Applied and Environmental
`Microbiology
`
`Microbiological Aspects of
`Ethylene Oxide Sterilization :
`|. Experimental Apparatus
`and Methods
`
`K. Kereluk, R. A. Gammon and R. S.
`Lloyd
`Appl. Microbiol. 1970, 19(1):146.
`
`Updated information and services can be
`found at:
`
`http://aem.asm.org/content/19/1/146
`
`These include:
`
`CONTENT ALERTS
`
`Receive: RSS Feeds, eTOCs, free email
`alerts (when new articles cite this article),
`more»
`
`
`
`
`
`tsenfiliqZLOZ‘vzMemqeju0/6J0'ulse'ulee//:d11uuJOJ;pepl20|UllAOG
`
`
`
`
`
`
`
`Information about commercial reprint orders: http:llaem.asm.org/site/misc/reprints.xhtm|
`To subscribe to to another ASM Journal go to: http://journals.asm.orglsite/subscriptions/
`
`JournalsASMcrg
`
`Regeneron Exhibit 1050.001
`
`

`

`APPLIED MICROBIOLOGY, Jan. 1970, p. 146-151
`Copyright © 1970 American Society for Microbiology
`
`Vol. 19, No. 1
`Printed in USA.
`
`Microbiological Aspects of Ethylene
`Oxide Sterilization
`
`I. Experimental Apparatus and Methods
`
`K. KERELUK, R. A. GAMMON, AND R. S. LLOYD
`Research Department, American Sterilizer Company, Erie, Pennsylvania 16512
`
`Received for publication 16 September 1969
`
`A specially built thermochemical death-rate apparatus is described which can be
`used to determine the resistance of microorganisms to ethylene oxide under con-
`trolled conditions. The apparatus was designed to provide instantaneous exposure
`of microorganisms to ethylene oxide and to eliminate variables that could result in
`errors when death kinetic reaction rates are calculated. The apparatus is used to ob-
`tain ethylene oxide resistance data which are useful in evaluating and developing
`sterilizing cycles for materials with known bacterial concentrations, as well as for
`calculating probability factors on which a given test condition can be expected to
`provide sterilization.
`
`Since 1949, when Phillips and Kaye (8) first
`published their work on the sterilizing capabilities
`of ethylene oxide, there have been numerous other
`reports on various aspects of this subject (I, 2, 6,
`7; H. El-Bisi, E. Thompson, and J. J. Perkins,
`Bacteriol. Proc., p. 34, 1962). Various tech—
`niques and apparatuses have been employed in
`investigating the conditions required for ethylene
`oxide sterilization. Ernst and Shull
`(1) used a
`cylindrical pressure vessel in their studies of the
`effects of temperature and concentration on
`ethylene oxide sterilization. Other investigators
`have used anaerobic jars (5, 6), as well as vapor-
`phase resistometers (H. El-Bisi, E. Thompson,
`and J. J. Perkins, Bacteriol. Proc., p. 34, 1962; T.
`Liu, C. R. Stumbo, and G. L. Howard, Ph.D.
`dissertation, Univ. of Mass, 1966) to study bac-
`terial death with ethylene oxide.
`This paper presents details of the methods and
`apparatus used in our laboratories to study the
`resistance of sporeforming and nonsporeforming
`bacteria to ethylene oxide. Preliminary data are
`presented on the use of this apparatus in studying
`the thermochemical death rate of Bacillus subtilis
`var. niger spores. Subsequent papers will de-
`scribe studies concerning the resistance of other
`microorganisms to ethylene oxide and the effects
`of certain environmental factors (sterilant con-
`centration,
`relative humidity,
`spore moisture
`content, temperature, and packaging materials)
`on ethylene oxide sterilization.
`MATERIAL AND METHODS
`
`Organisms. Both sporeforming and nonspore-
`forming microorganisms were used in these studies.
`
`The sporeformers were grown on selected sporulation
`media (9), and the nonsporeformers on appropriate
`media. In both cases, once a suitable population
`was obtained, the organisms were suspended in sterile
`distilled water, washed by centrifugation, resuspended
`in sterile distilled water, and stored at 4 C prior to
`exposure. The details of growth, conditions of incuba-
`tion, media used, and procedures of harvest and
`storage for specific microorganisms will be described
`as pertinent in subsequent articles.
`Selection of carrier. Studies have suggested that
`spores on a hygroscopic carrier are less resistant to
`ethylene oxide than are spores on a nonhygroscopic
`carrier
`(4, 6). To establish comparative data, we
`chose to work with both surface types.
`Two types of nonhygroscopic carriers were con-
`sidered. These were glass heads, 4 mm in diameter,
`and 0.635-cm square, glazed, ceramic tiles. Each
`carrier was prepared as follows.
`Glass beads. B. subtilis var. niger spores (in a
`2.5-ml distilled-water suspension) were placed in a
`250-ml polyethylene bottle containing 1,000 sterile
`glass beads. The bottle was connected to a drive
`shaft which rotated the unit (on its side) at 35 rev/min.
`During rotation, dry air at 45 C was circulated
`through the bottle to facilitate drying of the spores
`on the beads.
`Ceramic tiles. Each tile in three groups of 50 tiles
`each was inoculated with 0.1 ml of a B. subtiIis var.
`niger spore suspension. The inoculated tiles were
`dried in a hot-air oven at 55 C for 1.25 hr. Viable
`spore counts were then prepared.
`The results of this study (Table 1) led to the rejec-
`tion of glass beads as a suitable nonhygroscopic
`carrier, in favor of the ceramic tiles.
`The procedures developed and used to recover
`inocula from the glass beads and ceramic tiles will be
`described in detail.
`Strips 0.635 by 0.375 cm of filter paper (no. 4013,
`
`146
`
`
`
`
`
`isenfiliqZLOZ‘vzMemqejuo/6J0'uise'uiee//:dnquJOJ;pepl20|U/V\O(]
`
`
`
`
`
`Regeneron Exhibit 1050.002
`
`

