spore population, and BI lots used. Seventeen different spore crops were used that were produced over a period of five years; samples were tested from 29 commercial batches of BIs. The BIs tested had spores inoculated onto paper or stainless-steel carriers. Spores inoculated onto glass fiber discs were not tested. Specific types and lots of microbiological media were used for grow-out time testing to minimize variability associated with the recovery media aspect of the testing. Each BI lot was produced with a separate media lot; therefore, 29 different media lots were included in the testing.

Table I: Biological indicator (BI) test sample matrix.

Incubation of BIs. Unexposed controls and exposed BIs were incubated for seven days and the time to exhibit a nonsterile result was recorded. G. stearothermophilus BIs were incubated at 60 ± 2 °C in a Smart-Well incubator system (Mesa Laboratories). Grow-out of a BI was detected by measuring color change in the growth medium and was recorded in 0.01-hour increments. The test BIs were removed from the Smart-Well incubator after 24 hours and placed in a conventional incubator at 60 ± 2 °C for the remainder of the seven days. Bacillus atrophaeus BIs were incubated in a conventional incubator at 37 ± 2 °C and were monitored for nonsterility by visual inspection at intervals between 30 and 60 minutes when nonsterile BIs were frequently occurring. Intervals were extended to several hours when there were less frequent changes in numbers of nonsterile BIs. When it appeared that the majority of the nonsterile BIs were detected, observation intervals were significantly decreased for the remainder of the seven days. The delayed nonsterile BI would be detected, but the precise time delay would not be known.

Delayed nonsterile definition. A BI was classified as a delayed nonsterile unit if the timing of the observed growth was 150% or greater than the incubation time needed for 95% of the other BIs in the test set to indicate the nonsterile response.

Equipment. Moist-heat, H2O2 vapor, and EO gas sterilization processes were performed in resistometers that met the requirements of ISO 18472: 2006 (6). All CIO2 exposures were performed in a custom resistometer manufactured by ClorDiSys. There are no published guidelines describing a resistometer for chlorine dioxide gas exposures.

Exposure conditions. Unexposed controls: Unexposed BIs were used as a control. Sets of 100 BIs for each for the various BI configurations were incubated to provide a baseline response for each spore type, carrier, media, and incubation condition.

Calculated survival time exposure: The calculated survival time was based on the resistance of the specific spores/carrier combination to a particular sterilization mode (3–5). This exposure was designed to reduce the number of viable spores by three to four orders of magnitude, which resulted in each BI having approximately 100 surviving spores. The formula for calculation of the survival time exposure is:

Survival time exposure = D-value x (log10 of the population – 2)

Because the calculated survival time exposure was intended to yield a spore concentration of approximately 100 spores per BI, all BIs were expected to be nonsterile. If all BIs were not nonsterile, then one had to review the two input values required for this calculation—the D-value and the population count. The D-value is a more complicated value to measure accurately and if the D-value is overstated, the calculated survival time exposure would be longer in duration and thus might result in some/all of the BIs testing as sterile. For a surviving population of approximately 100 spores/BI, all exposed BIs should be nonsterile 100% of the time.

FDA RIT protocol exposure: The FDA RIT protocol exposure requires a sublethal process that, on groups of 100 BIs, yields at least 30 and no more than 80 positive BIs. The most probable number of spores per BI when 30 out of 100 BIs tested are nonsterile is 0.357. The most probable number of spores per BI when 80 out of 100 BIs tested are nonsterile is 1.609. This results in a practical range of 0–5 surviving CFU/BI. Additionally, some BIs in every set of 100 would not have any surviving spores (6).

Results and discussion

Moist-heat exposures. All moist-heat exposures were performed with self-contained BIs with 105 spores of G. stearothermophilus inoculated onto paper carriers (see Table I).

Unexposed controls: Nine lots of 100 and one lot of 80 BIs were incubated at 60 ± 2 °C to determine a baseline for time to nonsterile results (see Table II). The first nonsterile BI was detected at 2.27 hours. Ninety-five percent of all BIs exhibited nonsterile results by 3.9 hours. The incubation duration between the first nonsterile BI and 95% BIs nonsterile was 1.63 hours. There was a slight tail in the curve due to a small number of BIs exhibiting somewhat longer grow-out times. It was concluded that this distribution of grow-out times was due to natural variation often seen in biological systems.

