Confirmation of the Sterility of Medical Devices and Changes in the Characteristics of Polymers Upon Sterilization

Biocontrol Science, 2001, Vol.6, No.2, 63-68

Minireview

Confirmation of the Sterility of Medical Devices and Changes in the Characteristics of Polymers upon Sterilization

HIDEHARU SHINTANI

National Institute of Health Sciences, 1-18-1, Kamiyoga, Setagaya, Tokyo 151-0041, Japan Received 15 November 2000/Accepted 20 November 2000

Key words : Sterility/Sterilization/Medical Devices.

STERILITY be sterile complies with the requirements set forth in the individual monograph with respect to the test for Medical devices may be sold and used in two differ- sterility. In view of the possibility that positive results ent conditions in terms of their bioburden (that is, the may be due to faulty aseptic techniques or environ- number of viable microorganisms present on the de- mental contamination during testing, provisions are in- vice) : sterile or nonsterile. If they are intended to be cluded under the interpretation of sterility test results used in the sterile condition, the actual act of steriliza- for two stages of testing. tion may be either performed by the manufacturer or Alternative procedures may be employed to demon- by the user (though the latter is commonly the case strate that an article is sterile, provided the results ob- only for reusable devices and instruments, such as tained are equivalent or greater reliability. When a surgical instruments at a health care facility). If the difference appears or dispute occurs, evidence of mi- manufacturer has sterilized the instruments, then this crobial contamination must be confirmed by the vali- is specified in the Device Master File (and in product dated procedure. The result so obtained is conclusive related literature), along with the means for having of the failure of the article to meet the requirements of achieved a suitable degree of sterilization and specifi- the test. Similarly, failure to demonstrate microbial cation of a program to assure proper performance of contamination by the validated procedure is evidence the process. that the article meets the requirements of the test. The sterility assurance level (SAL) is the statistical The following considerations apply to sterilized de- probability that a device is not sterile even after going vices manufactured in lots, each consisting of a num- through an established sterilization process. Residual ber of units. Special considerations apply to sterile microorganisms are usually measured as colony devices manufactured in small lots or in individual forming units (CFUs), with the amount of sterilization units where the destructive nature of the sterility test required to reduce the number of CFUs by an order of renders such conventional sterility tests impractica- magnitude for a specific device being termed the ble. For these articles, validated modifications to the D10 (the sterilization period or radiation dose or pe- sterility tests must be made. riod to produce a tenfold reduction in bioburden). For articles of such size and shape as to permit D10 here is both a concentration (or, for radiation, a complete immersion in not more than 1000 ml of cul- power level) and the time interval the device (or ma- ture medium, test the intact article, using the validated terial) is exposed to that dose (Prince and Rubino, media, and incubate for a validated incubation period. 1984). For devices having hollow tubes, such as transfu- sion or infusion assemblies, or where the size of an Sterility tests item makes immersion impracticable and where only The following procedures are applicable for deter- the fluid pathway must be sterile, flush each lumen of mining whether a pharmacopeial article purporting to 20 units with a sufficient quantity of fluid thioglycollate medium and each lumen of the same 20 units with a Corresponding author. Tel : +81-3-3700-9268, Fax: +81- sufficient quantity of soybean-casein digest (SCD) 3-3707-6950. medium to yield a recovery of not less than 15 ml of 64 H. SHINTANI each medium. Then, incubate not less than 100 ml of test a validated number of units. each of the two media as directed under the validated procedure. For devices in which the lumen is so small INTERPRETATION OF STERILITY TEST that the fluid thioglycollate medium will not pass RESULTS through, substitute an alternative thioglycollate medium for fluid thioglycollate medium, but incubate the At the prescribed intervals during and at the