`

`VOL. 19, 1970
`
`ETHYLENE OXIDE STERILIZATION. I
`
`147
`
`TABLE 1. Comparison of spore counts of B. subtilis var. niger recovered from
`ceramic tiles and glass beads
`
`Group 1
`
`Group 2
`
`Group 3
`
`Carrier type
`
`.
`
`Population“
`Log value
`Population
`Log value
`Population
`; Log value
`F ratio”
`
`0.65
`5.41
`258 X 103 i
`5.40
`252 X 103
`5.42
`268 X 103
`Ceramic tile. . ..
`
`
`
`
`
`
`3.513.20 X 103 §3.928.40 X 1034.22Glass beads..... 16.7 X 103 7.50
`a Values represent the average spore counts obtained from 50 carriers per group.
`b Ratio of two estimates of the sample variance with the second estimate in the numerator.
`
`Numtdiiy
`indicator
`
`
`0 ®
`D
`D
`Q
`I D
`
`' inl,l.l' . r-I-IgI-I-Iq-l-lyl-li-‘rl:
`
`
`
`O
`
`
`
`Thermocoupln
`load Wire;
`
`
`
`
`
` EthyleneOxidtfnonMixluu
`
`
`
`l!»
`U
`
`FIG. 1. Diagramatic representation of the thermochemical death apparatus and a reaction container, with
`auxiliary and recording equipment. Key: 1, conditioning chamber; 2, manifold and heating tape; 3, water heater;
`4, water bath; 5, reaction container; 6, vacuum pump; 7, potentiomenter recorder; 8, gas analyzer.
`
`Schleicher & Schull Co., Keene, NH.) were used as
`the hygroscopic carriers. They were placed in glass
`petri dishes and sterilized in a hot-air oven for 2 hr
`at 170C. Subsequently, each sterilized carrier was
`inoculated with 0.01 ml of the spore- or vegetative-cell
`suspension which contained approximately 108 organ-
`isms per ml. The spore-inoculated carriers were
`dried at 55 C for 1.25 hr;
`those inoculated with
`nonsporeforming organisms were air dried at ambient
`temperature.
`
`Test apparatus. Figure 1 shows the main compo-
`nents of the test apparatus used to determine the
`resistance of the microorganisms on the carriers to
`ethylene oxide.
`Sterilizer. The sterilizer was a Cryotherm (American
`Sterilizer Company, Erie, Pa., model 1016). The
`chamber (25.4 cm in diameter by 40.64 cm in length),
`provided and maintained the required levels of tem-
`perature, humidity, and ethylene oxide concentration
`for the tests. Thermostatically controlled strip heaters
`
`
`
`
`
`isenfiliqZLOZ‘vzMemqeju0/6J0'uise'uiee//:d11quJOJ;pepeoIUAAoc]
`
`
`
`
`
`
`
`Regeneron Exhibit 1050.003
`
`