Table II: Time-to-nonsterile results for Geobacillus stearothermophilus biological indicators (BIs) exposed to moist heat.

Calculated survival time exposures at 121 °C: Five lots of 100 BIs were exposed to the calculated survival time process and incubated as previously described (see Table II). The average time for the first nonsterile BI for the five lots tested was 2.98 hours. Ninety-five percent of all BIs exposed to these conditions were nonsterile in 4.9 hours. The 4.9 hours of incubation required to observe 95% of the BIs as nonsterile was 20% longer than that found with the unexposed controls. No delayed nonsterile BI's were observed in this series of exposures.

FDA RIT protocol exposures: Three lots of 100 of BIs were exposed to a moist-heat sterilization process that resulted in 30 to 80 nonsterile BIs for each of the lots (see Table II).

The average time for the three lots exposed at 132 °C for the first nonsterile BI was 3.42 hours. Ninety-five percent of the BIs exposed to 132 °C were nonsterile at 7.0 hours. The incubation duration from first nonsterile BI to 95% nonsterile was 3.58 hours, which was nearly twice as long as that observed for the calculated survival time exposure.

At 134 °C, the average time for the three lots tested for the first nonsterile BI was 3.27 hours, which was slightly less than that observed at 132 °C. Ninety-five percent of the BIs exposed to 134 °C were nonsterile in 8.0 hours, which was one hour longer than at 132 °C. The incubation duration from the first nonsterile BI to 95% nonsterile was 4.73 hours, which was about one hour longer than that observed for the exposures at 132 °C.

At 135 °C, the average time in the three lots for the first nonsterile BI was 3.38 hours. Ninety-five percent of the BIs exposed to 135 °C were nonsterile at 7.9 hours. The incubation duration from the first nonsterile BI to 95% nonsterile was 4.73 hours, which was also approximately one hour longer than that observed for the exposures at 132 °C.

There were two delayed nonsterile BIs observed, one exposed at 132 °C and one exposed at 134 °C. The BI exposed to 132 °C grew out between 8.27 hours and 120 hours of incubation. The BI exposed to 134 °C grew out between 9.38 hours and 168 hours of incubation. The actual grow-out times for these two BIs is believed to be less than 120 and 168 hours, but these BIs were not monitored after 10 hours so the exact grow-out time was not determined.

Using the probability table shown in Part I of this study, the authors projected that 61% of the 335 nonsterile BIs exposed at 132 °C and 134 °C were likely to contain only one surviving spore (1). The two delayed grow-out results accounted for only 0.6% of the nonsterile BIs.

The grow-out time results for the moist-heat exposures at 121 °C, 132 °C, 134 °C, and 135 °C are presented in vertical scatter plots (see Figure 1, Panel A). The vertical scatter plots illustrate the variability in grow-out time for the different exposure temperatures and different lots of BIs. The difference in grow-out time results between the survival time and RIT exposures is clearly illustrated.

The 121 °C calculated survival time exposure results were much more consistent from first to last nonsterile BI grow out. There were a few BIs that were slower to grow out, but the delay was not significant. Lots 1, 2, 3, and 4 had one to three BIs that were noticeably slower in grow out; the delay average was approximately 0.5 hours.

All of the RIT exposure results were much more variable than the results for the calculated survival exposures. The grow-out time results for the moist-heat exposures are graphically illustrated in Figure 2, Panel A.

Hydrogen peroxide vapor exposures. All H2O2 exposures were performed with self-contained BIs with 106 spores of G. stearothermophilus inoculated onto stainless steel discs (see Table I).

Unexposed controls: Three lots of 100 BIs were incubated at 60 ± 2 °C to determine a baseline for the time required for BIs to exhibit nonsterile results (see Table III). The average time for the first nonsterile BI in the three lots was 2.98 hours. Ninety-five percent of all BIs tested exhibited nonsterile results by 5.3 hours. The incubation duration between the first nonsterile BI and 95% BIs nonsterile was 2.3 hours.