conclu- validated medium anaerobically. sion of the incubation period, examine the contents of Where the entire intact article, because of its size all of the vessels for evidence of microbial growth, and shape, cannot be tested for sterility by immersion such as the development of turbidity and/or surface in not more than 1000 ml of culture medium, expose growth. If no growth is observed, the article tested that part of the article most difficult to sterilize and test meets the requirements of the test for sterility. that part, or where possible, remove two or more parts When microbial growth is found, but a review of the each from the innermost part of the article. Aseptically sterility testing facility of the monitoring, materials transfer these article parts to the specified number of used, testing procedure, and negative controls indi- vessels of validated media in a volume of not more cates that inadequate or faulty aseptic technique was than 1000 ml and incubate for the validated incuba- used in the test itself, the first stage should be de- tion period. clared invalid and may be repeated. Where the presence of the test specimen in the me- If microbial growth is observed and there is no evi- dium interferes with the test because of bacteriostatic dence invalidating the first stage of the test, proceed or fungistatic action, rinse the article thoroughly with a to the second stage. minimal amount of rinse fluid. Recover the rinse fluid, The minimal number of specimens selected should and test as directed for devices under validated test be double the number tested in the first stage. The minimal volumes tested from each specimen and the procedures using membrane filtration. media and incubation periods are the same as those indicated for the first stage. If no microbial growth is Sterile empty or prefilled syringes found, the article tested meets the requirements of the Sterility testing of prefilled syringes is performed by test for sterility. If, however, it can be demonstrated employing the same techniques used in testing sterile that the second stage was invalid because of faulty or products in vials or ampuls. The direct transfer tech- inadequate aseptic technique in the performance of nique may be employed if the bacteriostasis and the test, the second stage may be repeated. Sterility fungistasis determination has indicated no adverse testing can be used as part of a quality assurance activity under the test conditions. Where appropriate, the membrane filtration procedure may be employed. program for a production lot or batch or as one of the quality control criteria for release of such a lot or For prefilled syringes containing a sterile needle, flush batch. the contained produce through the lumen. For syringes packaged with a separate needle, aseptically STERILIZATION attach the needle, and expel the product into the validated media. There are three major means employed for steriliz- Special attention should be paid toward demon- ing medical instruments (Nair, 1995), each of which strating that the outer area of the attached needle has advantages and drawbacks. (that part which will enter the patient's tissues) is sterile. For empty sterile syringes, take up sterile me- Heat sterilization dium or diluent into the barrel through the needle if at- Heat (steam) is effective and inexpensive but can tached, or, if not attached, through a sterile needle only be used for medical devices of a relatively small attached for the purpose of the test, and then express volume and materials, which are not degraded or de- the contents into the validated media. formed by it. Accordingly, it is typically useful only for Devices that are purported to contain sterile path- smaller metal devices and cannot always be success- ways may be tested for sterility by the membrane fil- fully applied to plastics and such. It is most commonly tration technique as follows. Aseptically pass a used for reusable steel instruments and glassware. sufficient volume of test fluid through each of not less Steam sterilization by autoclaving has traditionally than 20 devices so that not less than 100 ml is recov- been the most widely used method for medical instru- ered from each device. Collect the fluids in aseptic ments. Today, prevacuum, high-temperature steam containers, and filter the entire volume collected sterilization is considered to be the safest and most through membrane filter funnel (s). practical means of sterilizing the majority of surgical Where the