`

`148
`
`KERELUK, GAMMON, AND LLOYD
`
`APPL. MICROBIOL.
`
`QUICK DISCONNECT COUPLING
`COMPOUND PRESSURE-
`VACUUM GAUGE
`
`PIPE CAP
`
`OARING SE
`THERMOCOUPLE
`AL
`LEAD WIRES
`
`
`
` CONTAINER
`
`FIG. 2. Diagramatic representation of a reaction
`container used with thermochemical death apparatus.
`
`affixed to the exterior chamber walls were the heat
`source. The chamber also included a pressure relief
`valve (Fig. 1A), and a hand valve for admitting
`the sterilant (Fig. 1B).
`Instrumentation for the chamber included a pres-
`sure gauge, graduated in increments of 0.2 psi; a
`vacuum gauge, graduated in increments of 0.25 cm
`of mercury; and a humidity indicator, with a sensor
`inside the chamber.
`An adaptor (Conax Co., Buffalo, NY.) was in~
`stalled in the chamber door to facilitate removal of
`gas samples while the door was closed.
`Manifold. A metal manifold tube equipped with a
`shutoff valve (Fig. 1, C) and covered with double-
`element heating tape (GlasCol Apparatus Co., Terre
`Haute, Ind.) was connected to the sterilizer exhaust
`line and extended over the water bath. This tube
`contained six quick-disconnect couplers (Fig. 1, E)
`for the reaction containers. Each coupler (Snap-Tito
`Co., Oil City, Pa.) was fitted with a hand-valve shutoff
`(Fig. 1, D).
`Heated water bath. The ZS-gal water bath was
`augmented by a constant
`temperature heater and
`circulator
`(Precision Scientific Co., Chicago, 11].,
`no. 66567 and 66540, respectively).
`Reaction containers. The containers shown in Fig.
`2, hold the inoculated carriers for exposure to the
`gaseous atmosphere contained in the sterilizer cham-
`ber. The containers are brass cylinders 125 mm long
`and 4 mm thick with an inside diameter of 44 mm.
`One end of each container is closed with the opposite
`end threaded to receive a standard pipe cap. The pipe
`cap, with four 0.375-cm openings on the top, was
`
`TABLE 2. Amount of water required in chamber
`to obtain a specific reIative humidity
`in exposure system
`
`Humidity Distilled water
`
`ml
`
`%
`1 5
`30
`50
`60
`90
`
`
`19995"?OOO3.3o'c
`
`equipped with two 0.375-cm adaptors (Conax Co),
`a 0.375~cm shutoff valve, and a compound pressure
`vacuum gauge. A copper-constantan thermocouple
`for sensing the temperature within the container was
`inserted through one of the Conax adaptors. The
`other Conax adaptor was used for the removal of
`gas samples from the container for chromatographic
`analysis. In operation,
`the equipped pipe cap was
`screwed onto the reaction container (with inoculated
`carriers inside) until a seal was formed against a
`rubber O—ring located on the container. For connec-
`tion to and release from the gas manifold, the shutoff
`valve on the reaction container pipe cap was equipped
`with a quick—disconnect coupling. This arrangement
`provided for two shutofi' valves between the gas
`manifold and each reaction container.
`Monitoring equipment. The temperature inside
`the gas chamber was determined by thermocouples
`attached to the inside of the chamber. These thermo-
`couples and those in the reaction containers were
`connected by wire leads to a multipoint recorder
`(Minneapolis~Honeywell Co., Philadelphia, Pa.) for
`direct temperature monitoring throughout the tests.
`Chamber humidity was determined by a humidity
`sensor connected to a humidity indicator (El-Tronics,
`Inc., Mayfield, Pa., model 1106) on top of the cham-
`ber. The desired relative humidity within the chamber
`was attained by injecting a predetermined amount of
`distilled water into the chamber prior to each test run.
`A Lira infrared gas analyzer
`(Mine Safety Ap-
`pliance Co., Pittsburgh, Pa., model 300), connected
`to the gas chamber, was used to analyze the ethylene
`oxide concentration during the tests.
`Sterilant supply. The sterilant used for the tests
`was a gaseous mixture of 12% ethylene oxide and
`88% dichlorodifluoromethane by weight, contained
`in a 145-lb. cylinder
`(Pennsylvania Engineering
`Company, Philadelphia, Pa).
`Exposure procedure. The test procedures used
`with the experimental exposure apparatus were as
`follows, in the order stated. (i) The sterilizer chamber,
`manifold, and water bath were brought to 54.4 :i: 3 C.
`(ii) The gas analyzer was calibrated according to
`operating instructions. (iii) Distilled water (Table 2)
`was added to the chamber to provide the desired
`relative humidity.
`(iv) Five inoculated hygroscopic
`carriers and five inoculated nonhygroscopic carriers
`in individual, sterile, glassine envelopes, were placed
`in each of six reaction containers. The containers
`were sealed, connected to a vacuum pump, and
`
`
`
`
`
`isenfiliqZLOZ‘vzMemqejuo/6J0'uise'uiee//:d11quJOJ;pepl20|U/V\O(]
`
`
`
`
`
`Regeneron Exhibit 1050.004
`
`

