CLINICAL MICROBIOLOGY REVIEWS, July 2008, p. 403–425 Vol. 21, No. 30893-8512/08/$08.00�0 doi:10.1128/CMR.00014-08Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Evidence-Based Biosafety: a Review of the Principles and Effectivenessof Microbiological Containment Measures

Tjeerd G. Kimman,1* Eric Smit,2 and Michel R. Klein1

Center for Infectious Disease Control1 and International Team,2 RIVM, National Institute of Public Health and the Environment,P.O. Box 1, 3720 BA Bilthoven, The Netherlands

INTRODUCTION .......................................................................................................................................................403DEVELOPMENT OF CONTAINMENT MEASURES: BRIEF HISTORIC OVERVIEW................................404PRINCIPLES AND METHODS OF BIOSAFETY.................................................................................................405BIOSAFETY MEASURES .........................................................................................................................................406

Risk Assessment......................................................................................................................................................406Biological Containment..........................................................................................................................................407

Viruses ..................................................................................................................................................................407Bacteria and protozoa ........................................................................................................................................407

Physical Containment ............................................................................................................................................409Categorization of microorganisms (non-GMOs)............................................................................................409Categorization of GMOs and definition of harmful gene products and microorganisms .......................409Laboratory design and primary and secondary containment.......................................................................410Categorization of biosafety containment levels ..............................................................................................410

APPROACHES FOR BIOSAFETY EVALUATION ...............................................................................................412Compliance with Procedures and Training.........................................................................................................413

EXPERIMENTAL AND OBSERVATIONAL DATA ON THE EFFECTIVENESS OF CONTAINMENTMEASURES.........................................................................................................................................................413

Do Single Devices and Procedures Function Effectively? .................................................................................413BSCs .....................................................................................................................................................................413Cell sorters...........................................................................................................................................................414Respiratory protection devices ..........................................................................................................................415

Does the Laboratory as a Whole Afford Effective Containment? ....................................................................415Are Laboratory Workers and the Environment Protected against Infection? ...............................................416

Reviews .................................................................................................................................................................416Surveys .................................................................................................................................................................418GMO-associated laboratory accidents .............................................................................................................419Accidents with risk category 4 organisms .......................................................................................................420

DISCUSSION ..............................................................................................................................................................420ACKNOWLEDGMENTS ...........................................................................................................................................422REFERENCES ............................................................................................................................................................422

INTRODUCTION

Work with pathogenic microorganisms and genetically mod-ified microorganisms (GMOs) requires precautions that guar-antee the safety of humans and the environment, includinglaboratory personnel, patients treated with GMOs, and otherpersons who could be exposed to these microorganisms. Dur-ing the past decades, responsible authorities and researchershave therefore developed regulations and guidelines that insome detail describe containment measures and working in-structions (Fig. 1). For GMOs such regulations and guidelinesappear to be largely derived from those developed for workingwith the natural, genetically unmodified pathogenic microor-ganisms from which these GMOs have been derived.

Despite containment measures and guidelines, laboratoryinfections, usually involving non-GMOs, occur more or less

commonly, suggesting that biosafety rules are not always ef-fective or complied with. The guidelines and instructions forworking with GMOs appear to be largely effective, as therehave been no major accidents with GMOs or with their unin-tended release. Nonetheless, despite such regulations and thelack of major accidents with GMOs, there appears to be con-tinuing concern about the health and safety of individuals andthe environment exposed to potentially hazardous GMOs (65).It has also been noted that the laws and regulations governingthe biotechnology world are outdated, are not comprehensive,and span too many agencies (65). Indeed, while the natures ofthe risks and the measures to handle these risks are largelyidentical for GMOs and non-GMOs, in many countries thereare different regulations for GMOs and non-GMOs. This maybe because the latter were presumed to carry greater risks forcausing ecological disturbances upon unintended release. Forexample, in The Netherlands, the Ministry of the Environmentoversees work with GMO, while the Ministry of Social Affairsoversees work with human pathogens. The different regula-tions and overseeing authorities may be confusing to workers

* Corresponding author. Mailing address: Center for Infectious Dis-ease Control, RIVM-National Institute of Public Health and the En-vironment, P.O. Box 1, 3720 BA Bilthoven, The Netherlands. Phone:31 (0)30 274 2330. Fax: 31 (0)6 46081730. E-mail: tg.kimman@rivm.nl.

403

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

in the field. Moreover, it is unknown to what extent specificfactors contribute to a safe biosafety practice. Thus, it is oftenunclear if and to what extent measures aimed at providingbiosafety are based on documented evidence of their effective-ness.

A central question in this study is whether containmentmeasures are effective and evidence based. One may argue thatthe evidence for the effectiveness of containment measures isat best indirect, i.e., based on the lack of many overt labora-tory-acquired infections (LAIs). In addition, we could questionwhether the criteria to judge effectiveness are sufficiently de-veloped. Indeed, the objectives of containment measures areoften not explicitly defined, and without (quantifiable) objec-tives, evaluation of effectiveness is difficult. Furthermore, infinding evidence for the effectiveness of biosafety measures, itis important to judge the quality of the evidence. For compar-ison, in evidence-based medicine, systematic reviews, hypoth-esis-driven controlled laboratory experiments, and prospectivestudies provide a higher quality of evidence than case reportsand expert opinion.

In this review we give a brief historical overview of the

development of the current biosafety practice, we will try toidentify which principles and methods appear to underlie it,and we will describe this current biosafety practice. We thenpresent an approach for evaluating the effectiveness of bio-safety measures to contain pathogenic microorganisms, andfinally we summarize experimental and observational data onthe effectiveness of containment measures. Our primary goal isto evaluate the evidence-based containment measures forGMOs. However, because such measures are largely based oncontainment measures for non-GMOs and because more data,although scarce, are available for non-GMOs, we also examineevidence-based measures to contain non-GMO pathogens.These data may be extrapolated to GMOs. We thereby hope tocontribute to a conceptual framework that helps in furtherdeveloping an evidence-based biosafety practice.

DEVELOPMENT OF CONTAINMENT MEASURES:BRIEF HISTORIC OVERVIEW

While Robert Koch had already developed some kind ofbiosafety cabinet (BSC), A. G. Wedum of the U.S. Biological

FIG. 1. Context of biosafety measures. Based on a risk assessment, wild-type biological agents and GMOs are assigned to one of four riskcategories. Work is subsequently performed under conditions that reflect increasing containment demands, i.e., BSL-1 to -4. Risks are containedby a set of measures employing biological and physical barriers and laboratory practices.

404 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

Research Laboratories at Fort Detrick, MD, can be regardedas one of the pioneers in developing biosafety measures afterthe Second World War. He evaluated the risks of handlinghazardous biological agents and developed practices, equip-ment, and facility safeguards for their control (113, 147, 148).Following his initial work, it is now regarded as conventionalwisdom that enclosure and ventilation of contaminated workareas are important factors in eliminating LAIs. Besides safemicrobiological techniques, primary barriers (safety equipmentand personal protective equipment) and secondary barriers(facility safeguards) are now regarded as vital elements ofcontainment measures.

It was recognized early that examining LAIs could be infor-mative about the risks involved with laboratory work. Severalcomprehensive reviews of LAIs have therefore been compiled(116, 131). These early examinations recognized that the pri-mary route of transmission of many of the causative agents wasby aerosol, and they led to the development of laminar-flowBSCs. Legislation and guidelines that were introduced over theyears have probably reduced but not eliminated the risk ofoccupational exposure to infectious agents.

With the growing ability to manipulate DNA in the mid1970s, there was also growing concern about the potentialhazards associated with recombinant DNA research and tech-nology (9). GMOs could display the intended properties, butcould also have unpredictable and undesirable features. Therestill appears to be debate about the health and safety of labo-ratory workers and animals, as well as the environment, ex-posed to potentially hazardous GMOs (65, 98). At the Asilo-mar Conference in 1975, general principles for dealing withpotential biohazards related to GMOs were drafted. It wassuggested that containment should be an essential consider-ation in the experimental design and that the effectiveness ofthe containment should match the estimated risk. Adjustmentof the level of precaution to the level of risk would preventinfection without unduly impeding operations.

The first edition of the National Institutes of Health (NIH)guidelines for research involving DNA molecules appeared in1976. Now, 3 decades later, there are a number of authoritativeinternational guidelines, instructions, and recommendationsfor the safe handling and manipulation of hazardous biologicalagents, including GMOs. In 1984, the U.S. NIH and Centersfor Disease Control and Prevention (CDC) produced the firstedition of a guidebook, called Biosafety in Microbiological andBiomedical Laboratories, that is now considered a major refer-ence text. The NIH/CDC and the WHO manuals are based onhistorical accounts of incidents with infectious microorganismsand extensive experience of experts working in this field, andthey have been developed and improved over the last 30 years(152, 156). Legislation has been implemented, for example, inEuropean Union and national regulations (145). In Europe,national authorities have based their regulations on directivesfrom the European Union, such as the directive on the pro-tection of workers from risks related to exposure to biologicalagents at work and the directive on the contained use ofGMOs. A summary of important guidelines and manuals isgiven in Table 1.

PRINCIPLES AND METHODS OF BIOSAFETY

In judging the effectiveness of measures intended to ensurebiosafety, it may be helpful to know which goals, principles,and methods of biosafety measures have been employed and toevaluate the scientific basis of their effectiveness. Because thenatures of risks are largely identical for GMOs and non-GMOs, containment measures to handle these risks are largelyidentical for both. In the past biosafety measures have evolvedstep by step, and usually based on expert knowledge and ex-perience, but without a unifying set of guiding principles. Ex-plicit guiding principles are therefore usually lacking in mostlegal regulations and scientific papers. We here attempt to

TABLE 1. Summary of guidelines and directives on recombinant DNA research

Institution Publication Yr issued(edition) Website or publication no.

CDC/NIH Biosafety in Microbiological and BiomedicalLaboratories

2007 www.cdc.gov/od/ohs/biosfty/bmbl5/bmbl5toc.htm

EEC/EC Directive on the protection of workers fromrisks related to exposure to biologicalagents at work

1990 90/679/EEC, 89/391/EEC, 2000/54/EC

Directive on the contained use of geneticallymodified microorganisms

1990 90/219/EEC, 98/81/EC, 2000/608/EC

AS/NZS Safety in Laboratories (part 3, MicrobiologicalAspects and Containment Facilities)

2002 2243.3

PHAC Laboratory Biosafety Guidelines 2004 (3rd) www.phac-aspc.gc.ca/publicat/lbg-ldmbl-04/index.html

WHO Laboratory Biosafety Manual 2004 (3rd) www.who.int/csr/resources/publications/biosafety/WHO_CDS_CSR_LYO_2004_11/en

VROM Intergrale versie van de Regeling genetischgemodificeerde organismen en Besluitgenetisch gemodificeerde organismen(in Dutch)

2004 www.vrom.nl/docs/milieu/regelingGGO_inclusiefbijlagen_1okt2003.pdf

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 405

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

draft a hierarchy of such guiding principles and methods thatmay be applicable to both GMOs and non-GMO, as follows.

(i) Risk assessment is the first and central step, and includeshazard recognition and identification, understanding of expo-sure potentials, frequency of occurrence, evaluation of worktasks and equipment, and assigning protective measures to thespecific tasks involved.

(ii) The second principle is biological containment; whereverpossible, risks for the workers and the environment may beminimized by reducing exposure potentials and their conse-quences by using attenuated microorganisms (GMO or non-GMO) that have reduced replicative capacity, infectivity, trans-missibility, and virulence.

(iii) Concentration and enclosure, while simple and oftenoverlooked, obviously constitute one of the most importantways to provide biosafety, by “locking up” microorganisms asmuch as possible, limiting the microbiological work load, andconcentrating it in as few as possible work sites. For example,if work associated with a harmful microorganism can be re-stricted to one BSC, there will be less risk than if this work isdone in two BSCs. Detection of microorganisms by nucleicacid multiplication carries less risk than detection by culture.

(iv) The fourth principle is exposure minimization; whenenclosure has to be disturbed, exposure can be minimized by aset of behaviors collectively known as safe microbiologicaltechniques, including an orderly and disciplined work fashion,wearing gloves, wearing a mask and eye covering, prevention ofaerosol and droplet generation, no mouth pipetting, preven-tion of skin disruptions, etc. These measures in particular pro-vide operator protection (121). In practice they often make useof special equipment providing physical containment. Labora-tory workers should be trained in these aspects.

(v) In physical containment, further protection of the oper-ator and the environment is provided by physical barriers thatprevent or minimize escape of microorganisms from the work-ing place and laboratory. These include a range of designrequirements, such as doors and locks, and physical safetyequipment, such as BSCs, isolators, air filtration, wastewatermanagement systems, etc. In addition to protecting the labo-ratory workers, these measures provide environmental protec-tion. Obviously such physical containment is combined withand complemented by the previous methods. Physical barriershave been designated primary and secondary barriers. Primarycontainment measures minimize occupational exposure of lab-oratory workers and thereby limit transmission of microorgan-isms from these workers to others. The secondary barriersprovide supplementary microbiological containment, serving

mainly to prevent the escape of infectious agents when a fail-ure in the primary barriers occurs.

(vi) Hazard minimization includes a set of activities to re-duce the consequences of exposure should it occur; these mayinclude the availability of emergency procedures, a contin-gency plan, and health and medical surveillance, but also vac-cination to reduce the consequences of inadvertent exposure(152).

In the following sections we further elaborate on some ofthese principles and methods and on the way in which theyhave been incorporated in legislation and regulations.

BIOSAFETY MEASURES

Risk Assessment

From a thorough risk assessment procedure, considering allpotentially harmful effects for humans and the environment,follows the risk classification of microorganisms and subse-quent containment measures that should be taken to managethese risks. Both the nature and scale of activities need to beconsidered to estimate the possibility of exposure of humansand the environment and the consequences of such exposure.Examples of risk assessment procedures can be found at http://www.hse.gov.uk/biosafety/gmo/acgm/ecrisk.htm and in Euro-pean Union Directive 90/219/EEC.

As indicated below, there are four risk categories for haz-ardous biological agents and four containment levels. Workwith non-GMO microorganisms is usually done at the corre-sponding containment level. It is evident that a risk assessmentfor working with noncharacterized pathogens in a clinical lab-oratory may involve uncertainties. Through manipulationGMOs may acquire pathogenic properties that are unexpectedand/or not well characterized or understood, necessitating ahigher containment level than for work with the natural mi-croorganism from which the GMO has been derived, or addi-tional measures (see, for example, reference 145). Therefore,the nature of recombinant DNA sequences, vectors, and re-cipient organisms needs to be carefully evaluated, as well asany potential biohazard associated with particular experimen-tal settings. It is particularly important to address whether ornot genetic modification affects cell tropism, host range, viru-lence, or susceptibility to antibiotics or other effective treat-ments. Some considerations regarding the risk assessment andcategorization of GMO activities are given in Table 2.

