The Effect of Vaporous Phase Hydrogen Peroxide on Sterility Test Devices

Sterility testing can be performed using one of two methods, direct inoculation or membrane filtration.(1,2) With direct inoculation, the contents of a specified number of final product containers are transferred to a container of presterilized nutritive growth medium, which is subsequently incubated and examined for the presence of growth. With membrane filtration, the contents of a specified number of final product containers are filtered through a microporous membrane, which is then transferred to a container of presterilized growth medium. Membrane filtration is the preferred transfer method of the US and European Pharmacopoeias (USP and EP, respectively) and can either be an open funnel or a closed, disposable system. The closed sterility test system significantly reduces aseptic manipulations and false positive test results.

Sterility test devices

Currently, there is only one commercially available presterilized, closed, disposable sterility testing device, the Steritest (Millipore Corporation, Billerica, Massachusetts, USA). This device is a presterilized kit consisting of a needle adaptor connecting two incubation canisters via double-lumen PVC tubing. The incubation chamber has a vent filter to facilitate gas exchange and a 0.45 µm membrane filter is sealed in the base for sample filtration.

After sample filtration, the membrane can be rinsed and the nutritive growth medium can be transferred directly into each canister for incubation. The kit is packaged in a PVC blister sealed with a Tyvek (DuPont, Wilmington, Delaware, USA) sheet.

To further reduce the chances of false positives, the sterility test may be performed in a sterility test isolator. Sterility test devices can be staged in a transfer isolator for decontamination of the exterior packaging before transfer to the sterility test isolator. The transfer and sterility test isolators are decontaminated with a sterilant gas or vapour before use. In some cases, sterility test devices may require exposure to a further decontamination cycle, so devices may be treated with two cycles of sterilant gas.

Table I: Measurable test responses.

With proper validation, supplies, including the sterility test devices, may be stored in the isolator for up to 6 months. The HEPA filters on isolators require servicing, usually at 6 month intervals. To service the HEPA filters, the isolator integrity must be breached and all supplies removed from the isolator.

Because the sterility test device has gas-permeable packaging, residual hydrogen peroxide may be present on product contact surfaces. Currently, there are no published data on the effect of vaporous phase hydrogen peroxide (VPHP) on the performance of sterility test devices.

Figure 1: Sterility test device test variables.

Aim

The purpose of this study was to determine the following:

• Does exposure to two cycles of VPHP affect the performance of the sterility testing device?

• If there is an effect, are different microporous membranes such as polyvinylidene fluoride (PVDF) and mixed cellulose esters (MEC) affected differently?

• Is device performance affected with time after exposure to VPHP?

• Does the presence of an impermeable barrier (aluminium tape) affect performance?

Table I outlines the test responses selected to answer the questions posed in this study.

Experimental design

The study was separated into two phases. In Phase 1, a full factorial, three-factor (two-level) balanced block design was executed. The test response measured was the bubble point test. The test design for Phase 2 was established based on the results of Phase 1. The test responses measured in Phase 2 were bacteriostasis/fungistasis and gravimetric non-volatile residue (NVR). In both phases, sterility test devices were tested at Tinitial (the time of initial treatment with VPHP) and T7 months (7 months after treatment with VPHP).

Figure 2: The Steritest sterility test device.

All sterility test devices were presterilized by ethylene oxide. Untreated devices refer to those presterilized units that were not exposed to VPHP; treated devices were those presterilized units that were exposed to VPHP.

The following assumptions were made for the Phase 1 study:

• there would be variability from VPHP load to VPHP load; thus, all factors and levels were accommodated within the test block

• there would be a one-to-one comparison with the control (untreated) samples

• the evaluation of variation across lots was valid with two different lots per sterility test device type.

Figure 1 provides a schematic of the Phase 1 design. Phase 2 (Figure 2) also assumed there would be variability from VPHP load to VPHP load.

Materials and equipment

Materials. The sterility testing device for liquids in ampoules or collapsible bags with mixed esters of cellulose membrane (catalogue number TTHALA210, Steritest MEC) and the sterility testing device for solvents, creams, ointments and veterinary injectables with polyvinylidene fluoride membrane (catalogue number TLHVSL210, Steritest PVDF) were supplied by Millipore. Bacillus subtilis ATCC 6633; Candida albicans ATCC 10231; Clostridium sporogenes ATCC 11437; Staphylococcus aureus ATCC 6538P; and Staphylococcus epidermidis ATCC 12228 were purchased from American Type Culture Collection (Manassas, Virginia, USA).

Table II: Experimental design and sample parameters for Phase 1.

