Decontamination assessment of Bacillus anthracis, Bacillus
subtilis, and Geobacillus stearothermophilus spores on indoor
surfaces using a hydrogen peroxide gas generator

J.V. Rogers1, C.L.K. Sabourin1, Y.W. Choi1, W.R. Richter1, D.C. Rudnicki', K.B. Riggs1,
M.L. Taylor1 and J. Chang2

1Batte!le Memorial Institute, Columbus, OH, and 2US Environmental Protection Agency, TW Alexander Drive, Research Triangle Park,
NC, USA

ABSTRACT

J.V, ROGERS, C.L.K. SABOURIN, Y.W, CHOI, W.R. RICHTER, D.C. RUDNICKI, K.B. RIGGS, M.L. TAYLOR
AND J . CHANG. 2005,

Aims: To evaluate the decontamination of Bacillus anthracis. Bacillus subtilis, and Geobacillus stearothermophilus
spores on indoor surface materials using hydrogen peroxide gas.

Methods and Results: Bacillus anthracis, B. subtilis, and G. stearothermophilus spores were dried on seven types of
indoor surfaces and exposed to >1000 ppm hydrogen peroxide gas for 20 min. Hydrogen peroxide exposure
significantly decreased viable B. anthracis, B. subtilis, and G. stearothermophilus spores on all test materials except
G. stearothermophilus on industrial carpet. Significant differences were observed when comparing the reduction in
viable spores of B. anthracis with both surrogates. Hie effectiveness of gaseous hydrogen peroxide on the growth
of biological indicators and spore strips was evaluated in parallel as a qualitative assessment of decontamination.
At 1 and 7 days postexposure, decontaminated biological indicators and spore strips exhibited no growth, while the
nondccontam mated samples displayed growth.

Conclusions: Significant differences in decontamination efficacy of hydrogen peroxide gas on porous and
nonporous surfaces were observed when comparing the mean log reduction in B. anthracis spores with B. subtilis and
G. stearothermophilus spores.

Significance and Impact of the Study: These results provide comparative information for the decontamination
of B. anthracis spores with surrogates on indoor surfaces using hydrogen peroxide gas.

Keywords: Bacillus anthracis, Bacillus subtilis, decontamination, Geobacillus stearothermophilus, hydrogen peroxide,
spores.

INTRODUCTION

Bacterial endospores can survive in the environment for an
extended period of time, and are resistant to a wide-variety
of treatments such as heal desiccation, radiation, pressure
and chemicals (Nicholson etal. 2000). This spore resistance
is tlie result of various factors such as the thick proteina-
ceous spore coat, low water content in the spore core, and

Correspondence to: James V.Rogers, PhD, Battelle Memorial Institute SOS King
Avenue, JM-7 Columbus, OH 43201, USA (e-mail: rogenfv@battelk.org).

the a//?-type small, acid-soluble spore proteins (Setlow and
Setlow 1993; Nicholson etal 2000; Setlow etal. 2000). As
potential bacterial spore dccontaminants, ultraviolet light y-
irradiation, wet/dry heat, ozone, aqueous solutions and
mixtures, gels, and gases have been evaluated (Sagripanti
and Bonifacino '1996; Setlow and Setlow '1996; Khadre and
Youscf 2001; Rabcr and McGuire 2002; Spotts Whitney
et at. 2003; Young and Setlow 2003; Young and Setlow
2004). Gaseous forms of chlorine dioxide, ethylene oxide,
formaldehyde, hydrogen peroxide, methylene bromide,
ozone, peracetic acid and propylene oxide have been used

1


-------
for the inactivation of Bacillus spores (Spotts Whitney et al.
2003). These fumigating agents could be advantageous for
large-scale decontamination of a room or building, as
fumigants are easily dispersed and can potentially penetrate
a large volume, thereby accessing all indoor surfaces.
However, the toxicity, material compatibility, exposure
time, concentration and ventilation requirements vary
among gaseous decontaminants and considerations should
be made for each of these treatments with respect to the
intended decontamination application. Gaseous decontami-
nants also have the potential to yield better results than
conventional surface cleaning with detergent sanitizers, as
observed in a hospital environment decontaminated with
hydrogen peroxide (French etal. 2004).

In aqueous or gaseous forms, hydrogen peroxide exhibits
decontamination efficacy against bacterial spores, vegetative
bacteria, viruses, amoeba and prions (Klapes and Vesley
1990; Heckert etal. 1997; Hiti etal. 2002; Fichet etal. 2004;
French etal. 2004; Johnstone/ al. 2005). Hydrogen peroxide
is considered less toxic than other fumigants such as
chlorine dioxide, ethylene oxide and formaldehyde, and
breaks down into water and oxygen. Therefore, gaseous
hydrogen peroxide has been used as a decontaminant for
treating laboratory and medical equipment, pharmaceutical
facilities, hospital rooms and animal holding rooms (Klapes
and Vesley 1990; Heckert et al. 1997; McDonnell and
Russell 1999; Krishna etal. 2000; Krause etal. 2001; Fichet
etal. 2004; French etal. 2004; Hillman 2004; Wagenaar and
Snijders 2004).

