Use of a Mixture of Surrogates for Infectious Bioagents in a Standard Approach to Assessing Disinfection of Environmental Surfaces

Disinfectant testing with a single type of organism does not represent field conditions, where bioagents or other pathogens may be mixed with other contaminants. Such an approach also cannot predict the true spectrum of microbicidal activity of a given chemical, while the identity of the target pathogen(s) is often unknown. We used a mixture of Acinetobacter baumannii, Mycobacterium terrae (15), hepatitis A virus (HAV) (4), and the spores of Geobacillus stearothermophilus as surrogates for infectious bioagents, with an added soil load on disks (1 cm in diameter; 0.75 mm thick) of brushed stainless steel (AISI no. 430; Muzeen & Blythe, Winnipeg, MB, Canada), to better simulate environmental surface disinfection (1, 11). Table 1 gives details on the microbial strains, media used for their culture and recovery, and methods for preparing working stocks. The quantitative carrier test (QCT) method, ASTM standard E-2197 (1), was used to test the organisms singly and in a mixture. Each 200 μl of the inoculum contained 34 μl each of the four organisms, 40 μl of bovine mucin, 14 μl of yeast extract, and 10 μl of bovine serum albumin stocks.

Disinfectants tested were peracetic acid (PAA; 500 and 1,000 ppm), chlorine dioxide (CD; 500 and 1,000 ppm), and domestic bleach (DB; 2,500, 3,300, and 5,000 ppm). Buffered saline (pH 7.2) was the control fluid, eluent, and diluent. Hard water (400 ppm CaCO₃) was the diluent for disinfectants (1).

Each disk received 10 μl of the inoculum, dried and covered with 50 μl of test substance, or saline at 22 ± 2°C. At the end of the contact time, each disk was eluted in a neutralizer and the eluates were assayed (1, 9, 11, 12). The neutralizer consisted of 1% dextrose (Difco), 0.7% lecithin (Alfa Aesar), 0.25% sodium bisulfite (J. T. Baker), 0.1% sodium thioglycolate (Sigma), 0.6% sodium thiosulfate (Analar), 0.2% l-cysteine (Sigma), 0.5% tryptone (Oxoid), and 0.1% Tween 80 (Bioshop) in buffered saline (pH 7.2). In each experiment, three control and three test carriers were used, and all experiments were repeated thrice. The performance criteria for the tested substances were ≥3.0 log₁₀ reduction in PFU of the virus and ≥6.0 log₁₀ reductions in the CFU for the other three organisms. When the mixture of test organisms was used, the components were separated by first passing the mixture through a membrane filter (0.22-μm pore diameter) to retain all the organisms except the virus. The filtrate was subjected to plaque assays for HAV in FRhK-4 cells. For the three bacteria, separate filters were placed on appropriate agar plates (Table 1) and incubated.

The data for 5-min contact are given in Table 2. All levels of the disinfectants tested met the criterion for M. terrae and A. baumannii when tested individually or in mixture. Only 1,000 ppm of PAA was effective against the spores. Both levels of PAA were ineffective against HAV, while the other disinfectants could reduce its titer between 3.5 and 4 log₁₀. Only 1,000 ppm of PAA could consistently meet the criterion for sporicidal activity after 10 min (data not shown). Extending the contact time to 20 min allowed both levels of PAA and DB to meet the criterion for sporicidal activity, while 500 ppm of CD failed to do so; CD at 1,000 ppm barely met the criterion when tested alone against the spores but could not do so in the mixture (Fig. 1).

The study showed the feasibility of testing liquid chemicals against a mixture of suitable surrogates for infectious bioagents. This approach allowed standardized and simultaneous assessment of the spectrum of microbicidal activities of the test formulations under identical conditions that better simulate field conditions and that can be readily adapted to test foams and gaseous chemicals on other carrier materials. The surrogates selected covered the spectrum of microbicide resistances of all currently known classes of infectious bioagents.

A. baumannii is among the more environmentally stable and microbicide-resistant vegetative bacteria known (7, 13). M. terrae represented pathogens with generally higher resistance to microbicides (3) and possibly drug-resistant Mycobacterium tuberculosis and category C agents (6). HAV, a small, nonenveloped virus known for its stability and microbicide resistance (9), represented select agents (CBW, biological weapons classification, 2001 [http://www.selectagents.gov/Select%20Agents%20and%20Toxins%20List.html ]) and also food- and waterborne pathogens listed as biothreats (2, 10). The spores of G. stearothermophilus may be more resistant to oxidizing chemicals than the spores of Bacillus anthracis (8); their thermophilic nature made them safer to handle and easy to separate from the mixtures.

The disinfectants were selected for their commercial availability and broad-spectrum and relatively rapid action (5, 14). The last criterion excluded all but oxidizers because other common active agents are limited as microbicides and/or require hours of contact for sporicidal action.

For PAA tests, the recovery of infectious HAV in the absence of any viable spores is somewhat anomalous but not surprising. While we do not believe HAV to be more resistant than bacterial spores, the small size of the virus in the dried inocula likely afforded it significant protection. Compared to HAV, the mycobacterium proved more susceptible to all the disinfectants tested. This highlights a serious weakness in the traditional rankings of disinfectant susceptibility, where mycobacteria are often considered more resistant than nonenveloped viruses (5, 14).

In the initial trials with the mixtures, the titer of A. baumannii dropped sharply; using virus pools without antibiotics resolved the issue. The ability of A. baumannii to grow on 7H11 agar and thus interfere with the recovery of M. terrae was addressed by replacing the standard strain of M. terrae with one containing a kanamycin resistance gene (15). Incorporation of enough kanamycin in 7H11 suppressed the growth of A. baumannii while allowing the mycobacterium to grow.

Using a mixture of surrogates in QCT not only proved feasible but also highlighted the need to review certain long-held concepts about the relative sensitivities of classes of pathogens to disinfectants. The details reported should allow extension of the work to CL-3 and possibly CL-4 agents to confirm that the results obtained with the carefully chosen surrogates are indeed applicable to various classes of infectious bioagents.

• Open in new tab
• Download powerpoint

FIG. 1.

Reductions of G. stearothermophilus spores by the test formulations after 20 min of contact, individually and in a mixture at 22 ± 2°C.

• Copyright © 2010 American Society for Microbiology