TABLE 2. Summary of considerations in the risk assessment for GMOs

Category Factor(s) to be considered

Recipient microorganisms ....................Virulence, transmissibility, host range, susceptibility to antivirals or antibiotics, availability of prophylaxis,control and treatment

Vectors ....................................................Replicative capacity, integration into host genomeInsert or donor sequences ....................Toxicity, biological properties, replicative capacity, properties (known/unknown), gene-gene and gene-

environment interactionsActivities .................................................Scale, animal experiments, transportHost factors ............................................ImmunodeficienciesPopulation factors..................................(Vaccine-derived) immunity

406 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

Biological Containment

A subsequent step in the selection of measures to ensurebiosafety is to minimize biological hazards associated with thework by employing host microorganisms with a reduced hostrange, strains with natural or genetically modified characteris-tics that diminish their invading capacity or virulence, self-inactivating vectors, etc. Thus, employing biological contain-ment is not restricted to work with GMOs. Natural pathogenicmicroorganisms may be replaced by less pathogenic microor-ganisms; for example, Mycobacterium tuberculosis may be re-placed with nonvirulent mycobacteria, such as Mycobacteriumsmegmatis (27). Approaches for acquiring biological contain-ment by genetic modification are given in Table 3.

Viruses. Examples of biological containment include the useof highly modified vaccinia virus Ankara, which has a signifi-cantly reduced pathogenicity, and the use of avian poxviruses,including canarypox and fowlpox, that have a restricted hostrange and do not replicate in mammals (93, 108). These viruseshave a better safety profile than the classical vaccinia virus.

Biological containment for retroviral vectors has been ob-tained by providing gene products that are required for theproduction of progeny viruses, i.e., gag, pol, and env, in trans bypackaging cell lines that stably express these trans-acting func-tions. Formation of replication-competent human immunode-ficiency virus (HIV) is excluded when the env gene is missingfrom the packaging plasmid. Furthermore, the range of hostspecies that can be transduced with such vectors can be ma-nipulated by using particular env genes or by pseudotyping withenvelope proteins of other viruses (29, 39, 118). Various en-velope proteins are associated not only with a varying hostrange but also with varying stability and intrinsic toxicity. Aconcern has been that recombination events may give rise toreplication-competent lentiviruses, but further modificationshave been made to reduce this opportunity. These include theuse of different transcriptional units, further attenuation bydeleting nonessential genes, and self-inactivating vectors (17,29). Self-inactivating vectors contain an inactive long terminalrepeat, resulting in a lack of promoter activity and preventingpotential transcriptional activation of (onco)genes downstreamof the integration site. These vectors cannot be rescued bywild-type HIV. No replication-competent lentiviruses havebeen detected when such systems are used (29, 37). Altogether,the risk associated with the newly developed lentivectors istherefore minimal, while they have distinct advantages such asstable integration in nondividing and dividing cells, long-termexpression of the transgene, and absence of an immune re-

sponse. However, although viral infection using state-of-the artlentivectors is very unlikely, their high transduction efficiencymay increase the risk of accidental exposure of lab workers,leading to a positive anti-p24 HIV antibody response and ac-cidental transduction of potentially hazardous genes. Contacttransmission thus should still be avoided.

Similarly, replication-deficient adenoviruses have been de-veloped using a viral DNA vector and a helper cell line that hasbeen stably transfected with the E1 region (E1A and E1B) ofthe adenoviral genome. Vectors prepared from the cell linelack the E1 region and remain replication defective (72).

Safe replication-deficient herpesvirus vectors have been de-veloped by deleting the viral glycoprotein gD. Envelope glyco-protein gD is essential for virus entry but is not required forsubsequent steps in the viral replication cycle. Phenotypicallycomplemented gD null mutants can infect cells and can spread,both in vitro and in vivo, by direct cell-to-cell transmission.However, progeny virions released by the infected cells arenoninfectious because they lack gD (109). Thus, an accidentalinfection remains restricted to a single round of replication.The same principle has been used for other viruses, for exam-ple, respiratory syncytial virus. Other ways to generate safeherpesvirus vectors are deletion of genes that are essential forvirulence in vivo but that are nonessential in vitro or deletionof immediate-early genes that activate early and late geneexpression and subsequent propagation of the crippled virus incomplementing cell lines (68, 75). Replication-incompetentherpesvirus vectors have also been generated by the use ofcosmid DNAs to provide the necessary viral gene products forpropagation of defective viruses or amplicons. These ap-proaches reduce the opportunities for generation of replica-tion-competent viruses through recombination.

Safe vectors have been developed from the nonpathogenicadeno-associated virus. The vector lacks all viral genes andrequires coinfection with a helper adenovirus or a helper-freepackaging system (24, 120).

A high level of biological containment is also acquired byusing the Autographa californica nuclear polyhedrosis virus, amember of the baculovirus family. These viruses normally rep-licate in insect cells but not in mammalian cells. Furthermore,by deleting the nonessential polyhedron gene, the virus be-comes noninfectious for its natural host (106).

Bacteria and protozoa. The classical example of a biologi-cally contained bacterium is Escherichia coli K-12. E. coli K-12is a debilitated strain that does not normally colonize thehuman intestine. The strain survives poorly in the environment

TABLE 3. Principles and methods of establishing biological containment

Principle Method Example(s)

Attenuation Natural or genetically modified deletion ofvirulence genes

Modified vaccinia virus Ankara, herpesvirus vectors,E. coli K-12, Salmonella aro mutants, Vibrio ctxmutants, Lactococcus lactis thyA mutant

Host range restriction Natural host-restricted viruses Canarypox virus, fowlpox virus, baculovirusHost range alteration Ecotropic packaging cell lines, pseudotyping Retroviral vectorsUse of replication-defective vectors Deletion and provision of essential gene products

in transHerpesvirus vectors, alphavirus vectors, retroviral

vectors, adenovirus vectors, adeno-associatedvirus vectors, respiratory syncytial virus, lentivirusvectors, Salmonella aro mutants, E. coli K-12

Prevention of gene transfer Expression of suicide functions E. coli relF

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 407

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

and has a history of safe commercial use. E. coli K-12 is con-sidered an enfeebled organism as a result of being maintainedin the laboratory environment for over 70 years (154). E. coliK-12 is defective in at least three cell wall characteristics. First,the outer membrane has a defective lipopolysaccharide corewhich affects the attachment of the O-antigen polysaccharideside chains (26). Second, it does not have the type of glycocalyxrequired for attachment to the mucosal surface of the humancolon (S. Edberg, unpublished report, U.S. EnvironmentalProtection Agency, 1991) as a result of the altered O-antigenproperties noted above. Finally, K-12 strains do not appear toexpress capsular (K) antigens, which are heat-labile polysac-charides important for colonization and virulence. K-12 thus isnot able to recognize and adhere to the mucosal surface ofcolonic cells (26). The normal flora in residence in the colonthus can easily exclude K-12. Furthermore K-12 lacks othervirulence factors (26, 46; Edberg, unpublished report).

Several approaches have been used to develop safe bacterialvectors for use as vaccines or for gene or protein delivery, inparticular, strains of Salmonella, Shigella, Listeria, Mycobacte-rium, Vibrio, and lactic acid bacteria (71, 137). When develop-ing vector vaccines, the challenge is to develop strains that arewell tolerated by the recipient host and no longer persist in theenvironment, yet still induce protective immune responses.This is not always achieved, however (32, 141). Elucidation ofbiosynthetic pathways has led to the development of Salmo-nella vectors that were attenuated by disruption of genes en-coding metabolic functions or genes located in a pathogenicityisland (Salmonella pathogenicity island 2, which encodes a typeIII secretion system). Salmonella pathogenicity island 2 is re-quired for survival and growth within macrophages (67). Forexample, attenuating mutations in Salmonella strains includedmsbB, galE, via, rpoS�, aroCD, htrA, cya, crp, cdt, asd, phoPQ,purB, sifA, and ssaV (33, 67, 84, 90). The best-characterized liveattenuated salmonellae have mutations in the prechorismatepathway. These are the so-called aro mutants, which are de-fective in the production of chorismate, which is essential inthe synthesis of aromatic compounds (90). It is evident thatstrains carrying different mutations differ in properties such asinvasiveness and survival. To reduce the possibility of reversionto virulence, strains carrying at least two attenuating distantlylocated mutations have been produced (67, 141). Salmonellaenterica serovar Typhimurium VNP20009 was developed todeliver potential therapeutic proteins to tumor sites. It wascreated by chromosomal deletion of two genes, purI (purinebiosynthesis) and msbB (lipopolysaccharide biosynthesis) andwas attenuated at least 10,000-fold in mice compared with theparental wild-type strain (84).

A promising Listeria vector vaccine is an L. monocytogenesauxotrophic mutant with deletions in alanine racemase (daI)and D-amino acid aminotransferase (dat) genes, two genesrequired for the biosynthesis of bacterial cell walls. The strainwas highly attenuated in mice (142). The strain requires D-alanine to grow and survive. Another L. monocytogenes candi-date vaccine strain (LH1169) contained deletions in actA andplcB, genes that are necessary for cell-to-cell spread and escapefrom secondary vacuoles, respectively (2). Deleting lecithinaseactivity in L. monocytogenes also results in inhibited cell-to-cellspreading (31).

The virulence of Vibrio cholerae is due mainly to the expres-

sion of cholera toxin (CT). Hence, strategies for attenuating V.cholerae for use as an expression vector for heterologous an-tigens have been to engineer mutants in which the CT gene(ctx) or the CT genetic element has been partially or com-pletely deleted (71).

Stable mutant strains of Bacillus licheniformis, an industriallyexploited species, were obtained by introducing defined dele-tions in recA and/or an essential sporulation gene (spoIV).These strains are totally asporogenous and severely affected inDNA repair, and they therefore are UV hypersensitive. Inliquid media these strains grow equally well as the wild type.Hence, such genes appear to be suitable disruption targets forachieving biological containment (96).

Construction of a genetically modified Lactococcus lactisstrain for intestinal delivery of human interleukin-10 (IL-10)employed a biological containment system by replacing thethymidylate synthase gene thyA with a synthetic human IL-10gene. When deprived of thymidine or thymine, the viability ofthe strain dropped considerably, preventing its survival in theenvironment. Transgene escape through acquisition of an in-tact thyA gene is very unlikely and would recombine the trans-gene out of the genome. The system was validated in vivo inpigs (136) and was used in a gene therapy study (13).

Further improvement in enhancing the safety profile ofbacterial vectors for gene transfer can be achieved by re-moving undesirable properties of plasmids, such as a pro-karyote origin of replication and antibiotic resistancemarkers. These elements could lead to dissemination ofprokaryotic replicative recombinant DNA. Darquet et al.(28) therefore developed so-called minicircles, which aresupercoiled DNA molecules that lack such elements andcontain only an expression cassette carrying the gene ofinterest. Furthermore, efficient suicide functions have beendeveloped to ensure biological containment of bacteria (69,128). Such systems achieve their goals when the GMOsself-destruct by expression of killing genes after fulfillingtheir jobs. Suicide systems are based on “lethal genes” thatare triggered by preprogrammed conditions. Such systems,however, appear to differ in efficiency of the suicide func-tion, and the less efficient ones may lead to selection ofmutants that have lost their suicide function. However, nosystem can provide complete efficiency. One efficient systemwas based on the lethal E. coli relF gene, which prevents thetransfer of plasmids to wild-type bacteria (69).

To enhance the safety of genetically modified yeast (Saccha-romyces cerevisiae), genes encoding bacterial toxins have beenused for containment control. Expression of the E. coli relEtoxin gene was highly toxic to yeast cells, and this could becounteracted by expression of the relB gene (73).

In conclusion, we believe that the mechanisms underlyingbiological containment are usually well examined and under-stood. Evidence for their effectiveness is usually available, al-though only seldom in quantitative terms of infectivity andtransmission. It is also important to note that not all GMOs areby definition biologically contained. They have reduced trans-missibility or virulence only upon disruption of virulence fac-tors. For example, very virulent recombinant influenza virusesand herpesviruses have been regenerated (30a, 47).

408 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

Physical Containment

Categorization of microorganisms (non-GMOs). Central tobiosafety programs is the concept of universal precautions(16). For that purpose, microorganisms are categorized intofour risk categories on the basis of a risk analysis. Subsequentrisk containment is focused not on specific infectious agentsbut on standard practices for handling infectious material thatwill prevent the transmission of all pathogens of that risk cat-egory. It is important to note that the principles, guidelines,and recommendations are basically the same for natural patho-gens and GMOs. However, as would be logical, this conclusiondid not lead to a single set of guidelines and recommendationsfor GMOs and non-GMOs, and thus there is some redundancyin guidelines and regulations.

The four categories for biological agents are based on theirrelative risk to laboratory workers and the community. In general,the following factors are considered in classifying biologicalagents: (i) the virulence of the biological agent or the severity ofdisease (in humans) (18), (ii) the mode of transmission (spread inthe community and host range), (iii) the availability of effectivepreventive measures (e.g., vaccines), and (iv) the availability ofeffective treatment (e.g., antibiotics or antiviral drugs). The haz-ard of the infectious (non-GMO) agent increases from risk group1, consisting of microorganisms not associated with disease, torisk group 4. Risk group 4 microorganisms can cause seriousdisease and can be readily transmitted, and effective treatmentsare usually not available (123, 152, 156). However, there aredifferences in the exact definitions as used by certain countriesand/or organizations (such as NIH/CDC, WHO, and EuropeanUnion) (Table 4), which result in differences in the exact listingsof biological agents in each risk category (43; http://www.absa.org/XriskgroupsX/index.html). The main difference between the

NIH/CDC classification and the WHO classification is that thelatter includes hazards to animals and the environment. Anotherdifference is that for risk group 3, the NIH/CDC states that “. . .therapeutic interventions may be available,” whereas the WHOand European Union state that “. . . effective treatment and pre-ventive measures are available.” There also are differences in thedescription of transmission properties between the different clas-sifications. In assigning an agent to a risk group, one must takeinto account that there are in all groups of microorganisms nat-urally occurring strains that vary in virulence and that may thusneed a higher or lower level of containment (43). In general,regulators deal with this concept by taking the highest level ofvirulence into account.