`

`VOL. 19, 1970
`
`ETHYLENE OXIDE STERILIZATION. I
`
`149
`
`TABLE 3. Comparison of recovery from nonhygroscopic and hygroscopic carriers
`subjected to various recovery procedures
`
`i
`
`i
`
`Carrier“
`
`3.221331%;
`
`Time shaken
`hr
`
`be§(%1§;§§stsion
`
`Time in sonicator
`min
`
`Nonhygroscopic
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`
`'
`‘
`
`l
`.
`
`1
`2
`4
`15+
`1
`2
`4
`15+
`15+
`
`6
`6
`6
`6
`6
`6
`0
`6
`12
`
`0
`0
`0
`0
`5
`5
`5
`5
`10
`
`
`
`
`
`
`Recovery
`%
`
`63.0
`66 .0
`68 .0
`66 .0
`88 .0
`72 . 0
`90 .0
`93 .0
`99 .0+
`
`
`
`
`
`
`
`
`
`isenfiliqZLOZ‘vzMemqejuo/6J0'uise'uiee//:dnquJOJ;pepeoIUAAoc]
`
`Hygroscopic
`
`0
`0
`i
`l
`0
`1
`E
`2
`0
`1
`2
`3
`i
`a Minimum of five carriers was used for each set of conditions.
`5 Dilution blanks contained 1% Darvan in 99 ml of distilled water.
`f Substrate.
`
`Waterc
`Water‘
`Darvanc
`
`3
`
`100
`100
`100
`
`evacuated to 67.58 cm of mercury. They were then
`connected to the manifold via the quick-disconnect
`couplers and suspended in the water bath. The shutoff
`valves on the containers and their respective conneco
`tions were closed. Each container with inoculated
`carriers was suspended in the water bath and heated
`to 54.4 C (approximately 15 min) before the contents
`were exposed to ethylene oxide. (v) The shutoff valve
`between the chamber and the manifold was opened.
`The chamber and the manifold were then evacuated
`to 67.58 cm of mercury and charged to a preselected
`pressure with the sterilant. (vi) After a 6-min stabiliza-
`tion period,
`the sterilant concentration, chamber
`temperature, and humidity were measured by the
`instrumentation noted previously. Deviations of these
`factors from the preselected test conditions were
`adjusted at this time. (vii) Once the preselected test
`conditions were established, the sterilant was trans-
`ferred to each pretempered reaction container by
`opening the shutoff valve leading from the manifold
`to the reaction container and the shutoff valve on
`the reaction container. This initiated the exposure
`period for each container.
`At predetermined intervals during the exposure
`period, the two shutoff valves between the manifold
`and one of the reaction containers were closed. This
`reaction container was then disconnected from the
`manifold and chilled in ice water (5 to 10 min) while
`being evacuated, returned to atmospheric pressure,
`and opened. The inoculated carriers were then re-
`moved for survivor counts. Initial total viable spore
`counts were prepared from inoculated carriers placed
`in one of the reaction containers and removed from
`the manifold and water bath prior to introducing the
`sterilant into the remaining containers.
`Recovery procedures. Comparative studies were
`performed to develop procedures which would yield
`maximal
`recovery of viable organisms from the
`inoculated carriers.
`
`Inoculated nonhygroscopic carriers were trans-
`ferred to dilution blanks containing 99 ml of 1%
`aqueous Darvan (R. T. Vanderbflt Co., New York,
`N.Y.) and a variable number of 4-mm diameter glass
`beads. Darvan, a polymerized sulfonic acid salt, was
`used to enhance the dispersement of the inoculum.
`The dilution blanks were placed in a reciprocating
`shaker and then in a sonic disintegrator for various
`time intervals. Sonic treatment was employed for
`further dispersion of the inoculum from the carrier
`surface.
`To obtain a 99.9% recovery of viable inoculum
`from the inoculated hygroscopic carriers (filter paper
`strips), a 2—min mixing in a Waring blendor microcup
`was required.
`As a result of these studies (Table 3), the following
`recovery procedures were adopted for treating inocu-
`lated carriers after exposure to the test conditions.
`Immediately after exposure, each treated non-
`hydroscopic carrier was transferred to a dilution
`blank containing 99 ml of aqueous 1% Darvan solu-
`tion and 12 (4 mm in diameter) glass beads. The
`dilution blanks were placed in a reciprocating shaker,
`in water at 4 C, and shaken overnight. The dilution
`blanks were then placed in a sonic disintegrator and
`exposed to 20 kc/sec for 10 min.
`The treated hygroscopic carriers were transferred
`to 99 ml of sterile, distilled water in a Waring blender
`microcup and were blended for 2 min. After these
`recovery procedures, serial dilutions were then pre-
`pared and viable cell counts were determined in
`various plating media.
`
`RESULTS AND DISCUSSION
`
`An analysis of the variance (3) among the
`viable spore count values obtained from three
`different groups of the glass beads demonstrated
`
`Regeneron Exhibit 1050.005
`
`