In October 2001, the release of Bacillus anthracis spores
from envelopes mailed in Trenton, NJ led to subsequent
contamination of buildings including the mail processing
and distribution centres in Washington, DC and Trenton,
NJ, the Hart Senate Office Building, and a number of other
mail handling facilities. This contamination led to extensive
remediation and clean-up efforts and increased public
awareness regarding the possibility of future bioterrorism-
related attacks. Since this release of B. anthracis spores, there
has been a growing interest in methods of detection,
sampling and decontamination of B. anthracis spores from
surfaces, rooms and buildings. A recent review evaluated the
information in published literature regarding inactivation of
B. anthracis spores in which the authors of the review
concluded that more research is needed to evaluate the
potential application of methods used in decontaminating
laboratories and food industry facilities to larger buildings,
as well as choosing the most appropriate decontamination
technology (Spotts Whitney etal. 2003). These authors also
note that much of the data available for decontamination of
Bacillus spores is based on species other than B. anthracis;
therefore, more testing of Bacillus spores should be
conducted with or in correlation to B. anthracis (Spotts
Whitney etal. 2003).

In 2002, the US Environmental Protection Agency respon-
ded to the increasing concerns about US homeland security
by establishing the National Homeland Security Research
Center and expanding its Environmental Technology Veri-
fication(ETV) program withanewBuildingDecontamination
Technology Center (http://www.epa.gov/etv). The role
of this new Center is to perform testing to verify the
performance of commercially available technologies intended
to decontaminate indoor surfaces of buildings contaminated
with biological and chemical agents. Working through the
ETV Building Decontamination Technology Center, the
purpose of this study was twofold. First, wewantedto develop
and assess a laboratory-scale approach for evaluating decon-
tamination efficacy of B. anthracis spores deposited on typical
porous and nonporous indoor surface materials using com-
mercially available technologies. Secondly, we wanted to
compare the decontamination efficacy of B. anthracis with
selected surrogates. This paper describes the decontamination
efficacies against spores of B. anthracis and two surrogates,
B. subtilis and G. stearothermophilus, obtained for the
first technology (a hydrogen peroxide gas generator) tested in
the ETV Building Decontamination Center.

MATERIALS AND METHODS
Test organisms

Spores of the virulent B. anthracis Ames strain were
prepared using a BioFlo 3000 fermentor (New Brunswick
Scientific Co., Inc., Edison, NJ, USA). A primary culture of
B. anthracis Ames was grown overnight (16-18 h at 37°C) in
nutrient broth (BD Diagnostic Systems, Sparks, MD, USA)
on an orbital shaker set at c. 150-200 rev min)1. This
primary culture was used as an inoculum for a scale-up
culture that was grown for 6-8 h in nutrient broth on an
orbital shaker set at c. 150-200 rev min)-1. The scale-up
culture was used to inoculate Leighton-Doi Broth (BD
Diagnostic Systems) in the BioFlo 3000 fermentor after
which cultures were grown in the fermentor for approx.
24 h at 37°C. Cultures exhibiting >80% refractile spores, as
determined by phase-contrast microscopy, were centrifuged
at approx. 10 000-12 OOOXg for 15-20 min at 2-8°C. The
resultant pellet was washed twice and resuspended in ice-
cold, sterile water. The suspension was heat-shocked by
incubating at 60°C for 45-60 min to kill vegetative cells, and
centrifuged and washed a minimum of two times in ice-cold,
sterile water to remove cellular debris. The spore prepar-
ation was purified by centrifuging the sample through a
gradient of ice-cold, sterile 58% Hypaque-76 (Nycomed
Amersham, Princeton, NJ, USA) at 9000 Xg for 2 h at 2-
8°C. The resultant pellet was washed and resuspended in
ice-cold, sterile water and evaluated by phase-contrast
microscopy. Preparations having >95% refractile spores

2


-------
with <5% cellular debris were enumerated, diluted to
approx. 1-0 XIO9 colony-forming units (CFU) ml)1, and
stored at 2-8°C.

Commonly used B. anthracis surrogates for decontamina-
tion testing of spores are B. subtilis and G. stearothermophilus
(Klapes and Vesley 1990; Sagripanti and Bonifacino 1996;
Rutala el al. 1998; Khadre and Yousef 2001; Melly el al.
2002a; Sigwarth and Stark 2003; Young and Setlow. 2003).
For hydrogen peroxide decontamination, differences in
resistance to hydrogen peroxide gas have been reported
between G. stearothermophilus and B. subtilis, where
G. stearothermophilus is the most resistant (Klapes and
Vesley 1990). Therefore, stock suspensions of B. subtilis
(ATCC 19659) and G. stearothermophilus (ATCC 12980)
spores were purchased from Apex Laboratories, Inc. (Apex,
NC, USA) fortius study. Samples from these stock cultures
were enumerated, diluted to approx. 1-0 X IO9 CFU ml) in
sterile water, and stored at 2-8°C until use.

Test materials

Seven materials representing porous and nonporous indoor
surfaces commonly found in buildings were used for testing.
These materials included ShawTek EcoTek 6 industrial
carpet (Shaw Industries, Inc., Cartersville. GA, USA), bare
pine wood (Kingswood Lumber, Columbus, OH, USA),
painted (latex, semi-gloss) concrete cinder block ASTM C90
(Wellnitz, Columbus, OH, USA), glass ASTM C1036
(Brooks Brothers Glass & Mirror, Columbus, OH, USA),
white formica laminate with matte finish (Solid Surface
Design, Columbus, OH, U SA), galvanized metal ductwork
(Accurate Fabrication, Columbus, OH, USA), and painted
(latex, flat) wallboard paper (United States Gypsum
Company. Chicago. IL, USA). Samples of each test material
were cut from a larger piece of the representative materials
to form 1-9 X7-5 cm coupons. Visual inspection of the
physical integrity and appearance of the test material

coupons was performed and observations recorded before
and after decontamination in order to detect any damage to
the test materials.