There has been debate about the classification of particularmicroorganisms, in particular the Flaviviridae, variola virus,avian influenza A/H5N1 virus, and extremely drug-resistant M.tuberculosis strains, especially as to whether they should becategorized as category 3 or 4 biological agents (1, 20, 25, 40,70, 97, 143). Because vaccination was stopped in the 1970s,variola viruses are now classified as category 4 biologicalagents. Although avian influenza A/H5N1 virus strains initiallywere classified as risk category 4 biological agents, susceptibil-ity to antiviral drugs and the availability of effective vaccinesmay downgrade them to category 3 for further studies. Ex-tremely drug-resistant M. tuberculosis strains (25) should beregarded as risk category 4 biological agent. Of note, to date nomicroorganisms other than viruses have been classified in cat-egory 4.

Categorization of GMOs and definition of harmful geneproducts and microorganisms. As for non-GMOs, GMOs areassigned to specific risk categories based on the risk assesse-ment. For GMOs biosafety containment levels are assigned

TABLE 4. Classifications of infectious (non-GMO) agents into risk groups by NIH/CDC, WHO, and the European Uniona

Riskgroup

Classification according to:

NIH/CDC WHO European Union

1 Agents that are not associated withdisease in healthy adult humans

A microorganism that is unlikely to causehuman or animal disease

A microorganism that is unlikely to cause humandisease

2 Agents that are associated withhumans disease which is rarelyserious and for which preventiveor therapeutic interventions areoften available

A pathogen that can cause human oranimal disease but is unlikely to be aserious hazard to laboratory workers,the community, livestock, or theenvironment; laboratory exposures maycause serious infection, but effectivetreatment and preventive measures areavailable and the risk of spread ofinfection is limited

A pathogen that can cause human disease but isunlikely to be a serious hazard to laboratoryworkers; the risk of spread of infection islimited, and effective treatment and preventivemeasures are available

3 Agents that are associated withserious or lethal human diseasefor which preventive ortherapeutic interventions may beavailable

A pathogen that causes serious human oranimal disease but does not ordinarilyspread from one infected individual toanother; effective treatment andpreventive measures are available

A pathogen that causes serious human diseaseand poses a serious hazard to laboratoryworkers; it is likely to spread from oneinfected individual to another; effectivetreatment and preventive measures areavailable

4 Agents that are likely to causeserious or lethal human diseasefor which preventive ortherapeutic interventions are notusually available

A pathogen that usually causes serioushuman or animal disease and that canbe readily transmitted from oneindividual to another, directly orindirectly; effective treatment andpreventive measures are not usuallyavailable

A pathogen that usually causes serious humandisease and that can be readily transmitted inthe population; effective treatment andpreventive measures are not usually available

a Based on data from references 19, 145, 152, and 156.

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 409

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

depending on the risk category of the donor organism, unlessthe modification may result in a higher or unknown risk.Therefore, the nature and function of insert sequences and theproperties of acceptor microorganisms are considered. It isremarkable that definitions of properties of GMOs are some-times quite strict (for example, the toxicities of vertebratetoxins, which are expressed as 50% lethal dose/body weight),while other properties of harmful gene products are not exactlydefined. In the regulations, both the nature and level of viru-lence and transmissibility of GMOs are not well defined, atleast not in quantitative terms (such as the basic reproductionratio [R0] [14]). Sometimes host range is taken into account(for example, for baculovirus, ecotropic murine retroviruses,and papillomaviruses when they are used in nonpermissivehost/vector systems). Properties may be unknown, such as thecapacity of microbial DNA to integrate into the host genome(as, for example, for HIV) and the availability of vector organ-isms in the environment and therefore the possibility of per-sistence in the environment. A point for consideration is there-fore the possibility of monitoring replication and survival of theGMO. In case of scientific uncertainty, the precaution princi-ple is leading, resulting in higher categorization of the GMO oradditional measures on a case-by-case base. Although the pos-sibility of microbial transmission is taken into account, thepossibility that transmission is reduced by herd immunity, ei-ther vaccine derived or not, is not mentioned explicitly bylegislators. The assessment should include whether propertiesof inserted sequences can be expressed in the background ofthe host organism. Thus, gene-gene and gene-environmentinteractions should always be considered. As an example, IL-4is not normally considered a harmful gene product, but whenexpressed by a murine ectromeliavirus, it drastically enhancedthe virulence of this virus by inducing changes in cytotoxicT-cell function (58). Importantly, the regulations, being notcompletely detailed, largely function as a framework that mustbe elaborated by researchers and laboratory directors in theirrisk assessment, in the risk categorization of GMOs, and in theimplementation of procedures and regulations in their labora-tories (79). While the considerations in the risk assessment andsubsequent risk categorization may show differences forGMOs and non-GMOs, many of the subsequent containmentprocedures are similar.

Laboratory design and primary and secondary containment.Phillips and Runkle (113) describe some principles underlyinglaboratory design. They introduce two concepts in designinglaboratories. The first is the concept of primary and secondarybarriers as described above, and the second provides the de-signers a logical division of major functional zones within alaboratory building. (Note that although elements of primaryand secondary barriers are clearly recognizable, they are notmentioned as such in the European and Dutch regulations.)Phillips and Runkle (113) identify five functional zones in thefacility (clean and transition, research area, animal holding,laboratory support, and engineering support). Primary con-tainment measures minimize occupational exposure of labora-tory workers. In addition to strict adherence to good microbi-ological practice, the primary containment barrier includephysical separation of the biohazardous agent from the labo-ratory worker using closed vessels, personal protective equip-ment (e.g., gloves or full-body suits) and additional equipment

(e.g., BSCs, enclosed centrifuge containers, or pipetting aids).The secondary barriers provide supplementary containment,serving mainly to protect other facility employees and to pre-vent the escape of infectious agents from the laboratory if andwhen a failure occurs in the primary barriers. These provide aseparation between potentially contaminated areas in thebuilding and the outside community. These measures maycomprise special procedures (e.g., validated decontaminationmethods, training of personnel, strictly controlled access zones,interlocked doors, etc.) and special engineering and facilitydesign features (e.g., decontamination equipment, showers,autoclaves, dedicated air handling system with filters, etc).

Secondary containment is very strict in high-containmentlaboratories. The high-containment laboratories (biosafetylevel 3 [BSL-3] and BSL-4) are airtight and have airlocks anda unidirectional airflow so that potentially contaminated air iskept inside. Thus, BSL-3 and -4 laboratories need to be neg-atively pressured, resulting in an airflow from adjacent areasinto the laboratory and a filtered exhaust airflow outside thebuilding without recirculation. Ten to 12 air changes per hourhave been recommended, which removes approximately 99%of airborne particles in 23 min (92). Release of air into theenvironment is possible only through HEPA (high-efficiencyparticulate air) filters (rated 99.99% efficient with particles 0.3�m in diameter and larger) or ULPA (ultra-low-penetrationair) filters (rated 99.999% efficient with particles 0.12 �m indiameter). Filters are used for BSCs, autoclaves, incinerators,chemical decontamination showers, etc. In addition, HEPAfilters are used to provide clean air to laboratory workers infull-body suits.

Categorization of biosafety containment levels. Biosafetycontainment levels have been categorized in a range from 1 to4. As indicated above, the containment levels are assigneddepending on the risk group of the microorganism (GMO ornon-GMO) and the scale and nature of activities. Importantly,the BSL designations are based on a composite of the designfeatures, construction, containment facilities, equipment, andoperational procedures required for working with agents fromthe various risk groups (152). It is important to stress thatalthough handling microorganisms of a certain risk group usu-ally requires working at the accompanying BSL, a risk assess-ment should be made to take other specific factors into con-sideration. For example, particular experiments may generatehigh-concentration aerosols, requiring a higher degree ofsafety. Thus, professional evaluation, based on personal re-sponsibility, should always guide the BSL for the specific work(152). Sewell (131) formulated some broad recommendations.BSL-1 is recommended for teaching activities with agents thatare not associated with disease. BSL-2 practices are used indiagnostic laboratories that manipulate agents that are nottransmitted via aerosols (e.g., hepatitis B virus [HBV], HIV,enteric pathogens, and staphylococci). BSL-3 is recommendedwhen working with agents that are highly infectious and aretransmitted via aerosols (e.g., M. tuberculosis, Brucella spp.,and Coccidioides immitis) and for large-scale work with BSL-2agents. BSL-4 practices are required when working with un-usual agents that cause life-threatening infections for which notreatment is available.

Depending on country and/or regulation authority, there aredifferences between the exact requirements for each of the

410 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

four containment levels, not to mention the sometimes con-fusing differences in nomenclature for the (high-)containmentlaboratories. Nulens and Voss (104) reviewed the basic prac-tice, equipment, and facilities necessary for each of the BSL,based on European Union Directive 98/81 and the WHO bio-safety manual. This information is summarized in a generalway in Table 5. A comprehensive listing of all requirements

and equipment necessary for the four BSLs is presented inTable 6. This list is adapted from the WHO biosafety manual(152).

At BSL-1, safety is achieved mainly by applying good micro-biological techniques. To achieve a higher degree of contain-ment and thus a higher degree of protection against LAIs, thenumber of requirements increases up to the maximum con-

TABLE 5. BSLs, practices, and safety equipment necessary for the four levels of microbial containmenta

BSL Practices Safety equipment Facilities

1 Standard microbiological practices None required Open bench top and sinkb

2 BSL-1 practices plus biosafetywarning, biosafety manual,waste decontamination, medicalsurveillance policy

Class I or II BSC, laboratory coats, gloves,face protection; face mask optional

BSL-1 plus autoclave

3 BSL-2 practices plus controlledaccess, decontamination ofwaste and laboratory clothing,baseline serum sample

Class I or IIc BSC, protective clothing,gloves; respiratory protection optional

BSL-2 plus physical separation from corridors,self-closing double doors, no airrecirculation, negative airflow in laboratory,HEPA-filtered air outletc

4 BSL-3 practices plus clothingchange before entering, showeron exit, all materialdecontaminated on exit

Class III BSC or class I or II BSC incombination with full-body, air-suppliedpositive-pressure suits

BSL-3 plus separated building or isolatedzone, dedicated supply, vacuum, anddecontamination systems plus air filtration

a Adapted from reference 152 with permission of the publisher and based on EU Directive 98/81 and reference 104.b According to EU Directive 98/81, an autoclave is required for ML-I.c According to EU Directive 98/81.

TABLE 6. Requirements for equipment and design for laboratories of different BSLsa

Design factor or equipmentRequirement at BSL:

1 2 3 4

Isolation of laboratory No No Yes Yes

Rooms sealable for decontamination No No Yes Yes

VentilationInward airflow No Desirable Yes YesControlled ventilation system No Desirable Yes NoHEPA filtered exhaust No No Yesb/noc Yes

Double-door entry No No Yes Yes

Airlock No No No Yes

Airlock with shower No No No Yes

Anteroom No No Yes

Anteroom with shower No No Yes/nod No

Effluent treatment No No Yes/nod Yes

AutoclaveOn site Yesb/no Yesb/desirable Yes YesIn laboratory No No Desirable YesDouble ended No No Desirable Yes

BSCs No Yesb/desirable Yes Yes

Personnel safety monitoring capability No No Desirable Yes

a Adapted from reference 152 with permission of the publisher.b According to Dutch regulations (145).c Depending on location of exhaust.d Depending on agents used.

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 411

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

tainment at BSL-4. Besides equipment, there are several codesof practice for laboratory access, personal protection, andworking procedures. Scientific evidence for the efficiency ofthese measures is scarce, and most of these measures havebeen developed based on a long history of microbiologicalpractice and common sense. Table 7 describes the codes ofpractice regarding access and personal protection for BSL-1,-2, and -3 laboratories.

In The Netherlands, BSL-1 to -4 are called MLI to -IV forwork with GMOs. These are based on the European UnionDirectives 90/219 and 98/81 and are implemented in nationalregulations (145). Many technical requirements and access andpersonal protection rules are similar, but a major difference isthat according to Dutch regulations, at the MLII level (BSL-2)a class II BSC is not optional but is required. Another differ-ence is that the Dutch regulations provide more detail withrespect to procedures. For instance, it is stated, “prepare yourwork carefully limiting the necessarily movement from oneplace to another during the microbiological work. Collect allmaterial and equipment before you start and arrange them inan orderly fashion.”

The first BSL-4 high-containment laboratories were built inthe 1970s. Until then, researchers had handled extremely haz-ardous biological agents in so-called glove boxes: hermeticallysealed, transparent cabinets fitted with rubber gloves, compat-ible with a class III BSC. To day, most BSL-4 high-containmentlaboratories operate as suit laboratories, where researcherswear full-body positive-pressure (“space”) suits (156; Euro-

pean Union Directives 90/218/EC, 98/81/EC, and 2000/54/EC).Key features of BSL-4 high-containment laboratories are thesafeguards to prevent failure and faults of containment systemsand measures. There is thus redundancy of critical systems andbiosafety procedures.

Although the objectives that the legislators want to reach byusing containment levels and procedures are not always explic-itly mentioned, one could deduce some guiding objectives, aswe have summarized in Table 8.