devices are large and lot sizes are small, instruments, surgical dressings, fluids, fabrics, and CONFIRMATION OF STERILITY OF MEDICAL 65 other absorbent materials. The process should be PET has shown that oligomers and, in particular, the used with caution while sterilizing polymers and com- cyclic trimer increases on PET with repeated auto- posites as both heat and steam can drastically alter claving. Studies (Nair, 1995) of the PET materials in- their properties. The main deleterious effect from dicated increased values of heat of fusion and steam sterilization of polymeric materials occurs when percent crystallinity, which suggested reorganization hydrolysis of the polymer takes place leading to unde- of the amorphous phase. Generation of new crystal- sirable contaminants. Under prolonged steam auto- line regions was ruled out as there was no increase in claving, a few ppb of 4,4'-methylene dianiline (MDA) the intensity of the infrared peak of 973 cm' which or less has been detected (Shintani, 1989, 1991, indicates crystallinity. Chain scission of amorphous 1992 and 1995) in the aqueous extract of methyl regions, resulting in the formation of new amorphous diisocyanate (MDI)-based polyurethane which was regions, was, therefore, believed to be the reason for attributed to the hydrolysis of the polymer. the increased values of crystallinity. The changes in Researchers observed the release of contaminating the molecular weights of Mn (number average of mo- substances, whose structure was not elucidated, lecular weight) and Mw (weight average of molecular when polyvinyl chloride (PVC) -containing aqueous weight) suggested that though degradation would suspensions were autoclaved (Habermann and take place when autoclaving for 15 min, subsequent

Waitzova, 1985). Repeated autoclaving of PVC in- sterilization of 30 min results in solid-state polymeriza- tended for biomedical applications was found to in- tion or cyclization reactions. These cyclization reac- crease the mechanical properties of PVC of both tions were responsible for the formation of fresh cyclic covered and uncovered PVC samples. The change in trimer molecules which could migrate to the surface. the mechanical properties was attributed to the rear- As moisture and air enhance oxidation and hydrolysis rangements in the PVC macromolecular chains giving reactions, subsequent autoclaving for 60 min would rise to branches of varied lengths. Repeated autoclav- result in more degradation of amorphous regions. The ing of uncovered PVC samples for 150 minutes was increased temperature could also facilitate the easy seen to induce higher leaching of the plasticizer such migration of the cyclic trimer to the surface. as phthalate esters (Shintani, 1985 and 2000)

Seemingly stable polymers can also undergo Chemical sterilization changes in the surface morphology, especially if the Chemical sterilants include ethylene oxide, glass transition temperature of the polymer is ex- glutaraldehyde, and formaldehyde. These can all be ceeded during autoclaving. This may cause changes used for most devices that do not have any readily in biocompatibility or blood contacting properties of chemically reactive (degradable) components. the material. Oligomers or other inadvertent contami- Ethylene oxide (ETO) is the most widely used, but is nants introduced subsequent to steam sterilization a mutagen and carcinogen, and is neurotoxic and may have significant effect on the biocompatibility causes hypersensitivity responses in some individuals and the performance of the polymers. A scanning (Golberg, 1986). electron microscopy (SEM) study of different types of Ethylene oxide is an effective bactericide active at arterial prostheses made from Dacron found that all temperatures as low as 60 •Ž. As ethylene oxide can the prostheses were coated with a layer of oligomers easily diffuse into materials to be sterilized, it is possi- on autoclaving. The Cooley knitted Dacron grafts ble to sterilize heat- or moisture-sensitive materials show the maximal amount of oligomer crystals, which even through sealed plastic wrapping. One of the dis- was progressively increased with each subsequent advantages of this method for polymeric materials is successive resterilization procedure. This coating of that they retain varying amounts of ethylene oxide oligomers also resulted in increased hemolysis. (Shintani, 1979 and 1981). Residual ethylene oxide in Berger and Sauvage (1981) also noted late fiber de- sterilized plastic tubing has been reported to cause terioration in 493 Dacron arterial prostheses, which hemolysis of blood in heart-lung surgery (Golberg, had been implanted for 3-15 years. One of the contrib- 1986). Adverse hemolytic reactions were also caused uting factors to this deterioration was believed to be when blood was exposed to plastics that retained sig- autoclaving. nificant quantities of residual ethylene oxide