Decontamination procedure

All portions of this testing were performed under Biosafety
Level 3 conditions. Prior to inoculation with either
B. anthracis, B. subtilis. or G. stearothermophilus spores, the
test materials were cleaned by wiping with 70% isopropanol.
Each test coupon was laid flat in a Biological Safety Cabinet
(BSC) Class III and contaminated with c. 1-0 XIO8 spores.
For each type of test material, three coupons were used for
decontamination, three coupons were used as controls
(inoculated; not decontaminated), and two coupons were
used as blanks (not inoculated). Suspensions of the spores
were transferred to each coupon using a micropipette by
placing the suspension over the surface as small droplets
(Fig. 1). Following inoculation, the coupons were allowed to
dry overnight, undisturbed. The next day, the inoculated
coupons intended for decontamination (and one blank) were
transferred to a Plas-Labs Model 830-ABC Compact Glove
Box (Plas-Labs, Inc., Lansing, MI, USA; volume of approx.
317 1) and the coupons were placed lying flat, inoculated
surface side up on a wire rack lined with Pet-D-Fence
Screening (New York Wire Co.. Mount Wolf, NY, USA)
for support (Fig. 2).

Biological indicators (BI) containing B. subtilis (ATCC
19659) and G. Stearothermophilus (ATCC 12980) and spore
strips (SS) containing B. atrophaeus (ATCC 9372) were also
used to evaluate decontamination. The B. subtilis and
G. stearothermophilus BI consisted of approx. 1-8 X106 and
2-6 X 10" spores, respectively, on stainless steel discs sealed
in Tyvek pouches (Apex Laboratories, Inc.), and the SS
consisted of approx. 1-8 X 106 spores on filter paper strips
sealed inglassine envelopes (Raven Biological Laboratories,
Omaha, NE, USA). For B. anthracis decontamination, three

Fig. 1 Representative view of test coupons
inoculated with 100 jul droplets containing
approx. 1-0 XIO8 spores, (a) Image showing
droplets immediately after application of
spores, (b) Image of spores on test materials
following overnight drying

3


-------
HB

I

Fig. 2 Representative view of test coupons, biological indicators (BI),
and spore strip (SS) inside of the Plas-Labs Compact Glove Box

of each BI and SS were placed inside of the glove box during
each decontamination test day. For B. subtilis and G. stearo-
thermophilus decontamination, three of each respective BI
and SS were placed inside of the glove box during each
decontamination test day. Three of each BI and SS not
subjected to hydrogen peroxide were used as positive
controls.

The Claras™ C hydrogen peroxide gas generator
(Bioquell, Inc., Horsham, PA, USA) was used for the
decontamination testing. The Claras™ C unit generated
hydrogen peroxide gas by flash evaporation of a 30%
hydrogen peroxide stock solution (Sigma, St Louis, MO,
USA). Modifications were made to the glove box to
accommodate injection/exhaust ports and sensors for the
Claras™ C. The glove box was tested for leaks prior to the
initiation of each experimental decontamination ran. For
each test, a negative pressure equivalent to two inches of a
water column was generated by a vacuum pump in the glove
box and maintained for a minimum of 2 min. Following this
leak test, the decontamination cycle of the Claras™ C unit
was initiated and included conditioning (pressure and
relative humidity adjustment), gassing (hydrogen peroxide
injection), dwell (contact time) and aeration (hydrogen
peroxide gas breakdown) phases. For the purposes of testing
under the ETV program, operational parameters were
provided by Bioquell, Inc. which included cycle pressure
(20 Pa), conditioning (10 min), gassing (20 min), dwell
(20 min), hydrogen peroxide injection rate (2-0 g min-1),
hydrogen peroxide dwell rate (0-5 g min-1), and aeration
time (9999 min). The aeration time was set for 9999 min;
however, the actual aeration time was overnight (16-18 h)
and the Claras™ C was turned-off the next day when the
hydrogen peroxide concentration was at or below the
recommended 8-h exposure limit of 1 ppm (http://
www.cdc.gov/niosh). All decontamination testing was per-
formed at room temperature. Time, pressure, relative
humidity and hydrogen peroxide concentration were

monitored by specific sensors or detectors within the
Claras™ C, and recorded either electronically in real-time
using a personal computer or on a paper printout. The
Claras™ C hydrogen peroxide sensor possessed an upper
detection limit of 1000 ppm.

Sample processing and data collection

Decontamination efficacy of B. anthracis, B. subtilis and
G. stearothermophilus spores was quantified by measuring the
viable spores on both exposed and unexposed coupons.
Following exposure to hydrogen peroxide gas, each coupon
was placed in a 50 ml tube containing 10 ml of sterile
phosphate-buffered saline (PBS) to which 0-1% Triton X-
100 (Sigma) and approx. 200 ug catalase (Roche, Indiana-
polis, IN, USA) had been added. The purpose of the Triton
X-100 was to minimize clumping of spores, and the catalase
was used to neutralize any residual hydrogen peroxide. The
inoculated control (not decontaminated) and blank coupons
were also placed in a 50 ml tube containing 10 ml of sterile
PBS with 0-1% Triton X-100 and catalase. For spore
extraction, the tubes were agitated at 200 rev min1 on an
orbital shaker for 15 min at room temperature. Each tube
was then heat-shocked at 65 °C for 1 h to kill vegetative
bacteria. Following heat-shock, 1-0 ml of each extract was
removed, and a series of dilutions from 10 1 to 10 were
prepared in sterile water.

Spore viability was determined by dilution plating in
which 100 fi\ of the undiluted extract and each serial
dilution were plated onto tryptic soya agar plates (Remel,
Lexena, KS, USA) in triplicate, allowed to dry, and
incubated overnight at 37°C for B. anthracis, 35°C for
B. subtilis and 55-60°C for G. stearothermophilus. Following
18-24 h incubation, plates were enumerated and CFU ml-1
were determined by multiplying the average number of
colonies per plate by the reciprocal of the dilution. Data
were expressed as the mean ± standard deviation (SD) of
observed CFU.