APPROACHES FOR BIOSAFETY EVALUATION

The scientific literature on evaluation of the effectiveness ofbiosafety measures is very scarce and does not provide a con-sensus approach. The effectiveness of biosafety measures maybe evaluated by different approaches and at different levels. InTable 9 we have given a simple classification of approachesthat collectively may provide guidance in evaluation activities.This table presents questions and purposes of single evaluationactivities. A first level of evaluation may be directed at mea-suring the effectiveness of single containment equipment andprocedures, such as the filtering capacity of face masks andBSCs under experimental circumstances. Such an evaluationcould be directed at physical or, preferably, microbiologicalcriteria. Subsequently these single apparatuses and proceduresshould be evaluated during practical work. It is evident that bytaking this step from experimental challenge to practical work,unforeseen circumstances that may occur during practical work

TABLE 7. Codes of practice regarding access and personal protection for BSL-1, -2, and -3 laboratoriesa

BSL(s) Access Personal protection

1 and 2 The international biohazard sign must be displayed on doorsof laboratories in which microorganisms of risk group 2 orhigher are handled; only authorized personnel are allowedto enter the laboratories; laboratory doors should be keptclosed; children under age 16 are not allowed to enter;access to animal houses should be specially authorized;animals not involved in the work are not allowed; signsprohibiting smoking, drinking, and eating should bedisplayed in and outside the laboratory (not required insome countries, including The Netherlands)

Laboratory coveralls, gowns, or uniforms must be worn in thelaboratory; appropriate gloves must be worn during all workinvolving contact with infected material, and after usegloves should removed aseptically and hands must bewashed; personnel must wash their hands after handlinginfectious material or animals and before they leave thelaboratory; safety glasses or face shields must worn when itis necessary to protect the eyes and face from splashes,impacting objects, or artificial UV radiation; it is prohibitedto wear protective clothing outside the laboratory; open-toefootwear should not be worn; eating, drinking, applyingcosmetics, and handling contact lenses in the laboratory areprohibited; storing human foods or drinks in the laboratoryis prohibited; protective clothing should not be stored inthe same lockers as street clothing

3 The international biohazard warning symbol and signdisplayed on the laboratory access door must identify thebiosafety level and the name of the laboratory supervisorwho controls access and must indicate any specialconditions for entry into the area (e.g., immunization)

Laboratory protective clothing must be of the type with solid-front or wrap-around gowns, scrub suits, coveralls, headcovering, and, where appropriate, shoe covers or dedicatedshoes; front-buttoned standard laboratory coats areunsuitable, as are sleeves that do not fully cover theforearms; laboratory protective clothing must not be wornoutside the laboratory, and it must be decontaminatedbefore it is laundered; removal of street clothing andchange into dedicated laboratory clothing may be warrantedwhen working with certain agents (agricultural or zoonoticagents);open manipulations of all potentially infectiousmaterial must be conducted within a BSC or other primarycontainment device; respiratory protective equipment maybe necessary for some laboratory procedures or whenworking with animals infected with certain pathogens

a Based on data from reference 152. Note that for BSL-3 laboratories the codes of practice for BSL-1 and -2 apply as well, except when modified as described in thetable.

412 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

may be detected. For example, turbulences caused by move-ments of personnel during practical work may lower the pro-tection afforded by BSCs. Air leakage may occur along respi-ratory masks during work. Masks may not fit perfectly.

A subsequent level of evaluation would be the laboratory asa whole, including its design and construction, the equipment,and working instructions. Again, such an evaluation can bedone experimentally during a validation of the laboratory pro-cess or actually in a working laboratory setting. An experimen-tal approach may use the deliberate release of indicator par-ticles or model microorganisms. Evaluation of laboratorysafety under field circumstances may include analysis of envi-ronmental samples taken inside and near the laboratory. Inthis case the effectiveness of biosafety measures and workinginstructions are actually evaluated during practical work andincludes compliance of workers with working instructions,their experience and training, unintentional incidents, and ef-ficacy of containment measures.

Finally, one may evaluate the effectiveness of measures atthe clinical-epidemiological level, examining the overall effec-tiveness of measures in their capacity to prevent infection oflaboratory workers and others. It is evident that laboratoryworkers play a central role in such an evaluation, as they arethe persons who both are at high risk and may pass infectionsto others. Such epidemiological studies may follow a passive oran active searching approach. Clearly, for ethical reasons thislevel of evaluation is usually not suitable for an experimentalapproach.

Compliance with Procedures and Training

We consider that optimizing the training of personnel andmonitoring their compliance with procedures are important, asthe best biosafety measures are only as good as the participa-tion and discipline of laboratory workers themselves. In manyreports, extensive training of laboratory workers and theproper execution of guidelines are mentioned as one of themost important measures to prevent incidents (85). However,poor compliance has also been reported (45, 144). In Argen-tina, incidents with pathogenic microorganisms were reducedafter setting up a training program and providing protocols(85). The importance of training was also emphasized by ex-periences during the outbreak of severe acute respiratory syn-drome (SARS) in 2003 (81, 100). Laboratory escapes of thevirus from BSL-3 laboratories occurred in Singapore, Taiwan,and Beijing because of breaches in good laboratory practicerather than failure of the facilities. Extensive contaminationoccurred because gloves were inappropriately worn and con-taminated surfaces were not disinfected (81).

EXPERIMENTAL AND OBSERVATIONAL DATA ON THEEFFECTIVENESS OF CONTAINMENT MEASURES

Do Single Devices and Procedures Function Effectively?

BSCs. Because most infections in the laboratory occur viaaerosols and infected material and surfaces (123), equipmentdirected at minimizing airborne infections has received most

TABLE 8. Objectives of microbiological containment rulesa

Objective(s)Achievement at BSL:

1 2 3 4

Reduction of direct and oro-fecal transmission of nonpathogenic microorganisms to labpersonnel, reduction of transmission of nonpathogenic microorganisms outside thelaboratory, general hygiene

� � �

Reduction oro-fecal transmission of enteric pathogens to lab personnel � �Reduction of airborne transmission of pathogenic microorganisms to lab personnel � ��Reduction of transmission of pathogenic microorganisms outside the laboratory by

direct contact, environmental spread, and airborne spread� ��

Strict prevention of transmission of very virulent microorganisms to lab personnel ���Strict prevention of transmission of very virulent microorganisms outside the

laboratory, calamity procedures���

a This table specifies objectives that may be achieved by containment measures at BSL-1 to -4, as specified by EU 90/679 and VROM (145). At increasing safety levels,there are additional demands regarding equipment and procedures, while those of the lower levels are maintained. At BSL-1 equipment and procedures are directedat providing general hygiene, which probably mainly diminish infection by the oro-fecal route. At BSL-2 a BSC class II is optional or required (as in The Netherlands),providing mainly respiratory protection to the laboratory workers; environmental spread is diminished by keeping windows closed. At BSL-3 further environmentalprotection is provided by a sluice, disinfection facilities, and a unidirectional HEPA-filtered airflow. At BSL-4 all procedures and equipment are directed towardpreventing any microbial transmission. �, adequate; ��, good; ���, excellent.

TABLE 9. Approaches for biosafety evaluationa

Category Exptl approach Practical conditions

Equipment and procedures Do single devices and procedures function effectivelyupon exptl challenge with particles or modelmicroorganisms?

Do single devices and procedures function effectivelyduring practical work?

Laboratory Does the laboratory as a whole afford effectivecontainment upon exptl challenge with particles ormodel microorganisms?

Does the laboratory as a whole afford effectivecontainment during practical work?

Laboratory workers andenvironment

NA Are laboratory workers and the environmentprotected against infection?

a Classification of approaches to evaluate containment measures, giving examples of the major questions and purposes of the evaluation steps. NA, not applicable.

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 413

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

attention. There are three classes of BSCs with different levelsof protection, i.e., classes I, II, and III (152). In addition, withinBSC II there are three subtypes. Class II BSCs consist of achamber with a small open front in which an airflow is gener-ated to prevent microorganisms from escaping the chamber.Laboratory workers sitting behind the cabinet insert theirhands and arms into the chamber. All objects and the arms ofthe worker can disturb the airflow and cause the microorgan-isms to escape. Class III BSCs are basically similar, but thefront is completely closed and can be accessed via attachedgloves.

BSCs should meet legal standards, as, for example, definedby the European Union (EN12469). BSCs have been improvedsignificantly during recent years due to this EN12469 standard,among others. Unfortunately, literature on the containmentefficiency of class II BSCs is scarce and is addressed mainly inolder publications, and such literature is virtually absent forclass III BSCs.

In general BSCs provide a good level of protection whenoperated and maintained correctly (107). However, in severalolder studies, before the introduction of EN12469, it wasshown that personnel working with open-front BSCs can stillbe exposed to infectious doses of microorganisms (7, 74, 112).Barbeito and Taylor (6) investigated the efficiency of contain-ment of a BSC under three different closure conditions anddifferent air velocities. In the cabinet, between 105 and 106

microorganisms per cubic foot were released in 5 min. Whenthe glove panel was removed, a human infectious dose wasreleased, and the number of microorganisms that escaped con-tainment increased with decreased air velocity. Moreover, anincrease in human activity in the cabinet resulted in increasednumbers of microorganisms escaping the cabinet. When theglove panel was attached, no microorganisms could be de-tected outside the cabinet. A remarkable finding was that whenthe glove panel was installed without the gloves attached, nomicroorganisms escaped from the cabinet. Their main conclu-sions were that laboratory workers are protected from infec-tious microorganisms only when they use closed BSCs withhigh airflow velocities and that the effectiveness of BSCs iscompromised by the activity of the workers. Macher and First(86) performed measurements on exposure of workers to bac-terial spores using a class II BSC with an adjustable workopening. Aerosols of bacterial spores were created inside thecabinet, and the escaping spores were measured. The workopening height appeared to be a significant predictor of sporeconcentrations outside the BSC (86). Similarly, airflow velocitywas negatively correlated with the concentration of escapingspores. Human activity in the cabinet, such as hands movingthrough the opening, also resulted in the escape of spores.Spore concentrations in the operator’s breathing zone wereabout 24 times higher than acceptable levels. Surprisingly,working in the rear of the cabinet was less safe than working inthe front, since the close proximity of the body to the cabinetinfluenced the airflow. Thus, for safe working conditions it isessential to limit the movement of arms and hands by arrang-ing the equipment in the most practical way. Heidt (54) alsotested the efficiency of a class II BSC. This author concludedthat the cabinet provided sufficient protection, since microor-ganisms escaped only at the highest densities of the test aerosolcreated inside the cabinet. Since the number of bacteria de-

tected was very low, this was considered to be acceptable.Osborne and coworkers (107) investigated a number of BSCsand calculated the operator protection factors (OPFs), as as-sessed by still and latterly limited “in-use” KI-Discus tests. TheOPF is defined as the ratio of the exposure to airborne con-tamination generated on the open bench to the exposure re-sulting from the same dispersal of airborne contaminationgenerated within the cabinet (66). Most BSCs had OPFs higherthan 100,000, except when room pressure changed or whendrafts occurred in the laboratory. The performance of class IIBSCs was shown to be affected by the movements of theworker, and some movements reduced OPF results as foundbefore by Macher and First (86). However, the levels of failurewere marginal. The OPF tests revealed that a selected class IIunit provided the same OPF as a class I unit when properlyused.

Although the literature is not unequivocal, it appears thatthe use of BSCs decreases LAIs significantly (54, 86, 107, 119).However, in a recent publication, Rusnak et al. (119) examinedillness surveillance data archived from the U.S. offensive bio-logical warfare program (from 1943 to 1969) and concludedthat BSCs and other measures failed to sufficiently preventillness caused by agents with lower infective doses in a high-riskresearch setting.

Though required in some countries (including The Nether-lands), cabinet performance is not generally assessed. How-ever, on-site containment tests indicated that 37 class II BSCs(all with adequate type test certification and including 18 newinstallations) failed to meet the OPF requirements as definedin BS 7526. Thus, testing for containment using an OPF testappears to be essential both at commissioning and during rou-tine maintenance (22, 107).

While formaldehyde gas has been used for over 100 years fordecontamination, the efficacy of this process remains contro-versial (95). Moreover, because of its toxicity, the use of form-aldehyde itself requires containment procedures (66). Formal-dehyde decontamination of BSCs is usually validated usingspore strips and culture. Therefore, poliovirus, Mycobacteriumbovis strain BCG, or Bacillus spores have been used. Bacterialspores on stainless steel appear to be resistant to decontami-nation, and using bacterial spores to validate decontaminationis too slow. Therefore, commercial biological indicator tests havebeen developed, which may be an aid in detecting incompletedecontamination. Difficulties in obtaining effective decontamina-tion by using formaldehyde gas have been demonstrated. Factorscontributing to the effectiveness of decontamination by formal-dehyde include the formaldehyde level, the relative humidity, thetemperature levels, and the medium to be contaminated. Loca-tions beyond the exhaust filters of BSCs were the most difficult todecontaminate.

Cell sorters. Modern cell sorting equipment has become animportant tool in microbiological laboratories (77, 110). Be-cause cell sorters lead to aerosol formation and are not easilyaccommodated by regular BSCs, this kind of apparatus couldcause the operators to become contaminated (77, 110, 125). Inaddition, their high costs often prohibit their incorporationwithin BSL facilities. To solve this problem, Lennartz andcoworkers (77) integrated a fluorescence-activated cell sorter(FACS) into a specially developed class II BSC. Biosafety wassubsequently tested by using T4 bacteriophage aerosols and

414 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

shown to be excellent. Bacteriophages were readily detectedinside and outside when the airflow of the BSC was off, butwhen the BSC was turned on no bacteriophages could bedetected outside (77). Many FACS protocols include inactiva-tion steps, including the use of fixatives based on alcohols orformaldehyde. Some of these protocols have been evaluatedfor antimicrobial activity directed against specific pathogens, inparticular HIV. Formaldehyde at concentrations of 0.5 to 2%is effective in inactivating HIV, but the ability of fixatives toinactivate other microorganisms in FACS equipment, includ-ing HBV, has not been demonstrated. In addition some pro-tocols employ nonfixed cells. While analytic cytometers areengineered not to produce aerosols, jet-in-air cell sorters gen-erate droplets and microdroplets that may be aerosolized. Re-cently high-speed cell sorting using high operating pressureswith an increased potential for aerosol generation and an en-hanced risk of sample splashes at the sample introduction porthas become more prevalent. At the same time, instrumentmanufacturers have become more safety conscious and havedeveloped novel devices for containment of aerosols andsplashes, modified sample uptake ports on cell sorters, andinstalled mechanisms to stop sample flow in case of a nozzleclog to reduce operator risks (111, 125). A Vantage FACS wasthus modified for safe use with potentially HIV-infected cells.Safety tests with bacteriophages were performed to evaluatethe potential spread of biologically active material during cellsorting. The bacteriophage sorting showed that the biologicallyactive material was confined to the sorting chamber. A failuremode simulating a nozzle blockage resulted in detectable drop-lets inside the sorting chamber, but no droplets could be de-tected when an additional air suction from the sorting chamberhad been put on (133). While these observations may be reas-suring, some recommendations regarding the use of FACSequipment are important (124–126). The recently publishedInternational Society for Analytical Cytology biosafety stan-dard for sorting of unfixed cells states that droplet-based sort-ing of infectious or hazardous biological material requires ahigher level of containment than the one recommended for therisk group classification of the pathogen (126). Training has toinclude performing aerosol containment testing of instrumentsto be used for biohazardous sorting. In addition, waste fluidhas to be collected in 10% sodium hypochlorite, and fluid linesshould be disinfected using a 1:10 dilution of 5.25% sodiumhypochlorite. Notwithstanding their potential hazards, no doc-umented disease transmission through the use of a cytometerhas occurred (124, 125).