`

`150
`
`KERELUK, GAMMON, AND LLOYD
`
`APPL. MICROBIOL.
`
`“)0NO.
`
`SURVIVORS
`
`TIME IN MINWES
`
`spores of Bacillus
`FIG. 3. Survivor curves for
`subtilis var. m‘ger exposed to ethylene oxide (500 mg/
`liter; 54 :l: 3 C; 30 to 50% relative humidity). Fig. 3a,
`nonhygroscopic carrier. Fig. 3b, hygroscopic carrier.
`
`statistically the inconsistency of using the beads
`as inoculum carriers.
`irregularities were
`Gross viable spore count
`found among the bead groups. This did not occur
`among the ceramic tiles, as is indicated by viable
`spore counts and the derived F ratio. When the
`same statistical test was applied to the viable
`counts from three groups of tiles, the advantage
`of this type of carrier was clearly demonstrated
`as the variance among the three groups was
`negligible.
`The F ratio is the ratio of two estimates of the
`sample variance with the second estimate in the
`numerator (3). This type of analysis is one in
`which observations are classified into groups on
`the basis of a single property, for example, carrier
`populations. There were three population groups
`for each carrier. The sample variance for each
`group was obtained and the three values were
`averaged to give the first estimate of the variance;
`the sample variance for the three mean values of
`the three groups was also calculated to give the
`second estimate of the variance. The variance
`values were used to calculate the F ratio. Large
`F—value ratios would indicate significant differ-
`ences among the three populations.
`The tiles had an F ratio of 0.65 as compared to
`7.50 for the glass beads as noted in Table 1.
`Ernst and Shull (1) noted the variation in the
`count of individual beads and tried to minimize
`this variation by using three beads per tube since
`
`they did not utilize the bead carriers quan-
`titatively.
`The use of the specially designed thermo-
`chemical death rate apparatus aided in reducing
`some of the possible errors in death kinetic
`studies, once the conditions of exposure were
`established in the main chamber of the apparatus.
`The transfer of these conditions to the reaction
`container was instantaneous. Similarly, once the
`exposure period for a group of inoculated carriers
`was completed,
`the design of the apparatus
`allowed the immediate removal of the carriers
`from the test environment.
`Chromatographic analyses of samples of the
`chamber atmosphere demonstrated that once the
`main chamber and reaction containers had been
`charged with a specific concentration of ethylene
`oxide, it did not vary greatly from one reaction
`container
`to another. No difficulty was en-
`countered in maintaining constant humidity and
`temperature within the sterilizer chamber and
`reaction containers.
`
`From these investigations, 3. recovery procedure
`was developed which yielded 99% (or better)