Efficacy calculations and statistical analysis

To calculate the efficacy of the decontamination treatment,
the numberofviable spores extractedfromthe decontaminated
test coupons was compared with the number of viable
spores extracted from the control coupons. Efficacy for
biological agents was expressed in terms of a log reduction
using the following equation:

Log Reduction = log( Y / TV)

where N is the mean number of viable organisms recovered
from the control coupons (i.e. those not subjected to
decontamination), and Vis the number of viable organisms
recovered from each test coupon after decontamination. For

4


-------
decontaminated coupons where viable organisms were not
detected, the efficacy was calculated as the log of the mean
number of viable organisms recovered from the control
coupons. Using the calculated log reduction for each test
coupon, the mean (±SD) log reduction was calculated.
Mean (±SD) percent recovery was calculated for each type
of test material inoculated with each biological agent or
surrogate by dividing the number of biological organisms in
the treated coupon by the number of biological organisms in
the nondeconlaminated controls.

For statistical comparisons, the two-way ANOVAand /-tests
(SAS version 8-2; SAS Institute, Inc., Cary, NC, USA) were
used for data analysis. For each material and species
combination, log reduction was calculated as described
above. The two-way ANOVA was used to assess main effects
for each organism and test material and interactions were
fitted to the log reduction data. This model was used to
compare the mean log reduction for each bacterial species
tested, and compare tlie log reduction in B. sitblilis and
G. stearothermophilus spores to B. anthracis spores for each
test material. The /-tests or statistical contrasts were used for
the comparisons, with no adjustment for multiple compar-
isons. The ANOVA model was fitted using the SAS GLM
procedure. P < 0-05 was used as the level for significance.

RESULTS

A total of nine decontamination runs were conducted, which
included three tests for each test organism (including BI
and SS) on all test materials. The conditioning, gassing
and dwell phases were initiated and completed within 2 h.
In all tests, the hydrogen peroxide concentration inside the
glove box was maintained at or above the vendor-recom-
mended 1000 ppm during the dwell phase. Relative humid-
ity reached levels of 60-80% during the dwell phase, in
which condensation was prevalent on all surfaces within the
exposure chamber. The observed mean cycle pressure was
observed to be approx. 20 Pa throughout the entire
decontamination cycle. Following all experimental decon-
tamination runs, the test coupons were evaluated qualita-
tively for visible surface damage and no changes to any of
the test materials were observed.

Exposure of test coupons contaminated with B. anthracis
Ames, B. suhtilis, or G. slearolhermophilus spores to hydrogen
peroxide gas resulted in a reduction of viable spores that
varied according to the type of the test material
(Tables 1-3). With respect to the inoculated surface, three
of these test materials (industrial carpet, bare pine wood,
painted concrete) can be considered porous, while the other
four test materials (glass, decorative laminate, galvanized
metal ductwork, painted wallboard paper) can be considered
nonporous. The mean log reduction of detectable viable
B. anthracis Ames spores ranged from 3-0 to 7-9 for all seven

T able 1 Decontamination efficacy of Bacillus anthracis Ames spores
following hydrogen peroxide exposure*

lai /

Recov

l .og

reduction

tarpc-f

6-9 ± 0-32 x H)' DO ± 2-8

V\

8-3 ± 7-2 x 10+ 0-081 ± 0-063 3-0 ± 2-lt

¦rt; ± 14 x I if	9 ± j-3

3-3 ± 2-9 x I if	<0-0!

f-	j

3-8 ± I-7 x JO7	33 ± :S

xtted I-3 ± 2-6 x 10'"'	<0-01

(

S4 ± 2-2 x Mf	75 ± 20

0

Decorative laminate

7-0	± 1-0 x 107	61 ± 8-7
xitc-d 0	0

"-•¦-"'ai ductwork

3'5 ± 0-13 x 10'	31 ± 1 -1

0	0
1- re, paper

8-3	± 0-63 x 10°	7-7 ± 0-39
cited 0	0

\/\

3-7 ± 0-67+
V\

6-4 ±2-1 +
\/\

>7-9 Of

o-i-

\A

>7-5 ± 0+

07

s arc expressed as mean + SO men triplicate samples of each test
sal.

Androgen peroslde
i the Materials atx!

<-	 'he

calculation were Iwd on the number ot detectable viable spores it! the
t VIear: log	than zero (/•* St p-03)

test materials (Table 1). For all seven test materials, the log
reduction of detectable viable B. subtilis and G. stearother-
mophilus spores ranged from 1-6 to 7-7 and 0-81-6-0
respectively (Tables 2-3). No viable organisms were detec-
ted in any of the blank samples.

Statistical analysis of the data revealed that all mean
log reductions were significantly different from zero
(Tables 1-3) with the exception of G. stearothermophilus on
industrial carpet, indicating that exposure to hydrogen
peroxide gas significantly reduced (P < 0-05) the mean
number of B. anthracis, B. subtilis and G. stearothermophilus
spores on all but one of the indoor surface materials.
Comparisons within each material indicated that the two
selected surrogates had either similar or lower mean log
reductions than B. anthracis. The mean log reduction in
B. subtilis spores was significantly lower (P < 0-05) than
B. anthracis spores for industrial carpet and bare pine wood.