Respiratory protection devices. A few papers examined theefficacy of face respirators and surgical masks. For example,Balazy et al. (5) examined the performance of two types of N95half-mask, filtering face piece respirators and two types ofsurgical masks. The collection efficiency of these respiratoryprotection devices was investigated using MS2 virus (a non-harmful simulant of several pathogens) in a particle size rangeof 10 to 80 nm. Penetration of virions through N95 respirators,which are certified by the National Institute for OccupationalSafety and Health (NIOSH), can exceed an expected level of5%. The tested surgical masks showed a much higher particlepenetration of the MS2 virions: 20.5% and 84.5%.

Does the Laboratory as a Whole Afford EffectiveContainment?

The proper functioning of equipment should be evaluatednot only in isolation but also in the context of the entirelaboratory. In considering biosafety of laboratories, the properfunctioning of autoclaves may be overlooked. Barbeito andBrookey (7) and Marshall et al. (87) emphasized the potentialof autoclaves to release viable microorganisms into the atmo-sphere and the importance of proper sterilizer location,ventilation, containment of heavily contaminated loads, andadequate sterilizer maintenance.

One of the few studies to assess contamination of the labo-ratory environment with pathogens found in blood examined800 environmental samples taken from 10 clinical and researchlaboratories working at the BSL-2 level at the NIH. Thirty-onesamples from 11 work stations in three laboratories containedHBV surface antigen. Factors associated with environmentalcontamination included flawed laboratory techniques (mouthpipetting, splashing, placing pens in the mouth, improper useof equipment, and improper instrument design requiring ex-ternal wash steps) (odds ratio [OR], 9.78; 95% confidenceinterval [CI], 1.46 to 65.49), high work loads (OR, 5.06; 95%CI, 0.8 to 31.96), and inappropriate behaviors (including notwearing gloves) (OR, 2.75; 95% CI, 0.44 to 17.4). Flow cytom-etry was identified as the technique with the most frequentoccurrence of overt spills (38). Indeed, HBV infection wasamong the most commonly reported LAIs. Laboratory workersin urban medical centers may have been at almost three timesthe risk of acquiring HBV infection than other hospital em-ployees due to exposure to patients’ blood and at 7 to 10 timesthe risk than that of the general public (38).

Some evidence of the effectiveness of a BSL-3 laboratoryenvironment may be derived from experiences with a speciallydesigned BSL-3 laboratory for autopsies of patients with SARS(79). SARS coronavirus is highly infectious, and during theoutbreak of SARS more than 30% of the approximately 8,000infected persons were health care workers. The autopsy labo-ratory was established in Beijing Ditan Hospital (which wasdesignated the SARS hospital during the outbreak of SARS inChina) in May 2003. Remarkably, the efficiency of decontam-ination in this laboratory was evaluated by a sarin simulant test.A sarin simulant aerosol of 0.3-�m particles at 4 mg/liter wasgenerated and spread by a special device in the contaminatedarea. Sarin could not be detected in either the semicontami-nated area or the clean area, and particles of �0.3 �m in sizewere not detected in the exhaust air. Twenty-three pathologistsand technicians participated in 16 complete autopsies thatwere performed on patients with clinically confirmed or sus-pected SARS, of which seven cases were later confirmed to beSARS infections. None of these personnel demonstrated anyevidence of SARS infection.

The set of biocontainment measures that define BSL-4 bio-safety is comfortably the most comprehensive and stringent,but each setting and laboratory design is unique and compar-ative data on their containment effectiveness are nonexistent.A problem in assessing the effectiveness of BSL-4 containmentmeasures is that isolated containment measures are consideredinsufficient. Thus, individual components such as autoclaves,incinerators, chemical decontamination showers, gaseous de-

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 415

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

contamination systems, air ventilation systems, and HEPA fil-ters can be tested for physical parameters during normal op-eration and under extreme conditions, but it remains unclearhow closely the simulated test conditions resemble the real-lifesituation. The effectiveness of HEPA filters is typically vali-dated using bacterial spore strips or particles.

Are Laboratory Workers and the Environment Protectedagainst Infection?

The analysis of laboratory accidents may illustrate what cango wrong and point the way to improvements. Such accidentsare one of the most relevant parameters to evaluate the overalleffectiveness of integrated biosafety measures. However, theepidemiology of the incidence and severity of LAIs is largelyunknown, as there are neither national surveillance and mon-itoring systems with complete coverage nor many systematicstudies on their occurrence (60, 131). Denominator data thatare necessary to calculate the actual incidence of LAIs areusually lacking. In addition, LAIs may be subclinical and mayhave an atypical incubation period and route of infection, andlaboratory workers and directors may be reluctant to reportthem because of fear of reprisal and stigma (52, 131). There-fore, much information is obtained from anecdotal case re-ports and some retrospective questionnaires. Such case reportsdo not always report on possible failure of biosafety proce-dures or unintended accidents. While accidental parenteralinoculation of infectious material appears to be one of theleading causes of LAIs, most LAIs appear to occur even withthe best safety precautions in place (116, 131). A summary ofsome recent LAIs is given in Table 10.

Reviews. Most LAIs are caused by microorganisms that arevery pathogenic or that need a very low infectious dose, in-cluding arboviruses, Venezuelan equine encephalitis virus,hantavirus, HBV, HCV, Brucella sp., Coxiella burnetii, Fran-cisella tularensis, Mycobacterium tuberculosis, Salmonella spp.,Shigella spp., Chlamydia psittaci, Blastomyces dermatitidis, Coc-cidioides immitis, Cryptosporidium spp., and organisms causingtyphus, streptococcal infections, histoplasmosis, leptospirosis,tularemia, coccidiomycosis, and dermatomycosis (52, 116,149). A direct link to accidents or exposure events, such asaspiration, injection, cut, spill, or bite, appears to be apparentin only a minority of the LAIs, while the majority are likelycaused by undefined exposure to aerosols (52, 116, 149–151,159). Indeed aerosols have been responsible for major out-breaks of LAIs caused by Brucella spp., Coxiella burnetti (Qfever), Chlamydia psittaci (psittacosis), and M. tuberculosis (89,103, 129). The main hazards for inoculation (114–116, 129,131) include (i) parenteral inoculation, (ii) inhalation of infec-tious aerosols, and less commonly, (iii) accidental oral inges-tion and (iv) direct contact with mucous membranes or (bro-ken) skin. Special hazards occur when working with infectedanimals. The presence of highly pathogenic microorganisms inunknown clinical samples likely explains the high incidence oftuberculosis among laboratory workers. In different studies theincidence of tuberculosis in laboratory personnel is estimatedto be 3 to 100 times the frequency observed in the generalpopulation. The high infectivity of M. tuberculosis is related toits low infective dose (i.e., a 50% infective dose of �10 bacilli)(1, 117). Schellekens (123) calculated that 1 out of 100 to 1,000

laboratory workers per year are infected, but recent studiessuggest that the rate of LAIs per person per year is decreasing(107, 123, 156).

Pike (116) tabulated the most common sources of LAIs frompublished literature and survey data. In the period from 1924to 1977, there were 4,079 reported cases of LAIs with 168casualties. In the subsequent period from 1980 to 1991, therewere 375 reported cases with 5 casualties. At the time of Pike’ssurvey, most LAIs (59%) occurred in research laboratories,compared with 17% in diagnostic laboratories. The highestmortality rate (7.8%) was associated with psittacosis. At thattime, approximately 70% of LAIs resulted from work with theinfectious agents (21%) or animals (17%), exposure to aero-sols (13%), and accidents (18%). Less frequent sources ofinfection included clinical specimens (7%), autopsies (2%),and contaminated glassware (1%). Most causes of LAI wereunknown (82%), and in only 18% of the reported cases couldthe cause be attributed to accidents, associated with the use ofsharps such as needles (25%), injuries by glass (16%), splashesor spills (27%), mouth pipetting (13%), and bites by laboratoryanimals (14%). Many of the LAIs of unknown origin werelikely caused by exposure to an infectious aerosol.

A recent survey of symptomatic and asymptomatic LAIs hasbeen conducted by Harding and Byers (52), who reviewed 270publications from 1979 to 2004, a period during which much hasbeen done to improve laboratory safety while the work load inlaboratories increased. A decrease in the number of LAIs wouldtherefore be expected; however, knowledge on the total popula-tion at risk and the total number of infections would be needed.Harding and Byers (52) found a total of 1,448 cases and 36 deaths,6 of which were aborted fetuses. The infections occurred in clin-ical, research, teaching, public health, and production facility lab-oratories, with clinical and research laboratories accounting forapproximately 76%. In recent years more LAIs from clinical lab-oratories were reported, probably due to a more active employeehealth program, the absence of biosafety containment equipmentin a number of clinical laboratories, or the fact that during theearly stages of culture identification, personnel are working withunknowns and may not be using adequate containment proce-dures. Like earlier findings, the authors report that only a smallproportion of the LAIs resulted from actual accidents. Most wereacquired by simply working in the laboratory or by exposure toinfected animals.

Sewell (131) concluded that adherence to the guidelines pro-mulgated by the various regulatory agencies decreases the risk ofoccupational exposure to infectious agents. However, he also rec-ommended additional studies to evaluate the effectiveness ofother safety measures implemented or mandated in the labora-tory. Interestingly, Sewell (131) describes personal risk factors oflaboratory workers that are associated with accidental infections.Characteristics of persons who have few accidents include adher-ence to safety regulations, a respect for infectious agents, “defen-sive” work habits, and the ability to recognize a potentially haz-ardous situation. In contrast, persons involved in laboratoryaccidents tend to have low opinions of safety programs, to takeexcessive risks, to work too fast, and to be less aware of theinfectious risks of the agents they are handling. Also, men andyounger employees (17 to 24 years old) are involved in moreaccidents than women and older employees (45 to 64 years old).

While many reports emphasize the importance of personal

416 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

TABLE 10. Causes and factors involved in LAIs

Microorganism No. of cases Setting Possible primary cause Other associated factor(s) Reference(s)

GMORecombinant

vaccinia virus1 Research lab Needle stick injury while

handling mice59

1 Research lab Unknown; no apparentfailures in handling and nodisruptions of epidermalbarrier

Possible enhanced infectivitydue to immunomodulatinginsert (LFA/ICAM-1); longinterval after vaccination

95

1 Research lab Needle stick injuries 1051 Research lab Unknown Eczema as predisposing factor;

noncompliance withprocedures (no wearing ofgloves, no recent vaccination)

89

1 Research lab Unknown Infrequent eye protection,laboratory coat sleeves werenot elasticized and did notalways cover the wrist, wastepipettes were not disinfectedbefore removal from thebiosafety cabinet, work(including vortexing) outsidethe BSC, no vaccination

78

Non-GMOsVaccinia virus 1 Research lab Accidental cut wound on a

coverslipNo vaccination 158

1 Research lab Accidental needle stick injury No recent vaccination 94SARS coronavirus 1 Research lab Cross-contaminated West

Nile virus sample, but noclear recognized laboratoryaccident

Insufficient training andnoncompliance with BSL-3procedures

80

1 (90 persons quarantined) Research lab Spilling accident (in BSL-4laboratory)

101

2, plus 8 secondary casesoutside the lab, of which1 was lethal

Research lab Inadequate inactivation ofSARS virus batch followedby transport to low-safetylab

Noncompliance with procedures(no check on inactivation, nomonitoring of workers’ healthstatus); in addition, therewere problems with improperair circulation, poorly locatedautoclaves and freezers, andtraining and record-keeping(100)

36

2 Research lab Unknown; perhapsinadequate inactivationand infection outside theBSL-3 area

102

Marburg virus 25, plus 6 secondary casesamong medical staff (7deaths total)

Research lab Handling Vervet monkeysoriginating from Ugandabefore the virus wasdesignated a risk group 4organism

41, 132, 140

2 (Marburg, Germany, andBelgrade)

Research lab Accidental infection beforethe virus was designated arisk group 4 organism

10, 99

Ebola virus 1 Accidental needle inoculationwhile processing materialfrom patients in Africa

35

5 people exposed but not ill,likely because the strainsused were not veryvirulent or because ofinfinitesimal doses(Reston, VA, and FortDetrick, Frederick, MD)

Two accidents, occurringwhile handling nonhumanprimates and mice

3

1 (lethal) (Novosibirsk,Siberia)

Needle stick injury whilehandling guinea pigs

4

Lassa virus 1 Handling tissue cultures andinfected mice

76

Sabia virus 2 (Brazil and YaleUniversity)

Diagnostic andresearchlaboratories

Handling of unknown virus(in diagnostic laboratory)and leakage fromcentrifuging tubes (in BSL-3 laboratory)

8, 44, 82

Burkholderiamallei

1 Research lab Noncompliance withbiosafety practices (noroutine use of latex gloves)

Researcher had type 1 diabetesmellitus

135

E. coli O157:H7 5 Different diagnosticlabs

Not identified or notreported

Infections occurred before E.coli O157:H7 was reclassifiedas a risk group 3 organism

23

Continued on following page

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 417

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

protection, there are indications that extensive personal pro-tection by use of double gloves, face masks, and protectiveclothing is not the sole solution, since such measures can re-duce the dexterity of the laboratory worker, leading to in-creased accidents (122). This indicates that the use of sharpsshould be minimized when workers wear extensive personalprotection.

Laboratory-acquired parasitic infections, from both proto-zoa and helminths, have been extensively reviewed by Her-waldt (56). Importantly, because protozoa, in contrast to mosthelminths, multiply in humans, even a small inoculum cancause illness. The author summarizes 199 case reports on lab-oratory and health care workers. The most frequently reportedparasitic infections were caused by Trypanosoma cruzi, Toxo-plasma gondii, Plasmodium spp., Leishmania spp., and Crypto-sporidium parvum. Two cases (one of Chagas’ disease and oneof toxoplasmosis) were fatal. However, as with other infec-tions, accurate counts of accidental exposures and infectionsand information on the risk per person-year are unavailable.Some of the laboratory-acquired parasitic infections were di-rectly linked to accidents (e.g., a bite by an escaped infectedmosquito) and poor laboratory practices, such as recapping aneedle, removing a syringe from a needle, working bare-handed, mouth pipetting, and working too fast. For 105 casesan accident or a likely route of exposure could be presumed; 47(44.8%) of these included a percutaneous exposure via a sharpobject. In other cases no apparent accidents were recognizedor reported, suggesting that subtle exposures (e.g., contamina-tion of unrecognized microabrasions and exposure throughaerosolization or droplet spread) resulted in infection.