Commercial poly (ethyleneterephthalate) (PET) is (Golberg, 1986). Ethylene oxide also possesses toxic known to contain cyclic and linear oligomers as natu- properties and is a strong alkylating agent (Golberg, ral impurities, which are formed as a byproduct during 1986; Shintani, 1998). As the ethylene oxide consists polymerization and texturing (Shintani, 2000). Cyclic of two carbon atoms and an oxygen atom linked to- trimer is a major oligomer component, and the content gether in an unstable three-membered ring, it can re- of cyclic trimer is actually a criterion for the quality of act with various functional groups (Shintani, 1998). PET. The study of the effect of steam sterilization on These functional groups are the sulfhydryl, amino, 66 H. SHINTANI carboxyl, and hydroxyl groups of proteins and nucleic devices. If any chemical sterilants are used, residual acids. The lethal effect of the ethylene oxide arises levels must be determined and kept below (or from its alkylating effect on these susceptible mole- brought below by allowing the devices to vent) certain cules. However, due to the reactivity of the gas with established limits. such functional groups, this type of sterilization should only be carried out at room temperature with materials Radiation sterilization such as polyurethanes, polyesters, and other poly- An alternative to ethylene oxide or steam heat ster- mers with potentially reactive groups. Several of the ilization is the use of ionizing radiation to sterilize composite polymeric materials used these days, such medical devices. The advantages of radiation sterili- as albuminated Dacron grafts or heparinized grafts, zation are high efficiency, negligible thermal effect, are liable to lose the beneficial effect extended by the and the fact that articles may be packaged and sealed albumin coating or heparin coating by subjecting prior to sterilization. The disadvantages are that some these materials to ethylene oxide sterilization. plastics are degraded (Shintani, 2000), and certain Ethylene oxide sterilization of heparinized polymers drugs become less active due to radioinduced chemi- reduces their nonthrombogenic properties (Bruck, cal changes. Additionally, if a 60Co source is used, 1971), while Guidoin et al. (1985) reported that ethyl- the sterilization time is long. ene oxide sterilized albuminated polyester grafts were Irradiation of high polymers results in either cross- associated with a slower rate of healing when im- linking or degradation depending on the chemical na- planted in comparison to a similar graft sterilized by ir- ture of the system. Though the accepted dose for radiation. Though retention of residual ethylene oxide radiation sterilization is 25 kGy, it is known that cer- is believed to be the cause for this slower rate of heal- tain species require up to 60 KGy for complete de- ing, it is likely that the alkylating action of ethylene ox- struction. This must be validated to define the ide could denature the albuminated surface forming sterilization dose setting. The effects of irradiation of products that delay the healing process. increased doses or multiple sterilization treatments Most manufacturers of ethylene oxide sterilizer are assumed to be cumulative. Polymers which tend equipment routinely recommend at least five and pref- to cross-link show a steady increase in their molecular erably 14 days of aeration for ethylene oxide sterilized weights with increasing doses and with the formation plastic tubing (Shintani, 1982). However, the time for of branches until a three-dimensional network is aeration is dependent on the material sterilized formed, while polymers that undergo degradation ex- (Shintani, 1982). Highly porous materials such as cel- hibit chain scission, the molecular weight steadily de- lulose, paper, and natural rubber show very high val- creases with increasing radiation doses and ues of ethylene oxide solubility and low diffusion mechanical properties of the system simultaneously coefficients (Shintani, 1979 and 1982). A study (Vink undergo changes (Shintani, 1990, 1991 and 