Equipment. VPHP cycles were performed using a VHP V1000 generator and Vaprox 31% hydrogen peroxide (H2O2) sterilant (Steris Corporation, Mentor, Ohio, USA). Integrity tests were performed using the Integritest II automatic integrity tester (catalogue number XEIT11011, Millipore). Gravimetric NVRs were evaluated by Fourier transform infrared spectroscopy (FT-IR) using a Nicolet 5SXC spectrophotometer (Thermo Nicolet Corporation, Madison, Wisconsin, USA) and a Spectratech IR-Plan microscope (Spectratech International, Inc., Middleway, West Virginia, USA). VPHP residuals were tested at point of use with short-term detector tubes (catalogue number 8101041) for an H2O2 detection level of 0.1-3 ppm (20 strokes) and Accuro Bellows Pump (catalogue number 6400000), both purchased from Drager Safety, Inc. (Pittsburgh, Pennsylvania, USA).

Table III: VPHP cycle parameters.

Experimental

VPHP cycle treatment Phase 1 study The presterilized sterility test devices were loaded into the test isolator according to Table II. The number of devices treated at any one time was limited by the maximum isolator load size. The sterility test devices were processed through two 90 minute VPHP cycles according to the parameters in Table III.

VPHP cycle treatment Phase 2 study

To ensure that the presterilized sterility test devices from the replicates were not exposed to the same cycle, the three bacteriostasis/fungistasis replicates were divided into four VPHP runs (Table IV). Each grouping, for example, Replicate 1, MEC, VPHP Run #1, consisted of five testing device canister sets. One set was selected from each lot, together with one negative control set. Each run contained four such groupings. The sterility test devices were processed through two 90 minute VPHP cycles using the same cycle parameters used in the Phase 1 study (Table III).

Table IV: VPHP cycle run plan.

Bubble point test

The bubble point method for determining membrane integrity has previously been described.3,4 In summary, sterility test devices treated with VPHP and untreated controls were tested at Tinitial and T7 months post-VPHP treatment. Tinitial and T7 months samples were tested within one week of treatment. The week's lag in testing for the bubble point represents the time required to ship treated samples across the US.

Figure 3: Tree diagram of bacteriostasis/fungistasis test plan.

Samples were wetted with water before they were tested with the automatic integrity tester. The acceptance criterion was to meet the manufacturer's specifications for each of the devices tested.

Bacteriostasis/fungistasis tests

The bacteriostasis/fungistasis test is discussed in detail in the USP and EP.1, 2 Sterility test devices were tested at Tinitial and T7 months, either post-VPHP treatment or untreated. Having been treated with VPHP and tested for bacteriostasis/fungistasis at the same facility, these devices were not subject to a shipping lag time.

Figure 4 and Figure 5.

The rinsing scheme was as specified in the USP and EP. In summary, each sterility test canister was treated with two 100 mL rinses of 0.1% (w/v) peptamin (peptic digest of animal tissue) in water, followed by the micro-organism spike. The peptamin was produced by Difco Laboratories (Sparks, Maryland, USA). In the micro-organism spike, pure cultures of the specified micro-organisms were diluted to a final concentration of less than 100 cfu per sterility test canister. The micro-organism spike was delivered to each canister as a suspension in 100 mL of 0.1% (w/v) peptamin in water.

Following the micro-organism spike, 100 mL of sterile soybean casein digest broth (SCDB) was transferred to one sterility test canister and 100 mL of sterile fluid thioglycollate medium (FTM) was transferred to the second sterility test canister for each device. The sterility test devices were then aseptically sealed, separated and incubated. The SCDB canisters were incubated at 20-25 ºC for 14 days and the FTM canisters were incubated at 30-35 ºC for 14 days.

Figure 6 and Figure 7.

Gravimetric NVR test

An exhaustive extraction of both VPHP treated and untreated sterility test devices was performed at ambient temperature in ASTM Type 1 reagent-grade water for 24 h under static conditions. The extraction solvent was evaporated to dryness and the weight of the NVR was determined.

Table V.

Samples of the NVR were then evaluated using FT-IR spectroscopy. For microscope work, a sample of each residue was thinned under pressure. Each sample was separately placed on a zinc selinide crystal. The infrared spectrum was obtained on a Nicolet 5SXC spectrophotometer using Omnic software (Thermo Nicolet Corporation, Madison, Wisconsin, USA). A Spectratech IR-Plan microscope with transmission lens was used as the sampling accessory. Standard collection parameters were used.

Residual hydrogen peroxide

The concentration of hydrogen peroxide was measured in the test sample after treatment with two 90 minute VPHP cycles using a short-term detector tube for hydrogen peroxide. This device had a detection range of 0.1-3 ppm. Immediately after the completion of the second VPHP cycle, test samples were removed from the isolator. The Tyvek sheet was pierced to facilitate the entry of the detector tube into the package. The bellows pump was actuated as per the manufacturer's instructions to sample the packaging environment and the results were recorded.

Table VI.

