Radiation Sterilization: Dose Is Dose

Radiation is a scientific term that describes transmitting energy through space. This term includes microwaves, ultraviolet, electron beam (E-beam), gamma, and X-rays. However, ionizing radiation (i.e., gamma rays, X-rays, E-beam) typically is used to terminally sterilize product. With the expansion of the use of E-beam and X-rays, the potential impact on microbial inactivation associated with sterilization of medical devices has been a point of discussion during recent updates of industry standards.

Radiation sterilization relies on ionizing radiation to inactivate microorganisms. Absorption of a sufficient amount of radiation will negatively affect the microorganism's ability to reproduce. The interaction of the ionizing energy with matter is key to this process. All ionizing radiation modalities are capable of sterilization. The question is whether the same dose delivered by these modalities is equal in its ability to inactivate microorganisms.

Previous articles have reported conflicting results when comparing radiation resistance for different radiation modalities. When reviewing the details for the procedures used for these published results, the primary issues that might affect the results and conclusions were difficulty in appropriate delivery (i.e., narrow dose ranges) and accurate measurement of the radiation process. In addition, some studies had questionable methods for process validation and preparation of the test articles.

This study evaluated the gamma and E-beam radiation resistance of microorganisms using two microbial challenges: (1) natural product bioburden on a nonwoven cellulosic material (bandage) and (2) biological indicator (BI) using paper carriers inoculated with Bacillus pumilus spores (10⁶ population).

The bandage evaluated in this study was selected because of its previously characterized bioburden population. The number of production batches, product sample size, and incremental doses selected to evaluate the bandage material match the requirements in Method 2A (per ANSI/AAMI/ISO 11137-2:2013/(R)2019).¹  Method 2A does not require that product bioburden be performed to determine the sterilization dose; however, for the purposes of this study, product testing was performed to demonstrate stability of the bioburden over the time of the test period, and the sample size was selected to match the requirements detailed in ANSI/AAMI/ISO 11737-1:2018.²  Rather than determining the sterilization dose, this study was designed to compare the microbial lethality; therefore, only the incremental doses were performed.

The B. pumilus spore was selected as a test article to provide a positive control to demonstrate microbial lethality. The number of BI lots and incremental doses were selected to match the requirements for Method 2A. As BIs are manufactured to provide a consistent spore population, the sample size selected for BI testing was reduced from the requirements for Method 2A. In addition, the number of incremental doses was reduced due to the known resistance of the B. pumilus spore. The stability of the BI over the time of the test period was demonstrated using population counts prior to and following testing. This species was selected based on its demonstrated, consistent response to radiation, due to a higher radiation resistance than other Bacillus species. It is important to note that the use of a spore challenge model for validation of a product sterilization dose is not recommended (11137-1, sections 1.2.3 and A.1.2.3).³ 

The range of radiation dose rates and energies selected for this study are typical of those that might be utilized during verification testing or routine sterilization processing. The radiation resistances of the test articles were assessed using gamma radiation (emits two rays with energies of 1.17 and 1.33 MeV) delivered at two separate dose rates (0.37 and 12.9 kGy/h) and E-beam radiation (10 MeV) delivered at two separate dose rates (3,100 and 36,000 kGy/h).

The test articles consisting of nonwoven cellulosic bandages (2 × 3.5 inches) from each of three batches were individually packaged in pouches, and B. pumilus (ATCC 27142) BI paper carriers from each of three lots were individually packaged in pouches. All test articles were submitted for irradiation.

Test articles were irradiated together for each of the doses selected. Twenty bandages from each of three batches and five BIs from each of three lots were irradiated together. The bandages were irradiated at nine incremental doses (2, 4, 6, 8, 10, 12, 14, 16, and 18 kGy), and the BIs were irradiated at six incremental doses (2, 4, 6, 8, 10, and 12 kGy). Each of the incremental doses was delivered using four dose rates: (1) gamma radiation at a dose rate of 0.37 kGy/h, (2) gamma radiation at a dose rate of 12.9 kGy/h, (3) E-beam radiation at a dose rate of 3,100 kGy/h, and (4) E-beam radiation at a dose rate of 36,000 kGy/h.

For consistent measurement of the irradiation doses, a common dosimetry system was used. The irradiation dose delivered was measured using FWT-60 radiochromic film dosimeters (Far West Technology, Santa Ana, CA). During several of the irradiation runs, alanine reference dosimeters from the National Physical Laboratory (Middlesex, UK) were placed side-by-side with the FWT-60 film dosimeters to verify the traceability of the Far West Technology films.

After irradiation, all test articles were returned to the laboratory for testing. A test of sterility was conducted on each bandage by aseptic transfer into a container of sterile soybean casein digest (SCD) broth. The SCD containers were incubated at 28 to 32°C for 14 days per 11137-2.¹  The containers were periodically examined and the results documented as the number of units positive for growth/total number of units tested. A population count was conducted on each BI by aseptic transfer to a test tube containing 10 mL sterile water, and the paper carrier was homogenized using a sterile pestle. The BI suspensions were serially diluted in sterile water to obtain concentrations of approximately 30 to 300 colony-forming units (CFU) per milliliter. Pour plates were prepared using molten SCD agar and incubated at 30 to 35°C for three to five days. Upon completion of incubation, the plates were enumerated and the surviving populations documented.

The microbial populations present on the bandage material (i.e., bioburden) and BI paper carriers were determined preceding and subsequent to completion of radiation processing to demonstrate population stability of the test articles throughout the test period.