5


-------
Table 2 Decontamination efficacy oi Bacillus subtilis (ATCC 19659)
spores following hydrogen peroxide exposure*

Table 3 Decontamination efficacy of Geobacillus stearothermophilus
(ATCC 12980) spores following hydrogen peroxide exposure*

Test material/
treatment

Total spores
recovered (CFU)

Log

% Recovery reduction

Industrial carpet
None

h2o2

Bare pine wood
None

h2o2

Painted concrete
None

h2o2

Glass
None

h2o2

Decorative laminate
None

h2o2

Galvanized metal ductwork

None	3-6 ± 0-76 x 107

H202	3-3 ± 3-3 x 10

Painted wallboard paper

None	3-3 ± 2-5 x 107

h2o2	0

4-7 ± 0-19 x 10'
1-2 ± 0-42 x 106

8-8 ± 2-2 x 105
8-1 ± 6-1 x 103

1-3	± 0-16 x 10'

2-2	± 1-9 x 10

3-7	± 2-0 x 107
0

4-6 ± 0-85 x 10'
0

51 ± 2-0
1-2 ± 0-45

1-0 ± 0-24
<0-01

14 ± 1-7
<0-0001

41 ± 23
0

51 ± 9-5
0

44 ± 9-3
<0-0001

41 ± 31
0

NA

1-6	± 0-15f

NA

2-2	± 0-50f

NA

6-1 ± 0-88f

NA

>7-6 ± Of
NA

>7-7 ± Of
NA

6-4 ± 0-98f

NA

>7-5 ± Of

Values are expressed as mean ± SD from triplicate samples of each test
material.

NA, not applicable.

*Bacillus subtilis spores were subjected to hydrogen peroxide gas

exposure and assessed for viability as described in the Materials and

Methods. Each test material was inoculated with approx.

1-0 X 108 CFU and dried overnight. Spores were extracted from the

test materials and enumerated. Percent recovery and log reduction

calculation were based on the number of detectable viable spores in the

control and hydrogen peroxide-treated samples.

|Mean log reduction is significantly different than zero (P < 0-05).

Test material/
treatment

Total spores
recovered (CFU)

% Recovery

Industrial carpet
None

h2o2

Bare pine wood
None

h2o2

Painted concrete
None

h2o2

Glass
None

h2o2

2-7 ± 0-051 x 10'
4-3 ± 1-1 x 106

2-8	± 0-081 x 106

3-0	± 2-0 x 102

9-4 ± 1-1 x 106
2-9 ± 4-1 x 103

8-7 ± 0-58 x 106
2-5 ± 2-0 x 102

Decorative laminate

None	5-9 ± 1-1 x 106

H2Oz	1-3 ± 2-1 x 104

Galvanized metal ductwork

None	1-5 ± 0-37 x 107

H2Oz	1-6 ± 0-27 x 105

Painted wallboard paper

None	9-7 ± 0-81 x 106

H2Oz	2-2 ± 1-9 x 10

21 ± 0-40
3-3 ± 0-84

2-2 ± 0-063
<0-001

7-8	± 0-87
<0-01

6-8 ± 0-45
<0-001

6-2 ± 1-2
0-013 ± 0-023

12 ± 2-9
0-13 ± 0-021

8-1	± 0-68
<0-0001

Log

reduction

NA

0-81 ± 0-10

NA

4-1 ± 0-46f

NA

4-1 ± l-0f

NA

4-7 ± 0-42f

NA

3-8 ± l-4f

NA

2-0 ± 0-07f

NA

6-0 ± 0-88f

Values are expressed as mean ± SD from triplicate samples of each test
material.

NA, not applicable.

*Geobacillus stearothermophilus spores were subjected to hydrogen
peroxide gas exposure and assessed for viability as described in the
Materials and Methods. Each test material was inoculated with approx.
1-0 X 108 CFU and dried overnight. Spores were extracted from the
test materials and enumerated. Percent recovery and log reduction
calculation were based on the number of detectable viable spores in the
control and hydrogen peroxide-treated samples.
fMean log reduction is significantly different than zero (P < 0'05).

while the mean log reduction in G. stearothermophilus spores
was significantly lower (P < 0-05) than B. anthracis for
industrial carpet, painted concrete, glass, decorative lamin-
ate and galvanized metal ductwork (Fig. 3).

For all BI and SS exposed to hydrogen peroxide, no
growth was observed as determined by the lack of visibly
cloudy liquid cultures at 1 and 7 days postexposure. When
not exposed to hydrogen peroxide, all of the BI and SS
exhibited growth as determined by the presence of visibly
cloudy liquid cultures on both 1 and 7 days.

DISCUSSION

The results of this study show that a 20-min treatment with
>1000 ppm hydrogen peroxide gas resulted in varying
decontamination efficacy against B. anthracis, B. subtilis
and G. stearothermophilus spores dried on common indoor

porous and nonporous materials. In the present study, the
spore deposition from drying of an aqueous suspension on a
material surface is a different delivery mechanism than what
occurred during the intentional B. anthracis release in
Washington, DC and Trenton, NJ. The Bacillus anthracis
spores that were mailed in envelopes ranged from individual
particles to microscopic aggregates and had the consistency
of a fine powder. Therefore, it is possible that the spore
preparation, mode of deposition, and spore adherence of a
fine powder compared with an aqueous suspension could
affect the decontamination efficacy of hydrogen peroxide
gas.

The observed log reduction values for B. anthracis,
B. subtilis and G. stearothermophilus spores inoculated on
porous materials were consistently lower than on nonporous
materials, suggesting that porosity affects decontamination
efficacy. Hydrogen peroxide decontamination of B. anthracis

6


-------
9 r

8 -

7 -

Industrial Bare pine Painted Glass Decorative Galvanized Painted
carpet wood concrete	laminate metal wallboard

ductwork paper

Fig. 3 Mean log reduction values for
B. anthracis (¦), B. subtilis (~) and
G. stearothermophilus (El) spores on all test
materials exposed to hydrogen peroxide gas.