Surveys. Walker and Campbell (146) did one of the fewsystematic but retrospective studies. They carried out a retro-

spective questionnaire survey of 397 responding United King-dom laboratories covering 1994 and 1995. Approximately 75%of these were diagnostic laboratories, 14% were research lab-oratories, and 9% were teaching laboratories. Over 55,000person-years of occupational exposure were covered, and onlynine cases of LAI were identified, giving an overall infectionincidence rate of 16.2/100,000 person-years, compared with82.7 infections/100,000 person-years found in a similar surveycovering 1988 and 1989 which was conducted by Grist andEmslie (48). This decline in incidence continues the trendpreviously reported for the period from 1970 to 1989. Infec-tions were most common in females (in contrast to the findingsreported by Sewell [131]), in relatively young staff, in microbi-ology laboratory workers, and in scientific/technical employees.Gastrointestinal infections predominated, particularly shigel-losis, but few specific etiological factors relating to workingpractices were identified. These included a broken glass lead-ing to a hand cut, a rat bite, and aerosol contamination. Inmost cases no clear accident was reported. Lack of experiencewas cited as a definite factor in two of the cases. Single cases ofHCV, E. coli O157, and M. tuberculosis infection were identi-fied, in addition to single cases of nonspecified septicemia andgastroenteritis. The absence of any cases of HBV infection, asin 1988 to 1989, reflects a sharp decline since 1970 and wasattributed to increased awareness, better technique, and theavailability of immunization. Furthermore, the absence of eyeinfections and the paucity of skin infections may indicate goodtechnique and use of protective equipment. Despite the short-comings of this study (retrospective study design, no reliabledenominator, potential underreporting or underrecording, andno detection of asymptomatic infections), the authors con-cluded that the small number of cases identified indicates high

TABLE 10—Continued

Microorganism No. of cases Setting Possible primary cause Other associated factor(s) Reference(s)

4 Different diagnosticlabs

Noncompliance withbiosafety practices(handling without latexgloves, hands were notwashed each time thatgloves were removed, openlaboratory coat)

Sudden increase in volume ofspecimens, low infectiousdose, prolonged survival onstainless-steel surfaces (155)

134

Brucella melitensis 5 Diagnostic lab Not identified or notreported

49

7 Diagnostic lab No apparent failure ofrecommended safetypractices

Large number of isolates ofBrucella spp. handled peryear in area of endemicity

159

Brucella melitensis/B. abortus

75 (retrospective survey of30 years; attack rate,11.9 %)

Different diagnosticlabs

Break in biosafety measuresin 80% of the cases(including lack ofrecognition of an isolate ofBrucella sp. and failure towork in a biological safetycabinet)

Large number of isolates ofBrucella spp. handled peryear

12

Mycobacteriumtuberculosis

7 (retrospective survey) Different publichealth labs

Needle stick injury in 1 case;in the other cases, thesource of infection couldnot be determined

Inadequate isolationprocedures, high volume ofspecimens, faulty ventilation

63

Neisseriameningitidis

16 (retrospective survey) Different researchlabs

No respiratory protection in15/16 cases

130

2 Two clinical labs High volume of specimens andformation of microaerosols

50

5 (retrospective survey;relative risk for laboratoryworkers, 184 �95% CI, 60to 431�)

Different labs Working outside a BSC 11

418 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

standards of infection control, although they still recognizedroom for improvement. Finally, the study emphasized that thenotification system in place in the United Kingdom to reportLAIs is inadequate for the task of monitoring their true inci-dence in a comprehensive way, a conclusion that probablyholds true for many other countries. For comprehensive mon-itoring of the incidence of LAIs, it is necessary to establish aroutine, active surveillance program or prospective surveywhich has the support and commitment of the laboratoriesthemselves.

Recently, a report was drafted on the biosafety status ofclinical laboratories in Japan (160). Data were obtained from431 hospitals and 301 institutions. The authors found 28 casesof possible laboratory-associated tuberculosis infection, ofwhich 25 could be associated with the lack of BSCs, which arerequired for work with M. tuberculosis. Other risk factors wereinsufficiently skilled equipment operation and rupture acci-dents during centrifugation of blood. Within the last 5 years1,534 events of self-inflicted needle punctures were recorded(160).

A retrospective survey of incidents occurring during biotech-nological and clinical work in Flanders, Belgium, indicated thaton average 13.6 incidents occurred per year among 7,302 lab-oratory workers. As a result, 69 persons (�1%) were exposedto biological agents, resulting in 2 LAIs, caused, respectively,by L. monocytogenes and Brucella melitensis. Most incidentsoccurred in clinical laboratories, likely caused by the highernumber of working hours actually spent in clinical laboratoriesand the sometimes unknown nature of microorganisms. Han-dling of experimental animals and waste was considered risky.Most incidents were caused by human failure, including prickaccidents, spilling, breaking, and maintenance work carried outin the laboratory (30).

Sejvar et al. (130) undertook a systematic, retrospective eval-uation of the risk of meningococcal disease among clinicalmicrobiologists and an assessment of the laboratory proce-dures that might predispose technicians to infection. Cases ofsuspected or proven laboratory-acquired meningococcal dis-ease were identified by placing an information request one-mail discussion groups of infectious disease, microbiology,and infection control professional organizations. Sixteen casesof probable laboratory-acquired meningococcal disease occur-ring worldwide between 1985 and 2001 were identified, includ-ing six U.S. cases between 1996 and 2000. Nine cases (56%)were serogroup B; seven (44%) were serogroup C. Eight cases(50%) were fatal. In 15 cases (94%), isolate manipulation wasperformed without respiratory protection. An average of threemicrobiologists are estimated to be exposed to the 3,000 me-ningococcal isolates seen in U.S. laboratories, yearly resultingin an attack rate of 13/100,000 microbiologists between 1996and 2001, compared to 0.2/100,000 among U.S. adults in gen-eral. The case/fatality rate of 50% seen among survey cases issubstantially higher than that observed among community-ac-quired cases, which may be explained by ascertainment biasdue to underreporting of mild cases of disease. However, analternative possibility is that clinical microbiologists routinelywork with highly virulent strains and high concentrations oforganisms. All cases identified in this inquiry occurred amongmicrobiologists and not among workers in other areas of theclinical laboratory. This suggests that exposure to isolates of

Neisseria meningitidis, and not patient samples, represents theincreased risk for infection. In addition, all isolates were de-rived from sterile sites. None of the microbiologists identifiedwere working with isolates obtained from pharyngeal or respi-ratory secretions, suggesting that such pharyngeal isolates rep-resent a lower risk, presumably due to their lower pathogenic-ity. The authors concluded that prevention should focus on theimplementation of class II BSCs or additional respiratory pro-tection during manipulation of suspected meningococcal iso-lates. Following two cases that prompted this survey, CDC hasinstituted a prospective surveillance for laboratory-acquiredmeningococcal disease (21).

In The Netherlands, two surveillance systems monitor theoccurrence of labor-acquired infections, i.e., not exclusivelyLAIs. The number of reported LAIs is low, but both systemssuffer from serious underreporting and do not provide detailsof the transmission route or accidental cause of infection (55).

GMO-associated laboratory accidents. Fortunately, thenumber of accidental releases or LAIs with GMOs appears tobe very low. Jones et al. (59) reported the first case of acci-dental vaccination with a recombinant vaccinia virus (WesternReserve [WR] strain) expressing the nucleoprotein gene ofvesicular stomatitis virus. The infection had a relatively mildcourse, perhaps because the laboratory worker had receivedsmallpox vaccination 30 years before the accident or because ofattenuation of the virus by the insertional inactivation of thethymidine kinase gene. Openshaw et al. (105) reported anaccidental infection of a laboratory worker with recombinantvaccinia virus (WR strain) expressing proteins of respiratorysyncytial virus. The infection occurred through two separateneedle accidents during the same work session, although theworker was experienced in the procedure. The procedure wassubsequently modified to prevent further accidents. The labo-ratory worker had been vaccinated with standard smallpoxvaccine, a practice that may have restricted the severity ofsymptoms to local redness and swelling. Mempel et al. (89)reported the case of a recombinant vaccinia virus infection ina previously vaccinated researcher working with various genet-ically modified strains. The isolated virus carried a functionallyinactivated cytohesin-1 gene of human origin, which impairsleukocyte adhesion by interacting with the LFA/ICAM-1 axis.The immunomodulating nature of the inserted construct mighthave added to the infectivity of the virus. Although the paperdoes not detail safety procedures in the lab, the infectionoccurred while the handling was considered proper. Contactinfections were not reported. Lewis et al. (78) reported a caseof ocular vaccinia infection in an unvaccinated laboratoryworker. The infecting virus was a unique form of recombinantWR vaccinia virus constructed in the laboratory. Althoughlaboratory staff generally followed established biosafety pre-cautions, several opportunities for virus exposure were identi-fied. Experiments were performed partly outside a BSC. Staffinfrequently wore eye protection. Laboratory coat sleeves werenot elasticized and did not always cover the wrist. Waste pi-pettes were not disinfected before removal from the BSC.Instances occurred in which samples with low titers of live viruswere removed from the BSC, transported to other parts of thefacility, and manipulated. In addition, laboratory staff routinelyvortexed tubes containing live virus outside the BSC. Finally, alaboratory-acquired recombinant vaccinia virus infection run-

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 419

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

ning a severe course was reported in a research laboratorytechnician who had a long history of eczema. She had beenworking with a thymidine kinase-deficient strain of vacciniavirus for use as a vector for gene therapy. She did not report anaccidental inoculation, but she usually wore no protectivegloves. She had been vaccinated with smallpox vaccine as achild but not again before beginning her laboratory work withvaccinia virus (83).

Another case of laboratory-acquired vaccinia virus (nonre-combinant) infection was reported by Wlodaver et al. (158).This infection occurred in a laboratory technician who had notbeen previously vaccinated and who developed generalizedvaccinia. She had accidentally cut a finger on a coverslip whileworking with vaccinia virus. Evaluation of this accident in herlaboratory prompted a review of procedures for handling con-taminated glassware. Moussatche et al. (94) reported anotheraccidental needle stick inoculation of a laboratory worker withvaccinia virus. Although the patient had previously been vac-cinated against smallpox, severe lesions appeared on thefingers.

In total, there have been at least 19 reported cases of labo-ratory-acquired vaccinia virus infections, of which 5 were ac-cidental infections with recombinant vaccinia (83; this paper).Several researchers emphasize the necessity of vaccinia vacci-nation of laboratory workers. In the United States and Canada,specific recommendations exist for laboratory personnel whoconduct research with (recombinant) orthopoxviruses, includ-ing vaccinia virus (20, 153). The Advisory Committee on Im-munization Practices recommends revaccination at least every10 years for persons working with non-highly attenuated vac-cinia viruses, recombinant viruses developed from non-highlyattenuated vaccinia viruses, or other nonvariola orthopoxvi-ruses. To ensure an increased level of protection against morevirulent nonvariola orthopoxviruses (e.g., monkeypox virus),empirical revaccination every 3 years can be considered. Incontrast, mandatory guidelines with respect to vaccinia vacci-nation do not exist in Europe (57).

In conclusion, LAIs with GMOs appear very seldom andappear to be restricted to infections with recombinant vacciniavirus. Although laboratory accidents with other GMOs mayhave been unnoticed due to a subclinical course of infection,this situation seems to reflect that vaccinia virus is very widelyused. Perhaps more important, the recombinant virus is stillpathogenic, and this might be enhanced by certain gene inserts.Vaccinia virus infection can be established via several routes,including breaks in the skin, and the infectious dose is probablylow. Guidelines for working safely with vaccinia virus, whichinclude vaccination, are available (57). We consider it none-theless advisable to work with the highly attenuated strains ofvaccinia virus (modified vaccinia virus Ankara and NYVAC)or with avian poxviruses that have a restricted host range anddo not replicate in mammals (ALVAC and TROVAC) when-ever possible.

Accidents with risk category 4 organisms. LAIs with cate-gory 4 biological agents (filoviruses, arenaviruses, flaviviruses,and bunyaviruses) are extremely rare and usually occurredearlier in settings with lower levels of biocontainment and/orinvolved animal work (41, 42, 51, 91, 127, 138, 139, 114). Rarelaboratory incidents with New World arenaviruses have been

reported in earlier surveys more than 4 to 5 decades ago (e.g.,with Junin virus and Machupo virus) (51). Such experiencesillustrated the need for more effective measures to reducehazards.

While this low number of BSL-4 laboratory accidents may bereassuring, the number of BSL-4 labs and workers is increas-ing. This appears to be in defiance of the “concentration andenclosure” principle, because the risks associated with thiswork may increase with the number of facilities and workers(61, 62). Concern is further fuelled by several incidents thatincluded unreported infections (among others involving Bru-cella and Coxiella burnetti) and other biosafety breaches. In2006, a Department of Health and Human Services InspectorGeneral audit of security procedures found that 11 of 15 insti-tutions had “serious weaknesses,” such as unlocked doors andfreezers and lax inventory records (61, 62). Another incident,in 2007, was the escape of foot-and-mouth disease virus fromthe Pirbright facilities in the United Kingdom, which has beenlinked to an outdated effluent system and caused several out-breaks of this very contagious disease among cattle and sheep(53).

DISCUSSION

In this paper we have reviewed the principles underlyingbiosafety measures for work with pathogenic microorganismsand GMOs, and we have examined to what extent evidence fortheir effectiveness is available. Clearly the risks of working withGMOs are considered to be largely identical to those of work-ing with pathogenic non-GMO microorganisms, and hencemuch of the knowledge of and containment measures forGMOs are derived from the latter. Regulations appear to bemore strict for GMOs, however.

Among many reports on biosafety, we found only scarceinformation on the evaluation of effectiveness and on criteriato judge effectiveness. We must therefore keep in mind thatsafety cannot be expressed in absolute terms. It is a relativeconcept defined in terms of tolerability and acceptability limits(64). This notion implies that workers and regulators try to finda balance between the costs of safety measures and the poten-tial benefits of the work for society. For example, in microbi-ological work, safety measures and associated costs increasefrom BSL-1 to BSL-4. Indeed, safety measures at BSL-1 and -2are probably insufficient to prevent all infections with micro-organisms of the corresponding risk categories, but their con-sequences at these levels are considered to be acceptable ornegligible.