The bioburden population present on nonirradiated bandages was determined using a validated bioburden recovery procedure, where 10 samples from each of three batches of bandages were tested. Each bandage was immersed in sterile diluent and placed on a shaker table. The diluent with test samples were shaken at approximately 450 rpm for 15 minutes. Serial dilutions were performed from the recovery fluid using sterile water to obtain concentrations of approximately 30 to 300 CFU/mL. Pour plates were prepared using molten SCD agar and incubated at 20 to 25°C for three days, followed by incubation at 30 to 35°C for two days. Upon completion of incubation, the plates were enumerated and the initial bioburden populations documented for each batch of bandages (Table 1).

The population counts on nonirradiated BIs were determined from five BIs from each of the three lots of test articles. Each BI was aseptically transferred to a sterile blender jar containing 100 mL sterile water and homogenized for one minute. The contents of each blender jar were serially diluted using sterile water to obtain concentrations of approximately 30 to 300 CFU/mL. The dilution tubes then were exposed to a heat shock at 80°C for 10 minutes. Pour plates were prepared using molten SCD agar and incubated at 30 to 35°C for two to three days. Upon completion of incubation, the plates were enumerated and the populations documented (Table 2).

Negative controls were conducted to demonstrate that positive tests of sterility observed were attributed to the product tested and not due to testing and/or laboratory issues.

Ten or 20 bandages irradiated at 50 kGy or greater were tested each day using the same methods applied for the test articles to confirm the aseptic technique of the technicians performing the product tests of sterility for the Method 2A incremental doses. A total of 200 samples were tested for the two tests of sterility teams, and zero positive product tests of sterility were observed for the negative controls.

Method suitability testing was performed to verify that the bandage and BI paper carrier materials did not alter the growth-promoting properties of the culture media to the extent of preventing or inhibiting outgrowth of microorganisms, if present on the test articles.

Six bandages were irradiated at 50 kGy or greater and each immersed into a tube of sterile SCD broth. Duplicate broth tubes were inoculated with not more than 100 CFU each of the following microorganisms: Bacillus subtilis ATCC 6633, Candida albicans ATCC 10231, or Aspergillus niger ATCC 16404.⁴  The broth tubes were incubated at 28 to 32°C for a maximum of seven days and examined for turbidity (Table 3).

Five B. pumilus BIs were irradiated at 50 kGy or greater, immersed into tubes of sterile water, and homogenized using sterile pestles. Each tube was inoculated with 10 to 100 CFU of B. pumilus ATCC 27142 spores, and the entire contents of each tube were pour plated using molten SCD agar. The plates were incubated at 30 to 35°C for a maximum of three days and enumerated (Table 4).

The product bioburden data demonstrate that the population remained stable over the course of the study (Table 1), and therefore, no impact to the resistance analysis occurred as a result of die-off of bioburden organisms.

The first no positive (FNP) dose, first fraction positive (FFP) dose, difference between FNP and FFP doses, and DS kGy (results and calculations from the Method 2A experiment) are provided in Table 5, and the d* kGy values (an initial estimate of the dose required to achieve a sterility assurance level (SAL) of 10⁻² for an individual product batch) for the four dose rates evaluated are provided in Table 6.

The DS kGy is an estimate of the dose required to inactivate 90% of the organisms surviving a 10⁻² SAL dose (i.e., verification dose experiment). This provides an estimate of the most resistant portion of the product bioburden and is an estimate using a composite of the three batches tested for each of the four dose rates. As the study design was not intended to determine a sterilization dose, the verification dose experiment was not conducted. Therefore, the calculation for DS kGy was conducted using the following assumptions: D**kGy = DD*kGy; CD* = 2+/100; and FNP = DD*kGy. No difference in DS kGy values (Table 5) can be observed over the 5-log difference in dose rates across the two radiation source types (gamma and E-beam).

A further evaluation was conducted by reviewing the d* kGy data for each individual product batch (Table 6) using an analysis of variance (ANOVA) statistical technique (Figure 1). The analysis included a Tukey comparison (95% confidence) of values and this comparison indicates there is no significant difference in d* kGy values for the bandage material that could be detected over a 5-log difference in dose rates across the two radiation source types (gamma and E-beam).

The BI population count data demonstrate stability over the course of the study (Table 2), and therefore, no impact to the resistance analysis occurred due to the stability of the BI population.

The D₁₀ values (radiation dose required to reduce a microbial population by 90%) for the B. pumilus BIs irradiated using two irradiation sources (gamma and E-beam) and four dose rates (0.37, 12.9, 3,100, and 36,000 kGy/h) were calculated using linear regression and are reported in Table 7. The data were analyzed using an ANOVA statistical technique (Figure 2). The analysis indicated that no significant difference (95% confidence) in the D₁₀ values existed for the B. pumilus BIs that could be detected over a 5-log difference in dose rates across the two radiation source types (gamma and E-beam).

Based on the analysis of the data, no significant differences could be detected in the rate of microbial lethality across the 5-log difference in dose rates evaluated for the natural product bioburden or the BIs across the two radiation source types (gamma and E-beam).

This data indicate that the radiation resistance of microorganisms is not affected by any slight differences in energy levels and dose rates of the radiation sources typically used by sterilization facilities. Because of this, it can be concluded that the sterilization and verification doses can be safely transferred between modes of irradiation, as well as irradiation facilities, without requiring proof of equivalent microbial inactivation. As long as proof exists that the specified dose is delivered, dose is dose.

The authors thank the supporting companies and individuals involved in performing the testing. The companies included Baxter Healthcare Corporation (Round Lake, IL; product and testing), Sterigenics International, Inc. (Corona, CA; gamma irradiation), Titan Scan Technologies (San Diego, CA; E-beam irradiation), Far West Technology (Santa Ana, CA; FWT-60 radiochromic film dosimeters), and National Physical Laboratory (Middlesex, UK; alanine dosimeters). Note: Company names are as represented at the time of study collaboration.