Statistical comparisons were made between
B. anthracis and each of the two surrogates on
all test materials as described in the Materials
and Methods. Asterisks indicate that the mean
value is significantly different (P < 0*05) than
the mean value for B. anthracis

spores inoculated on nonporous surfaces resulted in
>6-0 log reduction. The log reduction in viable B. anthracis
spores was less (3-0 and 3-7) on the industrial carpet and
bare pine wood compared with the other five test materials.
For nonporous materials with relatively smooth and hard
surfaces, the inoculated spores should remain predomin-
antly at the material surface, enabling interaction of the
gaseous hydrogen peroxide and spores with limited
obstruction. However, when materials possessing surfaces
such as the porous weave of carpet or open-grain of bare
pine wood are inoculated with B. anthracis, the spores may
penetrate into the material and embed in pits and cavities.
Such penetration and embedding of spores into the test
materials could preclude the interaction of hydrogen
peroxide gas with the spores, thereby decreasing the
potential for spore inactivation. This may, in part, be the
result of the fact that hydrogen peroxide gas must penetrate
through carpet fibre or into the wood grain before it can
interact with B. anthracis spores. Similar to our observa-
tions, disinfection of Paenibacillus lar\>ae subsp. lar\>ae
spores in wood using chemical sterilants was not effective
because of a limited penetration capacity of the chemicals
into the wood (Dobbelaere et al. 2001). It is also possible
that some of the hydrogen peroxide could decompose
through chemical interactions within the carpet fibre or bare
pine wood matrices before it reaches the embedded spores.
The penetration of bacterial spores into the porous materials
is plausible; however, further work evaluating spore depos-
ition, chemical interactions of hydrogen peroxide with test
matrices, or possible chemical or physical interactions
between the test materials and components of the spore
coat will be required.

Most of the data available in the scientific literature on
Bacillus species spore decontamination utilizes avirulent

Test material

strains of B. anthracis or surrogates. An aim of this study
was to compare the decontamination efficacy of the virulent
B. anthracis Ames strain spores with that of B. subtilis and
G. stearothermophilus spores. With respect to virulent
B. anthracis spores, a surrogate should simulate the
behaviour or responses and result in comparable perform-
ance data to hydrogen peroxide gas under controlled
conditions. In the present study, B. subtilis exhibited
nonsignificant differences in log reductions to that of B.
anthracis for five of the seven materials tested. For the two
materials (industrial carpet and bare pine wood) that
exhibited statistical differences between B. subtilis and B.
anthracis, the mean log reduction values for B. subtilis were
lower than the mean log reduction values for B. anthracis.
When comparing the mean log reductions between B.
anthracis and G. stearothermophilus spores, nonsignificant
differences were observed for only two of the seven
materials tested. However, for the industrial carpet, painted
concrete, glass, decorative laminate, and galvanized metal
ductwork materials, the mean log reduction in G. stearo-
thermophilus spores was significantly lower than B. anthracis.
Therefore, statistical comparisons of the log reductions for
all three organisms show that as a surrogate, B. subtilis
spores appear to reflect B. anthracis spore resistance to
hydrogen peroxide gas more closely than G. stearothermo-
philus spores. It is possible that the G. stearothermophilus
spores are a surrogate that could result in relatively
conservative results as these spores appear significantly
more resistant to hydrogen peroxide gas decontamination
than B. anthracis spores on most surfaces tested in this
study. Our results are consistent with previous work where
G. stearothermophilus spores were reported to be more
resistant to being killed by hydrogen peroxide gas compared
with B. subtilis spores (Klapes and Vesley 1990).

7


-------
For the inactivation of Bacillus species spores, consider-
ations must be made for the length of exposure time,
temperature, decontaminant concentration, pH and relative
humidity (Spotts Whitney el al. 2003). In the current study,
only one combination of the operational parameters for the
Claras™ C unit was utilized. It should be noted that
the Claras™ C hydrogen peroxide sensor measured the
circulating hydrogen peroxide concentration inside the glove
box and not at the surface of the test materials. Therefore, it
is possible that the hydrogen peroxide concentration at the
material surface could be different than that measured in the
glove box air, which could potentially affect spore killing.
The 20-min dwell (contact) time promoted decreases in
viable spores for all three organisms on all test materials
except G. stearothermophilus on industrial carpet. The
observed log reductions in viable B. anthracis and B. subtilis
spores were >6-0 on painted concrete, glass, decorative
laminate, galvanized metal ductwork and painted wallboard
paper. A longer dwell time, such as 60 or 120 min, should
result in higher decontamination efficacy of B. anthracis and
surrogate spores on porous and nonporous surfaces. Such
time-dependent results for spore killing have been observed
using hydrogen peroxide and peroxide-based decontami-
nants (Klapes and Vesley 1990; Melly etal. 2002a; Young
and Setlow 2004). The relative humidity reached 60-80%
inside of the glove box during the dwell phase of testing and
was concomitant with the formation of condensation (fog-
ging) along the walls inside of the glove box. Similar micro-
condensation has been observed during decontamination of
an enclosed B SC in which hydrogen peroxide concentrations
exceeded 700 ppm (Hillman 2004). Condensation may
enhance the decontaminating effects of hydrogen peroxide
(Watling el al. 2002), which could promote the penetration
of hydrogen peroxide into porous materials to reach
embedded spores, or provide a microenvironment for
oxidative spore injury. Further investigation is needed to
determine whether the formation of this condensation layer
is an important aspect for the decontamination of B.
anthracis spores on porous materials.