The current biosafety practice gradually evolved during theprevious century. Therefore, it is not immediately obviouswhether and what principles have been employed to ensuresafe work and on which scientific basis they were built. In thispaper we have tried to identify some principles that appear tounderlie the current practice. Such principles clearly partlyoverlap and mutually enhance each other. A central activity,either implicit but preferably explicit, is a thorough risk assess-ment procedure that considers all potentially harmful effectsand their possibility of occurrence. Other important underlyingprinciples are the use of (wherever possible and appropriate)biological containment, concentration and enclosure, exposureminimization, physical containment, and hazard minimization.

420 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

Clearly, throughout the world regulators have adapted themodel of universal precautions based on a classification ofmicroorganisms in four hazard classes and accompanying stan-dard safety practices (16, 156). The advantage of this model isthat work with certain microorganisms can be grouped to-gether to comply with the accompanying containment rulesaccording to their classification. A disadvantage may be thatthis universal model may overlook the necessity to tailor safetymeasures for specific microorganisms or specific strains withparticular routes of transmission or virulence properties. Re-searchers should also reckon with variability in human immu-nocompetence. Therefore, risk assessment remains at the coreof any individual experiment. Such a risk assessment should inparticular be based on (preferably quantitative) parameters oftransmission, infectivity, and virulence. These should guide thesubsequent measures aimed at reducing the amount of micro-organisms to which individuals are exposed to below a minimalthreshold level of infectivity. Nonetheless, this universal modelof four biohazard classes appears to work well, but we recom-mend further harmonization of criteria for both non-GMOand GMOs and between different regulatory authorities, suchas the European Union, WHO, and CDC.

Altogether, the regulations specifying the biosafety contain-ment measures appear to be based on experience, expert judg-ment, and common sense. They are not motivated or sup-ported (at least not explicitly) by scientific literature, however,and often are not based on precisely defined or specified prop-erties of microorganisms and vector and insert sequences. Inaddition, the regulations do not exactly specify the level ofprotection that they aim to afford, for example, in terms ofdiminishing exposure of the laboratory workers below a thresh-old level of infectivity. Furthermore, it is clear that the physicalcontainment classes 1 to 4 afford increasing levels of contain-ment, but it is not sufficiently clear and scientifically supportedto what extent they provide effective protection with regard toprevention of infection of laboratory personnel, prevention ofairborne escape, etc. This, together with sometimes not verydetailed regulations, puts much responsibility on researchers,lab directors, advisors, and regulatory authorities in furtherdetailing working practices. The regulations also do not com-prise evaluation procedures to monitor the compliance or ef-fectiveness of the containment provisions. Table 11 summa-rizes our recommendations.

The hazard classification of work with GMOs follows theclassification of work with non-GMOs. This extrapolationshould be based on a risk estimation as precise as possibleconsidering the genetic modifications involved. In case ofdoubt or uncertainty about the properties of the GMO in-volved, regulators and biosafety experts will choose a higherrisk classification than the risk level of the microorganism fromwhich the GMO has been derived, or they will demand addi-tional safeguards. Basic research on transmission properties ofGMOs, in comparison with those of the nonmodified organ-isms, may be helpful in such a risk assessment to further definethe risks involved in the manipulation of GMOs. However,often properties, such as infective dose, may be difficult toobtain. In risk assessment we consider it important to takegene-gene and gene-environment interactions into account,because experience (for example, with the IL-4/ectromelia

construct) has shown that specific gene products may haveunwanted effects in a particular environment (58).

Regulators do not always require routine evaluation andmonitoring of biosafety aspects in laboratories, which wewould like to recommend. Routine monitoring of biosafetyaspects, including monitoring of compliance and educationaland behavioral aspects, may not easily be implemented, inparticular in the many clinical laboratories with their highworkload involving a wide variety of sometimes unknown mi-croorganisms, but it may enhance overall safety awareness. Forexample, validation experiments using T4 bacteriophage, bac-terial spores, or other indicator microorganisms could be use-ful for this purpose.

From the literature it appears obvious that there is littleexperience and no consensus on how the effectiveness of bio-safety practices should be evaluated. Clearly, the effectivenessof biosafety measures can be assessed at different levels andunder different circumstances that logically complement eachother; i.e., one could question whether a single piece of equip-ment is effective under experimental conditions or, conversely,whether the population has not been accidentally exposed toLAIs. Data on the biological containment efficiency of equip-ment and laboratories are scarce and fragmented and aremainly limited to technical specifications. Monitoring of LAIstherefore appears to play a pivotal role in evaluating the ef-fectiveness of containment and the potential exposure of lab-oratory workers and the population, but it suffers from seriousunderreporting throughout the world. Many reports of labora-tory accidents are only anecdotal. We therefore recommendoptimization of the systematic monitoring of laboratory acci-dents, including the serological monitoring of personnel. In-

TABLE 11. Summary of recommendations

Recommendation

Strengthen the evidence base wherever possible and feasiblein modernizing biosafety measures; this may enhance theeffectiveness of biosafety measures as well as compliancewith these measures

Develop one set of guidelines and regulations for GMOs andnon-GMOs

Develop explicit goals of biosafety measures that can be evaluatedKnowledge and measures of biosafety should be directed to

(preferably quantitative) parameters of infectivity andtransmission

Further harmonize biosafety guidelines between regulatoryauthorities (European Union, CDC, WHO)

Take gene-gene and gene-environment interactions into account inrisk assessment

Optimize the use of biological containment, in particular withrespect to the use of recombinant vaccinia virus

Monitor and evaluate biosafety aspects in laboratories and theircompliance regularly

Optimize systematic surveillance of laboratory accidents andimplement recommendations following such accidents; “blame-free” reporting may enhance the reporting rate, and serologicalmonitoring may support the detection of laboratory infections andshould match the risks involved

Promote education of laboratory personnel and compliance with therules

Collect data to support the evidence base of the biosafety practiceDevelop mathematical models to support the further development

of knowledge of biosafety, to detect gaps in our knowledge, andto support the development and evaluation of biosafety measures

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 421

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

fection with microorganisms, either GMO or not, that have ahigh infective dose or low virulence, usually belonging to riskcategory 1 and 2, may be difficult to detect. The extent ofserological monitoring should therefore depend on the risksinvolved. A passive sampling strategy, i.e., collecting serumsamples at the time of employment and following incidents,may be sufficient for work with low-virulence microorganisms,but an active sampling strategy at regular intervals may beconsidered for class 3 and 4 microorganisms. One clue tooptimizing monitoring of accidents may be the introduction of“blame-free” reporting, which aims for workers to share expe-riences without being punished. In addition to systematic mon-itoring, retrospective surveys may be very useful, as they mayidentify certain risk factors, as shown for the occurrence ofmeningococcal disease (130).

Despite the methodological imperfections, it is clear that thenumber of GMO-associated laboratory accidents is very scarcein comparison with the number of non-GMO-associated infec-tions and is practically restricted to accidental infection withrecombinant vaccinia virus. We interpret this finding to meanthat the biological containment obtained by attenuating GMOsis possibly a major factor in preventing their transmission.However, other factors contributing to the low number ofGMO-associated accidents may be that GMOs are well char-acterized, implying that the worker has knowledge of the prop-erties of the GMO and that the work load involving GMOs islikely much lower than that in clinical laboratories. Moreover,clinical samples may contain unknown pathogens. Other fac-tors that may contribute to this low number of accidents in-volving GMOs are the stricter regulatory framework and astricter compliance with containment rules. In many countries,including The Netherlands, both the researcher and the regu-lator make a risk assessment for each individual project thatinvolves GMOs, a practice that is less developed for work withnon-GMOs. In case of doubt or uncertainties regarding theproperties of GMOs, biosafety experts and regulators will de-mand a higher risk category or additional measures. Both alocal biosafety officer and a national inspectorate supervise thispractice. Whenever possible, we consider it important to fur-ther optimize the possibilities of employing genetic modifica-tion to enhance the safety of GMOs. In particular, we recom-mend further definition of the genetic properties underlyingthe transmissibility and infectivity of microorganisms, mea-surement of the influence of specific mutations on infectivedose and transmission properties of GMOs, and that contain-ment rules be based on such findings. This is not an easy taskbut would provide a further scientific basis for the phenotypicproperties of GMOs and the accompanying level of biologicalcontainment afforded by specific genetic alterations. Vacciniavirus or recombinant viruses developed from non-highly atten-uated vaccinia viruses appear to be less well suited as vectororganisms due to their retained virulence and low infectiousdose and should be replaced by safer poxvirus vectors whereverpossible.

In many reports of LAIs there has been a lack of compliancewith biosafety practices. This observation may be reassuringregarding the effectiveness of such biosafety practices, at leastif they are followed. It illustrates that education of laboratorypersonnel and compliance with the rules remain the top pri-ority. Increased attention to these aspects may have caused a

decrease in the rate of LAIs per person per year (107, 156). Onthe other hand, in the majority of cases of LAIs a direct causecould not be assigned (52, 116, 131, 149, 159), suggesting thata failure was not noticed in many cases or that containmentmay have been insufficient. This observation may warrant fur-ther research on the routes of exposure in such cases and onthe effectiveness of measures. Finally, although monitoring ofLAIs is an important element in evaluating the effectiveness ofcontainment measures, it may overlook the risks associatedwith nonreplicating agents, such as transduction by nonrepli-cating viruses.

Many countries, including The Netherlands, regulate workwith pathogenic microorganisms and GMOs differently. Be-cause the regulations are derived from the same underlyingprinciples and use the same instruments for biosafety andbecause the number of accidents involving GMOs is very low,we recommend harmonization, modernization, and simplifica-tion of the regulatory framework through developing a singleset of regulations for both non-GMOs and GMOs.

Despite their presumed overall effectiveness in providingbiosafety, it is often unclear to what extent the current set ofspecific biological or physical containment measures, alone ortogether, contribute to the prevention of transmission ofpathogenic microorganisms or GMOs. In further developingand modernizing the biosafety practice, we therefore recom-mend developing evidence-based practices and criteria to eval-uate effectiveness wherever this is possible and feasible. Thismay optimize and perhaps simplify future biosafety measuresand stimulate compliance with the rules. Although scientificresearch may strengthen the evidence base for biosafety mea-sures, such work is complicated and does not necessarily guar-antee new findings on which further improvements can bebased. To unravel complexities and to obtain further insightinto the contribution of specific elements to biosafety, mathe-matical modeling, which is directed at quantitative parametersof infectivity and transmission, may be supportive, but model-ing obviously needs confirmation by observational and exper-imental findings. Such an approach may, however, point to thedata that are needed to further guide the development ofevidence-based risk analysis and containment policy for bothnon-GMO pathogens and GMOs.

ACKNOWLEDGMENTS

This work was financially supported by the Dutch Committee forGenetic Modification (COGEM).

We thank Marja Agterberg, Marjolein van Esschoten, Ben Peeters,Erik Schagen, Gijsbert van Willigen, and Dick van Zaane for theirconstructive support.

REFERENCES

1. Andersen, P., and T. M. Doherty. 2005. The success and failure of BCG—implications for a novel tuberculosis vaccine. Nat. Rev. Microbiol. 3:656–662.

2. Angelakopoulos, H., K. Loock, D. M. Sisul, E. R. Jensen, J. F. Miller, andE. L. Hohmann. 2002. Safety and shedding of an attenuated strain ofListeria monocytogenes with a deletion of actA/plcB in adult volunteers: adose escalation study of oral inoculation. Infect. Immun. 70:3592–3601.

3. Anonymous. 2004. Ebola virus, laboratory accident—USA (Maryland).ProMED-mail, 2 February 2004. 20040220.0550. http://www.promedmail.org. Accessed 22 September 2007.

4. Anonymous. 2004. Ebola, lab accident death—Russia (Siberia). ProMED-mail, 22 May 2004. 20040522.1377. http://www.promedmail.org. Accessed22 September 2007.

5. Bałazy, A. M., Toivola, A. Adhikari, S. K. Sivasubramani, T. Reponen, and

422 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

S. A. Grinshpun. 2006. Do N95 respirators provide 95% protection levelagainst airborne viruses, and how adequate are surgical masks? Am. J.Infect. Control 34:51–57.

6. Barbeito, M. S., and L. A. Taylor. 1968. Containment of microbial aerosolsin a microbiological safety cabinet. Appl. Microbiol. 16:1225–1229.

7. Barbeito, M. S., and E. A. Brookey, Jr. 1976. Microbiological hazard fromthe exhaust of a high-vacuum sterilizer. Appl. Environ. Microbiol. 32:671–678.

8. Barry, M., M. Russi, L. Armstrong, D. Geller, R. Tesh, L. Dembry, J. P.Gonzalez, A. S. Khan, and C. J. Peters. 1995. Treatment of a laboratory-acquired Sabia virus infection. N. Engl. J. Med. 333:294–296.

9. Berg, P., D. Baltimore, S. Brenner, R. O. Roblin III, and M. F. Singer. 1975.Asilomar conference on recombinant DNA molecules. Science 188:991–994.

10. Borchert, M. 2001. Marburg hemorrhagic fever, 1967 to the present (02).ProMED-mail, 8 Feb 2001. 20010208.0248. ]http://www.promedmail.org.Accessed 22 September 2007.

11. Boutet, R., J. M. Stuart, E. B. Kaczmarski, S. J. Gray, D. M. Jones, and N.Andrews. 2001. Risk of laboratory-acquired meningococcal disease. J.Hosp. Infect. 49:282–284.

12. Bouza, E., C. Sanchez-Carrillo, S. Hernangomez, M. J. Gonzalez, et al.2005. Laboratory-acquired brucellosis: a Spanish national survey. J. Hosp.Infect. 61:80–83.

13. Braat, H., P. Rottiers, D. W. Hommes, N. Huyghebaert, E. Remaut, J. P.Remon, S. J. van Deventer, S. Neirynck, M. P. Peppelenbosch, and L.Steidler. 2006. A phase I trial with transgenic bacteria expressing interleu-kin-10 in Crohn’s disease. Clin. Gastroenterol. Hepatol. 4:754–759.

14. Breban, R., R. Vardavas, and S. Blower. 2007. Theory versus data: how tocalculate R0? PLoS One 2:e282.

15. Reference deleted.16. Buesching, W. J., J. C. Neff, and H. M. Sharma. 1989. Infectious hazards in

the clinical laboratory: a program to protect laboratory personnel. Clin.Lab. Med. 9:351–361.

17. Bukovsky, A. A., J. P. Song, and L. Naldini. 1999. Interaction of humanimmunodeficiency virus-derived vectors with wild-type virus in transducedcells. J. Virol. 73:7087–7092.