Table III: Time-to-nonsterile results for Geobacillus stearothermophilus biological indicators (BIs) exposed to hydrogen peroxide vapor.

Calculated survival time exposures: Three lots of 100 BIs were exposed to the calculated survival time and incubated as previously described (see Table III). The average time for the first nonsterile BI in the three lots was 3.54 hours. Ninety-five percent of the BIs were nonsterile in 6.0 hours. This incubation duration from the first nonsterile BI to 95% nonsterile BIs was 2.46 hours. This result was very similar to that observed with the unexposed controls. Two of the three lots tested did not yield 100 nonsterile BIs; one lot had 99 nonsterile BIs and the second lot had 98 nonsterile BIs. It is believed that the calculated survival time exposure was inappropriately long due to inaccuracy of the D-value determination. If the D-value is overstated, the average number of surviving spores would be less than approximately 100 resulting in some of the BIs being sterilized in the survival time exposure.

One delayed nonsterile BI was observed in one of the lots tested. For this BI, the grow-out time was 4.11 hours longer than that of the adjacent nonsterile test result. This delayed nonsterile BI took 223% longer to exhibit nonsterility than 95% of the nonsterile BIs. This was the only delayed nonsterile BI observed for the calculated survival time exposures for any of the processes examined in this study.

FDA RIT protocol exposures: Three lots of 100 BIs were exposed to H2O2 vapor that resulted in 30 to 80 BIs nonsterile per lot (see Table III). The average time for the first nonsterile BI for the three lots was 4.7 hours. Ninety-five percent of the BIs were nonsterile by 9.9 hours of incubation. The duration of incubation from the first nonsterile BI to 95% nonsterile was 5.2 hours, which was nearly twice as long as that for the BIs exposed to the calculated survival time exposure.

There were also three delayed nonsterile BIs observed; one BI from one lot and two BIs from another lot. The delayed grow-out time was approximately 5.5 hours longer than that of the adjacent nonsterile test results. This incubation duration was 190% of the time to 95% of the nonsterile BIs.

It appeared that significant nonlethal spore damage occurred at this exposure condition, which was designed to yield a surviving population of 0 to 5 CFU (7–9). Using the probability table in Part I of this study, the authors projected that 66% of the 160 nonsterile BIs would contain only one surviving CFU (1). However, only 2.5% of the nonsterile BIs had a delayed response.

All grow-out time results for all H2O2 exposures are presented in vertical scatter plots in Figure 1, Panel B. They are also presented graphically in Figure 2, Panel B.

Ethylene oxide gas exposures. All EO exposures were performed with self-contained BIs with 106 spores of B. atrophaeus inoculated onto paper carriers (see Table I).

Unexposed controls: Three lots of EO BIs were incubated at 37 ± 2 °C in a conventional incubator to determine a baseline value for the time required for BIs to exhibit nonsterile results (see Table IV). The average time for the first nonsterile BI in the three lots was 7.5 hours. Ninety-five percent of all BIs were nonsterile in 10.2 hours. The incubation time between the first nonsterile BI and 95% nonsterile BIs was 2.7 hours; no delayed nonsterile BIs were observed.

Table IV: Time-to-growth positive results for Bacillus atrophaeus biological indicators (BIs) exposed to ethylene oxide gas.

Calculated survival time exposures: The calculated survival time exposures were performed using three lots of 100 BIs. The average time for the first nonsterile BI in the three lots was 18.2 hours. Ninety-five percent of all BIs exposed to these conditions were nonsterile in 21.3 hours. The duration of incubation between the first nonsterile BI and 95 % nonsterile was 3.1 hours. This duration was approximately 15% longer than that found with the unexposed controls. No delayed nonsterile BIs were observed.

FDA RIT protocol exposures: Six lots of BIs were exposed to yield 30 to 80 BIs nonsterile per lot (see Table IV). The average time for the first nonsterile BI for the six lots tested was 21.96 hours. Ninety-five percent of all BIs exposed to these conditions were nonsterile in 62.9 hours. The duration of incubation from the first nonsterile BI to 95% nonsterile was 40.94 hours, which was approximately 13 times longer than that found with the calculated survival time exposures.