1995). and Pleijsier, 1986) of residual ethylene oxide content These properties include tensile strength, elastic in different polymers revealed that for a number of hy- modulus, shear strength, elongation, and occasion- drophilic materials, the residual content was well ally, color changes (Shintani, 1990, 1991 and 1995). above the levels which are currently considered to be Though cross-linking and chain scission processes safe even after aeration for 14 days. For a reliable de- may occur simultaneously, usually one or the other termination of aeration times after ethylene oxide ster- predominates. In the case of vinyl polymers, if the car- ilization of medical devices, the type of the material bon atoms of the main chain carry at least one hydro- from which the device is made and, in particular, its gen atom, the polymer tends to cross-link, whereas if thickness as well as chemical characteristics should the carbon atoms are fully substituted, the polymer be considered. tends to degrade. There is a correlation between Another point worth noting on the subject of cross-linking tendency and the heat of polymerization, deaeration is the procedure for estimation of residual showing that polymers which exhibits a high heat of ethylene oxide after sterilization. In a project involving polymerization (above approximately 16 kcal/mole) the interlaboratory comparisons of the procedures for cross-link, while polymers that exhibit a low heat of the estimation of ethylene oxide, it was observed polymerization (less than 16 kcal / mole) degrade. (Marlow et al. 1987) that the average estimated total Irradiation of polymers can also result in gas evolution coefficient ranged from 8 to 22% even when following double bond formation, and the production of trapped the same standard procedure of residual ethylene ox- free radicals. The presence of impurities can acceler- ide estimation, emphasizing the need for standard and ate the degradation process with the possibility of reliable procedures for estimating the residual ethyl- producing irritants or other undesirable products. Free ene oxide content. The concern with all of the chemi- radicals produced in polymers by irradiation in air con- cal sterilants is residuals left on and (permeated) in vert to peroxidic radicals. Additionally, the irradiation CONFIRMATION OF STERILITY OF MEDICAL 67 process may result in discoloration of polymeric prod- increased stability of Dacron or PET fabrics coated ucts. with albumin during irradiation, but the results of the While general effects of irradiation on polymeric investigation also indicate a leakage of albumin with materials have been extensively studied (Chapiro, irradiation. The amount of albumin released was 1962) , relatively small changes in physicochemical, found to be dependent upon the nature of fabric used. mechanical, and biological properties may be toler- The migration of chicken embryonic cells was also able in the short term, whereas similar changes may observed to be decreased for the irradiated fabrics. lead to catastrophic failures in long-term application. Another group of polymers used extensively in bio- Besides, seemingly unchanged materials may also medical applications is the polyurethanes. The y irra- suffer some minor damages which affect their biologi- diation of polyurethane based on 4,4'-diphenyl cal and mechanical performance. In one case, the sta- methane diisocynate (MDI) has been reported to pro- bility of cellulosic polymers was recognized up to 200 duce the carcinogen of 4,4'-methylene dianiline kGy, however in another case, the regenerated cellu- (MDA) with simultaneous detection of polyurethane lose membranes were not always maintained charac- oligomers (Shinatani, 1989, 1991, 1992, 1993, 1995 teristics with radiation sterilization. Rose et al. (1984) and 1996). The y irradiation of poly (DL -lactide/ have observed that the wear of polyethylene exposed glycolide)-type microspheres has been reported to re- to y irradiation increased with dosage and contact sult in decreased molecular weights with the degrada- stress, becoming measurable in many cases only af- tion continuing on storage. An inadvertent effect is the ter a critical dose or stress was exceeded. The most change in the release pattern of the drug loaded in the significant effect noted was the pressure dependence microspheres. Higher molecular weight polyglycolide of the wear rate. The increase in wear rate appeared sutures which were initially more resistant to enzy- to