`viable inocula from the inoculated and exposed
`carriers (Table 3).
`Survivor curves were prepared for spores and
`cells of microorganisms dried on both hygro-
`scopic and nonhygroscopic surfaces and exposed
`to ethylene oxide in the test apparatus. Each point
`used in preparing the curves represents an average
`number of survivors. The straight-line portion of
`all curves was located by linear
`regression.
`Analysis of death kinetic reaction rates, when
`plotted on semilogarithmic paper,
`is usually
`expressed in terms of the decimal reduction, or D
`value. The D value is the time required to destroy
`90% of the bacterial cell or spore population
`under a given set of conditions (9). This value
`can be useful
`in determining the theoretical
`probability of bacterial cells or spores surviving a
`given sterilizing agent or process. From the
`curves, the decimal reduction times (D values at
`54.4 C—concentration of ethylene oxide in milli-
`grams per liter) were taken as the time (in min-
`utes) to kill 90% of the spores. Survivor curves
`of B. subtilis var. niger spores, deposited on both
`hygroscopic and nonhygroscopic surfaces and
`exposed to ethylene oxide, are presented in Fig. 3,
`which represents typical data obtained.
`LITERATURE CITED
`
`1. Ernst, R. R., and I. I. Shull. 1962. Ethylene oxide gaseous
`sterilization 1. Concentration and temperature efi'ects. Appl.
`Microbiol. 10:337-341.
`2. Gilbert, G. L., V. M. Gambil, O. R. Spinner, R. K. Hoffman,
`
`
`
`
`
`isenfiliqZLOZ‘vzMemqeju0/6J0'uise'uiee//:d11quJOJ;pepl20|U/V\O(]
`
`
`
`
`
`
`
`Regeneron Exhibit 1050.006
`
`

`

`VOL. 19, 1970
`
`ETHYLENE OXIDE STERILIZATION. I
`
`151
`
`and C. R. Phillips. 1964. Effects of moisture on ethylene
`oxide sterilization. Appl. Microbiol. 12:496—503.
`3. Heel, P. G. 1960. Elementary statistics. John Wiley & Sons,
`Inc., New York.
`4. Kay. 3., and C. R. Phillips. 1949. The sterilizing action of
`gaseous ethylene oxide. IV. The efi‘ect of moisture. Amer. J.
`Hyg. 50:296-306.
`5. Liu, T., G. L. Howard, and C. R. Slumbo. 1968. Dichlorodi.
`fluoromethane—ethylene oxide mixture as a sterilant at ele-
`vated temperatures. Food Technol. 22:86—89.
`
`6. Opfell, J. B., J. P. Honmann, and A. B. Latham. 1959. Ethyl-
`ene oxide sterilization of spores in hygroscopic environ-
`ments. J. Amer. Pharm. Ass. 48:280-282.
`7. Phillips, C. R. 1952. Relative resistance of bacterial spores and
`vegetative bacteria to disinfectants. Bacteriol. Rev. 16:135‘
`138.
`8. Phillips, C. R., and S. Kaye. 1949. The sterilizing action of
`gaseous ethylene oxide. 1. Review. Amer. J. Hyg. 50:270~279.
`9. Stumbo, C. R. 1965. Thermobacteriology in food processing.
`Academic Press Inc., New York.
`
`
`
`
`
`isenfiliqZLOZ‘vzMemqeju0/6J0'que'uJee//:d11quJOJ;pepl20|U/V\O(]
`
`
`
`
`
`
`
`Regeneron Exhibit 1050.007
`
`