Table V : Time-to-nonsterile results for Bacillus atrophaeus biological indicators (BIs) exposed to chlorine dioxide gas.

Delayed nonsterile BIs were observed with two of the lots. Two BIs were observed in one lot and one BI was observed in the second lot. The delayed nonsterile BIs grew out between 72 hours and 168 hours of incubation. Using the probability table in Part I of this study, the authors projected that 66% of the nonsterile BIs would contain only one surviving CFU (1). However, only 0.97% of the nonsterile BIs exhibited a delayed grow-out time.

The grow-out time results for all BIs exposed to EO gas are illustrated in the vertical scatter plots in Figure 1, Panel C. A graphical expression of these data appears in Figure 2, Panel C.

Chlorine dioxide gas exposures. All ClO2 exposures were performed with paper spore strips and tubed media culture sets. The paper strips contained 106B. atrophaeus spores and were packaged in Tyvek Mylar envelopes (see Table I).

Unexposed controls: The unexposed controls consisted of three lots of spore strip culture set BIs which were incubated in a conventional incubator at 37 ± 2 °C. The average time for the first nonsterile BI for the three lots was 5.5 hours. Ninety-five percent of all BIs were nonsterile in 6.1 hours. The incubation duration between the first nonsterile BI and 95% nonsterile was 0.6 hours. No delayed nonsterile BI's were observed in this series of tests.

The grow-out time for the ClO2 unexposed controls was faster than the time required for the EO gas unexposed controls. The spores, carrier, and incubation conditions were the same. The recovery media used to culture the spore strip had a different formulation for the CIO2 BIs, which resulted in the faster grow-out time.

Figure 1: Vertical scatter plots of grow-out time results for exposure to moist heat (Panel A), hydrogen peroxide vapor (Panel B), ethylene oxide gas (Panel C), and chlorine dioxide gas (Panel D). RIT = reduced incubation time. (ALL FIGURES COURTESY OF AUTHORS)

A vertical scatter plot for each of the exposures is shown in Figure 1, Panel D. A graphical representation of these data is illustrated in Figure 2, Panel D.

Calculated survival time exposures: Two lots of ClO2 spore strip culture sets were used for these exposures. The average time for the first nonsterile BI for the two lots tested was 12.75 hours. Ninety-five percent of all BIs exposed to these conditions were nonsterile in 15.5 hours. The incubation duration from the first nonsterile BI to 95% nonsterile was 2.75 hours, which was 4.5 times longer than that observed with the unexposed controls. No delayed nonsterile BIs were observed.

Figure 2: Graphical illustration of grow-out time results for exposure to moist heat (Panel A), hydrogen peroxide vapor (Panel B), ethylene oxide gas (Panel C), and chlorine dioxide gas (Panel D). RIT = reduced in incubation time.

FDA RIT protocol exposures: The RIT exposures were performed using four lots of spore strip culture set BIs. The average time for the first nonsterile BI for the three lots tested was 15 hours. Ninety-five percent of all BIs exposed to these conditions were observed nonsterile in 32.25 hours. The incubation duration from the first nonsterile BI to the 95% nonsterile was 17.25 hours, which was 6.27 times longer than that observed with the calculated survival time.

Three delayed nonsterile BIs were observed. Two BIs exhibited a delayed response in one lot and one BI in each of the other lots tested. Using the probability table in Part I of this study, the authors projected that 53% of the nonsterile BIs would contain only one surviving CFU (1). However, only 1.2% of the nonsterile BIs exhibited a delayed response. One lot of BIs was tested twice in this series, thus there are four sets of results but only three separate lots of BIs.

The grow-out time results for all BIs exposed to ClO2 gas is illustrated using vertical scatter plots in Figure 1, Panel D. A graphical representation of these data sets is shown in Figure 2, Panel D.

Conclusions

The study showed several key findings as outlined below:

  • Grow-out time results, regardless of the mode of sterilization, approximates a normal distribution for all exposure conditions.