The growth assessments of various BI are often used as a
qualitative evaluation of decontamination performance
(Heckert el al. 1997; Sigwarth and Moirandat 2000; Sig-
warth and Stark 2003; French et al. 2004; Hillman 2004;
Johnston et al. 2005). In the present study, we employed
this type of growth assessment and evaluation in which the
nondecontaminated control BI and SS displayed growth in
the liquid cultures at both 1 and 7 days. However, when the
BI and SS were exposed to hydrogen peroxide, no growth
was observed in the liquid cultures at 1 and 7 days. These
results suggest that exposure to >1000 ppm hydrogen
peroxide for 20 min inactivated both the BI and SS, all of
which contain spore loads of approx. 1-8-2-6 • 106 spores.
Using these results, a >6-2 log reduction can be calculated

based on spore density for the BI and SS. However,
<6-0 log reduction was obtained for B. subtilis on industrial
carpet and bare pine wood, while none of the test materials
inoculated with G. stearothermophilus exhibited a greater
than 6-2 log reduction. When compared with the test
coupons used in this study, the observed complete inacti-
vation of purchased BI and SS may reflect differences in
inoculation techniques, spore load/distribution and varia-
tions in carrier materials of the BI and SS. The spore load
on the purchased BI and SS was much less than on the
inoculated material coupons prepared in this study, sug-
gesting that the spores may be more evenly distributed with
less clumping on BI and SS than on test materials and this
even distribution may be a factor in facilitating effective
decontamination. Differences in spore carrier materials can
affect the performance of BI (Shintani and Akers 2000;
Johnston et al. 2005), which may also play a role in
hydrogen peroxide gas-induced decontamination. It is
possible that when using BI and SS as qualitative indicators
of decontamination performance, these indicators may
overpredict the log reduction of B. anthracis spores achieved
on porous surfaces.

Spore production conditions and heat treatment of the
spores at 65°C could potentially contribute to the spore
killing observed in this study. Differences in spore produc-
tion methods, including growth medium and temperature,
have been shown to influence spore resistance (Palop et al.
1999; Cazemier et al. 2001; Melly et al. 2002b). In the
present study, the B. subtilis and G. stearothermophilus spores
were purchased from a commercial source where the growth
conditions may have been different from those used in
producing the B. anthracis Ames spores. Such differences
could affect the relative resistance of B. subtilis and G.
stearothermophilus spores to hydrogen peroxide treatment
when compared with B. anthracis spores. Bacillus subtilis
spores pretreated with any of eight different oxidizing agents
(including hydrogen peroxide) were more sensitive to killing
by subsequent incubation at 84°C (Cortezzo etal. 2004). In
this study, it is possible that viable spores extracted from the
material coupons exposed to hydrogen peroxide gas could be
killed during the 65 °C heat shock. However, further
experimentation is needed to determine the effect of the
65°C incubation step decreases the viability of spores
damaged by the hydrogen peroxide treatment.

This study demonstrates the decontamination efficacy of
hydrogen peroxide gas on spores from B. anthracis Ames
and the two simulants B. subtilis and G. stearothermophilus.
The capacity for decontamination appears to be influenced
in part by the porosity of the contaminated surfaces and
differences in decontamination efficacy are prevalent
between the biological warfare agent B. anthracis and
surrogates. The hydrogen peroxide gas exposure demon-
strated significant efficacy in spore decontamination on the

8


-------
seven types of surface materials evaluated; however, future
studies will be needed to determine an acceptable reduction
in spores (e.g. 6 logs) that would enable safe re-entry back
into a building that has been decontaminated. The results of
this study support the need for more building decontam-
ination technology testing against Bacillus spores conducted
with or in correlation to B. anthracis spores.

ACKNOWLEDGEMENTS

This work was funded by the National Homeland Security
Research Center of the US Environmental Protection
Agency as part of the Environmental Technology Verification
(ETV) program. The Clarus™ C unit was kindly
provided by Bioquell, Inc. for this testing. The authors
thank Rick Tuttle, Nancy Niemuth, Claire Matthews,
Ed Heller and Nicole Caudill for their excellent assistance.

REFERENCES

Cazemier, A.E., Wagenaars, S.F.M. and ter Steeg, P.F. (2001) Effect
of sporulation and recovery medium on the heat resistance and
amount of injury of spores from spoilage bacilli. J Appl Microbiol 90,
761-770.

Cortezzo, D.E., Koziol-Dube, K., Setlow, B. and Setlow, P. (2004)
Treatment with oxidizing agents damages the inner membrane of
spores of Bacillus subtilis and sensitizes spores to subsequent stress.
JApplMicrobiol97, 838-852.

Dobbelaere, W., de Graaf, D.C., Reybroeck, W., Desmedt, E., Peeters,
J.E. and Jacobs, F.J. (2001) Disinfection of wooden structures
contaminated with Paenibacillus larvae subsp. larvae spores. J Appl
Microbiol 91, 212-216.

Fichet, G., Comoy, E., Duval, C., Antloga, K., Dehen, C., Charbon-
nier, A., McDonnell, G., Brown, P. et al. (2004) Novel methods
for disinfection of prion-contaminated medical devices. Lancet 364,
521-526.

French, G.L., Otter, J.A., Shannon, K.P., Adams, N.M.T., Watling,
D. and Parks, M.J. (2004) Tackling contamination of the hospital
environment by methicillin-resistant Staphylococcus aureus
(MRSA): a comparison between conventional terminal cleaning
and hydrogen peroxide vapour decontamination. J Hosp Infect 57,
31-37.