be a combination of chain scission and oxidation matic degradation become more prone to enzymatic suggesting that the radiation should be carried out in attack as a consequence of the altered physical and an inert atmosphere. chemical structure obtained from y irradiation. Study of the effects of aging of y radiation- Melberg et al. (1988) also report that the irradiation of sterilized isotactic polypropylene revealed that though PVC catheters of an external insulin pump produced both cross-linking and degradation occur simultane- chemical transformation products and damaged the ously during irradiation and postirradiation, branching insulin solution. takes place during aging (Shintani, 1990, 1991 and Bacterial endotoxins were found to be unaffected 1995). A dose of 25 KGy was sufficient to introduce by y irradiation, and, hence, gloves and software changes in crystallinity. These changes were attrib- sterilized by irradiation have been observed uted to the semicrystalline nature of the polymer and (Shumnes and Darby, 1984) to contain significant were observed to be different for the covered and un- amounts of bacterial endotoxins. As another inadver- covered samples. Additionally, the transitions be- tent effect, it was also observed that the endotoxin tween short- and long-range order and short- and levels were increased when the bacterial counts were long-duration stiffness in both covered and uncovered elevated, in some cases leading to contact dermatitis samples were affected by the formation of branches (Shumnes and Darby, 1984). The y irradiation of in the backbone of the polymer. A higher degree of intraocular lens, however, decreased the inflammatory branching was observed in covered samples resulting reactions observed on implantation of ethylene oxide- in greater long-duration stiffness in the transition be- sterilized lens. Ludwig et al. (1988) attributed this de- tween short- and long-duration stiffness. The higher creased response not only to less absorption of the degree of branching in the covered samples was at- toxic agent, but also to polymerization of any residual tributed to the high diffusion of energized oxygen into monomeric methylmethacrylate. The y radiation or the polymer matrix when compared to the uncovered ultraviolet (UV) light are both used, with the former samples. being very predominant on the industrial scale. PET, used extensively in biomedical applications, Radiation has significant advantages, but there is a has been believed to withstand radiation sterilization significant capital investment involved, safety precau- without significant degradation. However, a study of tions must be vigorously enforced, and some poly- PET fibers revealed that a single dose of 25 KGy was mers are degraded (though this usually translates sufficient to induce chemical changes, which were re- only to being discolored) by the absorbed energy in a flected in the changes of crystallinity and breaking gamma field. load of the fibers (Shintani, 2000) . These changes were attributed to the degradation of PET in the amor- REFERENCES phous regions and the recombination of degraded ali- phatic segments. Baquey et al. (1987) report on the Baquey, C., Sigot-Luizard,M. F., Friede, R. E., Prud'hom, R. 68 H. SHINTANI

E., and Guidoin, R. G. (1987) Radiosterilization of Appl. Polym. Sci., 41, 661-675. albuminated polyester prostheses. Biomaterials., 8, 185- Shintani, H. (1991) Degradation and cross-linking of polyu- 189. rethane irradiated by gamma-rays. Polymer Degrad. Berger, K., and Sauvage, L. R. (1981) Late fiber deteriora- Stability, 32, 191-208. tion in dacron arterial grafts. Ann. Surg. 193, 477-91. Shintani, H. (1991) Effects of gamma-ray irradiation on the Bruck, S, D. (1971) Sterilization problems of synthetic change of characteristics of medical-use polyurethane. biocompatible materials. J. Biomed. Mater. Res., 5, 139- Polymer Degrad. Stability, 32, 17-30. 