Results and discussion

Bubble point test

All the canisters met the acceptance criteria for bubble point (Tables V and VI). The sterility testing device with PVDF was 24.1-26.3 psig (specification of >22 psig) and the sterility testing device with MEC was 31.1-35.3 psig (specification of >30 psig).

A multivariate analysis (MANOVA) was used to evaluate these data. For the variables of time (Tinitial versus T7 months), membrane type (MEC versus PVDF) and tape (tape versus no tape), the following conclusions were drawn. Neither the VPHP sterilization nor the use of tape had an effect on the bubble point values for the sterility test devices. VPHP had no effect upon the bubble point of the sterility test devices.

Table VII.

Bacteriostasis/fungistasis tests

Growth of all test micro-organisms in all sterility test canisters met the acceptance criteria for the tests being conducted and no static or cidal effects were observed. Thus, none of the variables examined had an effect upon microbial growth (Tables VII-X).

Table VIII.

The acceptance criteria for this series of tests were as follows:

• The negative controls shall demonstrate no growth as determined visually after incubation for 14 days.

• The inoculum concentration of each test micro-organism delivered to the specified sterility test device shall be less than 100 cfu as verified by the standard plate count methodology.

• The positive controls shall demonstrate growth of the specified test micro-organism in not more than 7 days.

• The inoculated test canisters shall demonstrate growth that is visually comparable with the positive control.

Gravimetric NVR test

The NVR test only provides quantitative results. To obtain qualitative results, samples of the NVR were tested using FT-IR spectroscopy. The water extraction study showed a significant difference in the amount of NVR for the VPHP treated devices with PVDF membrane compared with the other values. The FT-IR spectra of the extractables profiles of all devices suggest the presence of inorganic materials, such as salts that may be attributed to sample preparation ( Table XI and Figures 4-7). The presence of these inorganic materials had no effect on microbial growth as demonstrated by the acceptable growth response in the bacteriostasis/fungistasis tests, and no effect on the integrity of the test devices as demonstrated by the bubble point test.

Table IX.

Residual hydrogen peroxide

The residual hydrogen peroxide concentration was measured after exposure to two 90 minute VPHP cycles. Samples were tested at the end of the aeration phase and the sterility testing device with MEC was selected as the one to test. The packaging between the device with MEC membrane is the same as the packaging for the device with PVDF membrane. The results for the device in the original primary packaging (blister and Tyvek) were 3 ppm of hydrogen peroxide, whereas those for the sterility test devices in the original primary packaging and aluminium tape were 0.1 ppm of hydrogen peroxide.

The values generated in this test were at the high limits of detection (untaped devices) and the low limit of detection (taped devices).

However, there was a significant difference between the VPHP treated devices with and without the aluminium tape. These results suggest that aluminium tape is an effective barrier to VPHP penetration compared with the absence of a barrier. The presence of residual VPHP had no effect on microbial growth, demonstrated by the acceptable growth response in the bacteriostasis/fungistasis tests and no effect on the integrity of the test devices, demonstrated by the bubble point test.

Table X.

Conclusion

Phase 1 was performed to determine whether or not the VPHP sterilization process could cause physical damage to the sterility test devices. The bubble point test is the most sensitive measure of damage, whether to the membrane or to the other components of the device. Bubble point results suggest that no membrane structural changes were caused by VPHP treatment. The use of aluminium barrier tape does not impair the function of the sterility test device.

In Phase 2, microbiological and chemical evaluations were performed to determine whether or not the presence of residuals of the VPHP decontamination process would affect microbial growth. Bacteriostasis/fungistasis results show no evidence of inhibitory residuals from VPHP treatment, thus no false negatives. A decrease in residual hydrogen peroxide was observed when the Tyvek portion of the primary packaging was covered with aluminium tape. Any residuals present in the devices tested in this study had no effect on bubble point or bacteriostasis/fungistasis. If cidal or static effects are observed, the use of impermeable barriers to the sterilant gas may be of value. If such a barrier is used, it must also be validated to ensure that it does not cause cidal or static effects.

Table XI: Non-volatile residue gravimetric extractables results.

These facts allow us to conclude that the function of the devices evaluated in this study was not impaired by two 90 minute VPHP cycles. These results were independent of the materials of construction of either sterility test device.

When performing sterility tests for the release of pharmaceutical or biopharmaceutical preparations in isolators, it is important to assess the effect of the isolator decontamination.5 Each isolator is unique in its intended use, design, load configuration and sterilization cycle. The results of this study are valid only for the isolator, sterilant gas, decontamination cycle and load configuration used in this study. It is imperative that the items used in an isolator for sterility testing be studied for potential cidal or static effects from the decontamination procedure. Any use of impermeable barriers should be documented as part of the unique decontamination cycle validation of each isolator.