Heckert, R.A., Best, M., Jordan, L.T., Dulac, G.C., Eddington, D.L.
and Sterritt, W.G. (1997) Efficacy of vaporized hydrogen peroxide
against exotic animal viruses. Appl Environ Microbiol 63, 3916-3918.
Hillman, D. (2004) Vapor phase hydrogen peroxide gas decontamina-
tion of a BSC. Perform Rev 10, 10-11.

Hiti, K., Walochnik, J., Haller-Schober, E.M., Faschinger, C. and
Aspock, H. (2002) Viability of Acanthamoeba after exposure to a
multipurpose disinfecting contact lens solution and two hydrogen
peroxide systems. Br J Ophthalmol 86, 144-146.

Johnston, M.D., Lawson, S. and Otter, J.A. (2005) Evaluation of
hydrogen peroxide vapour as a method for the decontamination of
surfaces contaminated with Clostridium botulinum spores. J Microbiol
Methods 60, 403-411.

Khadre, M.A. and Yousef, A.E. (2001) Sporicidal action of ozone and
hydrogen peroxide: a comparative study. Int J Food Microbiol 71,
131-138.

Klapes, N.A. and Vesley, D. (1990) Vapor-phase hydrogen peroxide as
a surface decontaminant and sterilant. Appl Environ Microbiol 56,
503-506.

Krause, J., McDonnell, G. and Riedesel, H. (2001) Biodecontamina-
tion of animal rooms and heat-sensitive equipment with vaporized
hydrogen peroxide. Contemp Top Lab Anim Sci 40, 18-21.

Krishna, A.K., Lodhi, S.A. and Harris, M.R. (2000) Isolation
technology for research and development applications: from concept
to production. Pharm Dev Technol 5, 507-520.

Melly, E., Cowan, A.E. and Setlow, P. (2002a) Studies on the
mechanism of killing B. subtilis spores by hydrogen peroxide. J Appl
Microbials, 316-325.

Melly, E., Genest, P.C., Gilmore, M.E., Little, S., Popham, D.L.,
Driks, A. and Setlow, P. (2002b) Analysis of the properties of spores
of Bacillus subtilis prepared at differ enttemperature. J ApplMicrobiol
92, 1105-1115.

Nicholson, W.L., Munakata, N., Horneck, G., Melosh, H.J. and
Setlow, P. (2000) Resistance of Bacillus endospores to extreme
terrestrial and extraterrestrial environments. Microbiol Mol Biol Rev
64, 548-572.

Palop, A., Manas, P. and Condon, S. (1999) Sporulation temperature
and heat resistance of Bacillus subtilis spores: a review. J Food Saf
19, 57-72.

Raber, E. and McGuire, R. (2002) Oxidative decontamination of
chemical and biological warfare agents using L-Gel .J Hazard Mater
93,339-352.

Rutala, W.A., Gergen, M.F. and Weber, D.J. (1998) Comparative
evaluation of the sporicidal activity of new low-temperature
sterilization technologies: ethylene oxide, 2 plasma sterilization
systems, and liquid peracetic acid. Am J Infect Control 26, 393-398.

Sagripanti, J.-L. and Bonifacino, A. (1996) Comparative sporicidal
effects of liquid chemical agents. Appl Environ Microbiol 62, 545-
551.

Setlow, B. and Setlow, P. (1993) Binding of small, acid-soluble spore
proteins to DNA plays a significant role in the resistance of Bacillus
subtilis spores to hydrogen peroxide. Appl Environ Microbiol 59,
3418-3423.

Setlow, B. and Setlow, P. (1996) Role of DNA repair in Bacillus subtilis
spore resistance. JBacteriol 178, 3486-3495.

Setlow, B., McGinnis, K.A., Ragkousi, K. and Setlow, P. (2000)
Effects of major spore-specific DNA binding proteins on Bacillus
subtilis sporulation and spore properties. JBacteriol 182,6906-6912.

Shintani, H. and Akers, J.E. (2000) On the cause of performance
variation of biological indicators used for sterility assurance. PDA J
Pharm Sci Technol 54, 332-342.

Sigwarth, V. andMoirandat, C. (2000) Development and quantification
of H202 decontamination cycles. PDA J Pharm Sci Technol 54,
286-304.

Sigwarth, V. and Stark, A. (2003) Effect of carrier materials on the
resistance of spores of Bacillus stearothermophilus to gaseous hydro-
gen peroxide. PDA J Pharm Sci Technol 57, 3—11.

Spotts Whitney, E.A., Beatty, M.E., Taylor, T.H., Weyant, R., Sobel,
J., Arduino, M.J. and Ashford, D.A. (2003) Inactivation of Bacillus
anthracis spores. Emerg Infect Dis 9, 623-627.

9


-------
Wagenaar, C.L. and Snijders, J.M.A. (2004) Decontamination of
broilers with hydrogen peroxide stabilized with glycerol during
processing. Int J Food Microbiol 91, 205-208.

Watling, D., Ryle, C. and Christopher, M. (2002) Theoretical analysis
of the condensation of hydrogen peroxide gas water vapour as
used in surface decontamination. PDA J Pharm Sci Technol 56,
291-299.

Young, S.B. and Setlow, P. (2003) Mechanisms of killing of Bacillus
subtilis spores by hypochlorite and chlorine dioxide. JAppl Microbiol
95, 54-67.

Young, S.B. and Setlow, P. (2004) Mechanisms of killing Bacillus
subtilis spores by Decon and Oxone™, two general decontaminants
for biological agents. J Appl Microbiol 96, 289-301.

10


-------