158. Shintani, H. (1991) Formation of 4,4'-methylenedianiline in Chapiro, A. (1962) Radiation Chemistry of Polymeric polyurethane potting materials by either y -ray or auto- Systems. Wiley-lntersciences, New York. clave sterilization. J. Biomed. Mater. Res., 25, 1275-1286. Golberg, L., (1986) Hazard Assessment of Ethylene Oxide. Shintani, H. (1991) Mechanism of degradation and CRC Press, Boca Raton, FL. crosslinking of polyurethane when irradiated by gamma- Guidoin, R., Snyder, J., and King, M. (1985) A compound rays. J. Appl. Polym. Sci., 42, 1979-1987. arterial prosthesis: The importance of the sterilization pro- Shintani, H. (1992) Solid-phase extraction (SPE) and HPLC cedure on the healing and stability of albuminated polyes- analysis of toxic compounds and comparison of SPE and ter grafts. Biomaterials., 6, 122-128. liquid-liquid extraction. I. Analysis of 4,4'- Habermann, V., and Waitzova, D. (1985) On the early methylenedianiline in serum. II. Analysis of components of evaluation of extracts from synthetic polymers used in dental materials. J. Liquid. Chromatogr.,15, 1315-1335. medicine. Arch. Toxicol., Suppl., 8, 45-460. Shintani, H. (1992) Gamma radiation and autoclave sterili- Ludwig K., Scheiffarth O. F., and Von-Meyer, L. (1988) zation of polyurethane. J. Rad. Sterilization, 1, 11-28. Reducing the amount of monomers in intraocular lenses Shintani, H. (1993) The Monitoring and evaluation of degra- through sterilization by gamma radiation. Ophthalmic. dation and cross-linking in polyurethane when irradiated Res., 20, 301-307. by gamma-rays during sterilization. In Ohio Science Marlow, D. E., Lao, N. T., Eaton, A. R., Page, B. F., and Lao, Workbook: Polymers (Elfner, L.E., ed.) 84-87, The Ohio C, S. (1987) Interlaboratory comparison of analytical Academy of Science, Ohio. methods for residual ethylene oxide in medical device ma- Shintani, H. (1995) Relative safety of gamma-ray, autoclave terials. J. Pharm. Sci., 76, 333-337. and ethylene oxide gas sterilization of thermosetting Melberg, S. G., Havelund, S., Villumsen, J., and Brange. J. polyurethane. Biomed. lnstrum. Technol., 29, 513-519. (1988) Insulin compatibility with polymer materials used Shintani, H. (1995) Gamma-ray irradiation, autoclave and in external pump infusion systems. Diabetic Med., 5, 243- ethylene oxide sterilization to thermosetting polyurethane. 247. Rad. Phys. Chem., 46, 377-381. Nair, P. D. (1995) Currently practiced sterilization methods- Shintani, H. (1995) HPLC analysis of toxic additives and re- Some inadvertent consequences. J. Biomat. Appl., 10, sidual monomer from dental plate. J. Liquid Chromatogr. 121-135. Clin. Anal., 18, 613-626. Prince, H. N. and Rubino, J. R. (1984) Bioburden dynamics: Shintani, H. (1995) Formation and elution of toxic com- The viability of microorganisms on devices before and af- pounds from irradiated medical products. J. Biomater. ter sterilization. Med. Dev. Diagnos. Ind. July. Appl., 10, 23-58. Rose, R. M., Goldfarb, E. V., Ellis, E., and Crugnola, A. N. Shintani, H. (1995) Gamma-ray irradiation to medical de- (1984) Radiation sterilization and the wear rate of poly- vices. In SterllizationSystems, (Szyche, M. and Sharma, ethylene. J. Orthop. Res., 2, 393-400. C. P, eds.) 158-196, Technomic Publication, Basel, Shintani, H. (1979) Residue of ethylene oxide in artificial di- Switzerland. alysis devices. Jpn. Med. Instrumen., 49, 19-22. Shintani,H. (1996) Formation and residue of toxic com- Shintani, H. (1981) Residue of ethylene oxide and their re- pound in sterilized medical products. Rad. Phys. Chem., action compounds in medical devices. Jpn. Hyg. Sci. 27, 47, 139-148. 285-295. Shintani, H., ed., (1998) Sterilization Validation of Shintani, H. (1982) Residue of ethylene oxide in several Pharmaceutical Compounds and Medical Devices (in sorts of medical-use polymers. Jpn. Med. lnstrumen., 52, Japanese). Yakugyojiho, Tokyo, Japan. 140-144. Shintani, H. (2000) Production of degraded products upon Shintani, H. (1985) Determination of phthalic acid, mono- gamma-ray sterilization to medical devices (in Japanese). (2-ethylhexyl) phthalate and di- (2-ethylhexyl) phthalate Bokin in human plasma and in blood products. J. Chromatogr., Bobai, 28, 727-730. 337, 279-290. Shintani, H. (2001) Pretreatment and chromatographic Shintani, H. (1989) Analysis of carcinogen, 4,4'- analysis of phthalate esters and their biochemical behav- methylenedianiline (MDA) in gamma-ray and in ior in blood products. Chromatographia, 52, 721-726 autoclave-sterilized thermoplastic polyurethane. Shumnes, E. and Darby, T. (1984) Contact dermatitis due to Fresenius J. Anal. Chem., 333, 637-640. endotoxin in irradiated latex gloves. Contact Dermatitis, Shintani, H. (1989) Analysis of carcinogen, 4,4'- 10, 240-244. methylenedianiline, from thermosetting polyurethane dur- Vink. P. and Pleijsier, K. (1986) Aeration of ethylene oxide- ing sterilization. J. Anal. Toxicol., 13, 354-357. sterilized polymers. Biomaterials, 7,30-48. Shintani, H. (1990) Effects of gamma-ray irradiation on the change of characteristics of medical-use polyurethane. J.