  • This study indicates that the dynamics of spore germination and out-growth are very similar regardless of the sterilization mode to which the BIs are exposed. This study included moist heat at 121 °C, 132 °C, 134 °C, and 135 °C as well as H2O2 vapor, EO gas, and ClO2 gas. It is unlikely that different bacterial endospores would respond differently to other sterilizing agents.

  • The data continue to support the conclusion that there is an inverse relationship between the number of surviving spores on a BI and the overall range of grow-out time. The time required to obtain an acceptable cell density and/or cumulative metabolic activity requires less time when the starting level of viable spores is higher than when it is lower.

  • Grow-out times were shorter for G. stearothermophilus spores than for B. atrophaeus spores.

  • The timing of the calculated survival time exposures yielded nonsterile results faster and more consistently than the FDA RIT protocol exposures. No delayed nonsterile results were observed in the calculated survival time exposures that provided results with all BIs nonsterile.

  • Delayed nonsterile BIs were observed with all sterilization modes tested. All delayed nonsterile BIs were observed when the exposures yielded dichotomous results (some BI replicates nonsterile and some BI replicates sterile).

  • Delayed nonsterile BIs almost certainly contained a single viable spore. If two or more spores are present on the BI and only one exhibited delayed germination and/or outgrowth, the other spore would mask this condition and yield a nonsterile result in a time similar to a BI with a single spore that was not severely damaged.

  • Delayed nonsterile results were found in a very small portion of the BIs exposed. The total number of BIs exposed that yielded a dichotomous result was approximately 2400. The number of BIs that were nonsterile was 1188. The total number of delayed nonsterile BIs observed was 13 or 1.1% of all nonsterile BIs. Delayed nonsterile results for each sterilization mode evaluated were as follows:

—Moist heat: 0.44% delayed nonsterile BIs

—H2O2: 2.5% delayed nonsterile BIs

—EO: 0.97% delayed nonsterile BIs

—ClO2: 1.2 % delayed nonsterile BIs.

*John R. Gillis, PhD, is retired president of SGM Biotech, Inc., 10 Evergreen Dr., Bozeman, MT, gillisphd@gmail.com; Gregg A. Mosley is retired president of Biotest Laboratories, Minneapolis, MN; John B. Kowalski, PhD, is principal consultant of SteriPro Consulting, Sterigenics International, Inc., Oak Brook, IL, Garrett Krushefski is vice-president, biological indicator operations; Paul T. Nirgenau is contract studies coordinator; Kurt J. McCauley is director of R&D and Apex Labs, all at Mesa Laboratories, Bozeman, MT; and Philip M. Schneider is a senior consultant with LexaMed, 705 Front Street, Toledo, OH 43604, pschneider@lexamed.net, and is Convener of ISO/TC 198 WG4 Biological Indicators.

*To whom all correspondence should be addressed.

Submitted: July 11, 2012. Accepted: July 18, 2012.

References

1. J.R. Gillis et al, Pharm. Technol. online http://www.pharmtech.com/pharmtech/Peer-Reviewed+Research/Understanding-Biological-Indicator-Grow-Out-Times/ArticleStandard/Article/detail/651903, Jan. 2, 2010, accessed May 6, 2013.

2. FDA Guide for Validation of Biological Indicator Incubation Time (1986).

3. USP 32–NF 27 General Chapter <1211>, "Sterilization and Sterility Assurance of Compendial Articles," 720–723.

4. ISO 11138–1: 2006, Sterilization of health care products—Biological indicators—Part 1: General requirements, (ISO, Geneva, 2006).

5. ISO 14161: 2009, Sterilization of health care products—Biological indicators—Guidance for the selection, use and interpretation of results (ISO, Geneva, 2006).

6. ISO 18472: 2006, Sterilization of health care products—Biological and chemical indicators—Test equipment (ISO, Geneva, 2006).

7. H.O. Halverson and N.R. Ziegler, J Bacteriol. 25 101-121 (1933).

8. A.D. Russell, The Destruction of Bacterial Spores (Academic Press, London, 1982).

9. A. Hurst and G.W. Gould, The Bacterial Spore 2 (Academic Press, London, 1983).