18. Casadevall, A., and L. A. Pirofski. 1999. Host-pathogen interactions: rede-fining the basic concepts of virulence and pathogenicity. Infect. Immun.67:3703–3713.

19. CDC. 1974. Classification of etiological agents on the basis of hazard. U.S.Public Health Service, Washington, DC.

20. CDC. 2001. Vaccinia (smallpox) vaccine: recommendations of the AdvisoryCommittee on Immunization Practices (ACIP), 2001. Morb. Mortal. Wkly.Rep. 50:1–25.

21. CDC. 2002. Laboratory-acquired meningococcal disease—United States,2000. Morb. Mortal. Wkly. Rep. 51:141–144.

22. Clark, R. P. 1997. Standards for safety cabinets. Nature 390:550.23. Coia, J. E. 1998. Nosocomial and laboratory-acquired infection with Esch-

erichia coli O157. J. Hosp. Infect. 40:107–413.24. Collaco, R. E., X. Cao, and J. P. Trempe. 1999. A helper virus-free pack-

aging system for recombinant adeno-associated virus vectors. Gene 238:397–405.

25. Corbett, E. L., C. J. Watt, N. Walker, D. Maher, B. G. Williams, M. C.Raviglione, and C. Dye. 2003. The growing burden of tuberculosis: globaltrends and interactions with the HIV epidemic. Arch. Intern. Med. 163:1009–1021.

26. Curtiss, R. 1978. Biological containment and cloning vector transmissibility.J. Infect. Dis. 137:668–675.

27. Danelishvili, L., L. S. Young, and L. E. Bermudez. 2006. In vivo efficacy ofphage therapy for Mycobacterium avium infection as delivered by a non-virulent mycobacterium. Microb. Drug Resist. 12:1–6.

28. Darquet, A. M., B. Cameron, P. Wils, D. Scherman, and J. Crouzet. 1997.A new DNA vehicle for nonviral gene delivery: supercoiled minicircle.Gene Ther. 12:1341–1349.

29. Debyser, Z. 2003. Biosafety of lentiviral vectors. Curr. Gene Ther. 3:517–525.

30. De Cock, B., and G. van Eetvelde. 2007. Biologische Veiligheid. Verken-nende bio-incidentenanalyse in Vlaanderen. Report, Vlaamse milieu-maatschappij, MIRA/2007/01. www.biorisico.UGent.be.

30a.de Wind, N., B. P. Peeters, A. Zuderveld, A. L. Gielkens, A. J. Berns, andT. G. Kimman. 1994. Mutagenesis and characterization of a 41-kilobase-pair region of the pseudorabies virus genome: transcription map, search forvirulence genes, and comparison with homologs of herpes simplex virustype 1. Virology 200:784–790.

31. Dietrich, G., A. Bubert, I. Gentschev, Z. Sokolovic, A. Simm, A. Catic, S. H.Kaufmann, J. Hess, A. A. Szalay, and W. Goebel. 1998. Delivery of antigen-encoding plasmid DNA into the cytosol of macrophages by attenuatedsuicide Listeria monocytogenes. Nat. Biotechnol. 16:181–185.

32. Dilts, D. A., I. Riesenfeld-Orn, J. P. Fulginiti, E. Ekwall, C. Granert, J.Nonenmacher, R. N. Brey, S. J. Cryz, K. Karlsson, K. Bergman, T. Thomp-son, B. Hu, A. H. Bruckner, and A. A. Lindberg. 2000. Phase I clinical trials

of aroA aroD and aroA aroD htrA attenuated S. typhi vaccines; effect offormulation on safety and immunogenicity. Vaccine 18:1473–1484.

33. DiPetrillo, M. D., T. Tibbetts, H. Kleanthous, K. P. Killeen, and E. L.Hohmann. 1999. Safety and immunogenicity of phoP/phoQ-deleted Salmo-nella typhi expressing Helicobacter pylori urease in adult volunteers. Vaccine18:449–459.

34. Reference deleted.35. Emond, R. T., B. Evans, E. T. Bowen, and G. Lloyd. 1977. A case of Ebola

virus infection. Br. Med. J. 2:541–544.36. Enserink, M., and L. Du. 2004. SARS. China dumps CDC head, probes lab.

Science 305:163.37. Escarpe, P., N. Zayek, P. Chin, F. Borellini, R. Zufferey, G. Veres, and V.

Kiermer. 2003. Development of a sensitive assay for detection of replica-tion-competent recombinant lentivirus in large-scale HIV-based vectorpreparations. Mol. Ther. 8:332–341.

38. Evans, M. R., D. K. Henderson, and J. E. Bennett. 1990. Potential forlaboratory exposures to biohazardous agents found in blood. Am. J. PublicHealth 80:423–427.

39. Farson, D., R. McGuinness, T. Dull, K. Limoli, R. Lazar, S. Jalali, S.Reddy, R. Pennathur-Das, D. Broad, and M. Finer. 1999. Large-scale man-ufacturing of safe and efficient retrovirus packaging lines for use in immu-notherapy protocols. J. Gene Med. 1:195–209.

40. Fenner, F., D. A. Henderson, I. Arita, Z. Jezek, and I. D. Ladnyi. 1988.Small pox and its eradication. WHO, Geneva, Switzerland. http://whqlibdoc.who.int/smallpox/9241561106.pdf.

41. Feldmann, H., W. Slenczka, and H. D. Klenk. 1996. Emerging and re-emerging of Filoviruses. Arch. Virol. Suppl. 11:77–100.

42. Feldmann, H., S. Jones, H. D. Klenk, and H. J. Schnittler. 2003. Ebolavirus: from discovery to vaccine. Nat. Rev. Immunol. 3:677–685.

43. Flemming, D. O. 2000. Risk assessment of biological hazards, p. 57–64. InD. O. Flemming and D. L. Hunt (ed.), Biological safety: principles andpractices, 3rd ed. ASM Press, Washington, DC.

44. Gandsman, E. J., H. G. Aaslestad, T. C. Ouimet, and W. D. Rupp. 1997.Sabia virus incident at Yale University. Am. Ind. Hyg. Assoc. J. 58:51–53.

45. Gershon, R. R., D. Vlahov, S. A. Felknor, D. Vesley, P. C. Johnson, G. L.Declos, and L. R. Murphy. 1995. Compliance with universal precautionsamong health care workers at three regional hospitals. Am. J. Infect. Con-trol 23:225–236.

46. Gorbach, S. L. 1978. Recombinant DNA: an infectious disease perspective.J. Infect. Dis. 137:615–623.

47. Grimm, D., P. Staeheli, M. Hufbauer, I. Koerner, L. Martınez-Sobrido, A.Solorzano, A. Garcıa-Sastre, O. Haller, and G. Kochs. 2007. Replicationfitness determines high virulence of influenza A virus in mice carryingfunctional Mx1 resistance gene. Proc. Natl. Acad. Sci. USA 104:6806–6811.

48. Grist, N. R., and J. A. N. Emslie. 1991. Infections in British clinical labo-ratories, 1988–1989. J. Clin. Pathol. 44:667–669.

49. Gruner, E., E. Bernasconi, R. L. Galeazzi, D. Buhl, R. Heinzle, and D.Nadal. 1994. Brucellosis: an occupational hazard for medical laboratorypersonnel. Report of five cases. Infection 22:33–36.

50. Guibourdenche, M., J. P. Darchis, A. Boisivon, E. Collatz, and J. Y. Riou.1994. Enzyme electrophoresis, sero- and subtyping, and outer membraneprotein characterization of two Neisseria meningitidis strains involved inlaboratory-acquired infections. J. Clin. Microbiol. 32:701–704.

51. Hanson, R. P., S. E. Sulkin, E. L. Beuscher, W. M. Hammon, R. W.McKinney, and T. H. Work. 1967. Arbovirus infections of laboratory work-ers. Extent of problem emphasizes the need for more effective measures toreduce hazards. Science 158:1283–1286.

52. Harding, A. L., and K. B. Byers. 2006. Epidemiology of laboratory-associ-ated infections, p. 53–77. In D. O. Flemming and D. L. Hunt (ed.), Bio-logical safety: principles and practices, 4th ed. ASM Press, Washington,DC.

53. Health and Safety Executive. 2007. Final report on potential breaches ofbiosecurity at the Pirbright site 2007. http://www.hse.gov.uk/news/archive/07aug/finalreport.pdf.

54. Heidt, P. J. 1982. Bacteriological testing of a modified laminar flow micro-biological safety cabinet. Antonie van Leeuwenhoek 48:373–381.

55. Heimeriks, K. 2007. Surveillance van arbeidsgerelateerde infectieziekten.Report LCI/CIb. RIVM, Bilthoven, The Netherlands.

56. Herwaldt, B. L. 2001. Laboratory-acquired parasitic infections from acci-dental exposures. Clin. Microbiol. Rev. 14:659–688.

57. Isaacs, S. N. 2004. Working safely with vaccinia virus: laboratory techniqueand the role of vaccinia vaccination. Methods Mol. Biol. 269:1–14.

58. Jackson, R. J., A. J. Ramsay, C. D. Christensen, S. Beaton, D. F. Hall, andI. A. Ramshaw. 2001. Expression of mouse interleukin-4 by a recombinantectromelia virus suppresses cytolytic lymphocyte responses and overcomesgenetic resistance to mousepox. J. Virol. 75:1205–1210.

59. Jones, L., S. Ristow, T. Yilma, and B. Moss. 1986. Accidental humanvaccination with vaccinia virus expressing nucleoprotein gene. Nature 319:543.

60. Kahn, L. H. 2004. Biodefense research: can secrecy and safety coexist?Biosecur. Bioterror. 2:81–85.

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 423

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

61. Kaiser, J. 2007. Accidents spur a closer look at risks at biodefense labs.Science 317:1852–1854.

62. Kaiser, J. 2007. Lawmakers worry that lab expansion poses risks. Science318:182.

63. Kao, A. S., D. A. Ashford, M. M. McNeil, N. G. Warren, and R. C. Good.1997. Descriptive profile of tuberculin skin testing programs and laborato-ry-acquired tuberculosis infections in public health laboratories. J. Clin.Microbiol. 35:1847–1851.

64. Kappeli, O., and L. Auberson. 1997. The science and intricacy of environ-mental safety evaluations. Trends Biotechnol. 9:342–349.

65. Keatley, K. L. 2000. Controversy over genetically modified organisms: thegoverning laws and regulations. Qual. Assur. 8:33–36.

66. Kennedy, D. A., and C. H. Collins. 2000. Microbiological safety cabinets:selection, installation, testing, and use. Br. J. Biomed. Sci. 57:330–337.

67. Khan, S. A., R. Stratford, T. Wu, N. Mckelvie, T. Bellaby, Z. Hindle, K. A.Sinha, S. Eltze, P. Mastroeni, D. Pickard, G. Dougan, S. N. Chatfield, andF. R. Brennan. 2003. Salmonella typhi and S. typhimurium derivatives har-bouring deletions in aromatic biosynthesis and Salmonella pathogenicityisland-2 (SPI-2) genes as vaccines and vectors. Vaccine 21:538–548.

68. Kimman, T. G., A. L. Gielkens, K. Glazenburg, L. Jacobs, M. C. de Jong,W. A. Mulder, and B. P. Peeters. 1995. Characterization of live pseudora-bies virus vaccines. Dev. Biol. Stand. 84:89–96.

69. Knudsen, S., P. Saadbye, L. H. Hansen, A. Collier, B. L. Jacobsen, J.Schlundt, and O. H. Karlstrom. 1995. Development and testing of im-proved suicide functions for biological containment of bacteria. Appl. En-viron. Microbiol. 61:985–991.

70. Kobasa, D., A. Takada, K. Shinya, M. Hatta, P. Halfmann, S. Theriault, H.Suzuki, H. Nishimura, K. Mitamura, N. Sugaya, T. Usui, T. Murata, Y.Maeda, S. Watanabe, M. Suresh, T. Suzuki, Y. Suzuki, H. Feldmann, andY. Kawaoka. 2004. Enhanced virulence of influenza A viruses with thehaemagglutinin of the 1918 pandemic virus. Nature 431:703–707.

71. Kochi, S. K., K. P. Killeen, and U. S. Ryan. 2003. Advances in the devel-opment of bacterial vector technology. Expert Rev. Vaccines 2:31–43.

72. Kost, T. A., J. P. Condreay, and C. A. Mickelson. 2000. Biosafety and viralgene transfer factors, p. 579–597. In D. O. Flemming and D. L. Hunt (ed.),Biological safety: principles and practices, 3rd ed. ASM Press, Washington,DC.

73. Kristoffersen, P., G. B. Jensen, K. Gerdes, and J. Piskur. 2000. Bacterialtoxin-antitoxin gene system as containment control in yeast cells. Appl.Environ. Microbiol. 66:5524–5526.

74. Kruse, R. H. 1962. Potential aerogenic laboratory hazards of Coccidioidesimmitis. Am. J. Clin. Pathol. 37:150–158.

75. Laquerre, S., W. F. Goins, S. Moriuchi, T. J. Oligino, D. M. Krisky, P.Marconi, M. K. Soares, J. B. Cohen, J. C. Glorioso, and D. J. Fink. 1999.Gene transfer tool: herpes simplex virus vectors, p. 173–208. In T. Fried-mann (ed.), The development of gene therapy. Cold Spring Harbor Labo-ratory, Cold Spring Harbor, NY.

76. Leifer, E., D. J. Gocke, and H. Bourne. 1970. Lassa fever, a new virusdisease of man from West Africa. II. Report of a laboratory-acquiredinfection treated with plasma from a person recently recovered from thedisease. Am. J. Trop. Med. Hyg. 19:677–679.

77. Lennartz, K., L. Mengli, M. Flasshove, T. Moritz, and U. Kirstein. 2005.Improving the biosafety of cell sorting by adaptation of a cell sorting systemto a biosafety cabinet. Cytometry 66A:119–127.

78. Lewis, F. M., E. Chernak, E. Goldman, Y. Li, K. Karem, I. K. Damon, R.Henkel, C. Newbern, P. Ross, and C. C. Johnson. 2006. Ocular vacciniainfection in laboratory worker, Philadelphia, 2004. Emerg. Infect. Dis. 12:134–137.

79. Li, L., J. Gu, X. Shi, E. Gong, X. Li, H. Shao, X. Shi, H. Jiang, X. Gao, D.Cheng, L. Guo, H. Wang, X. Shi, P. Wang, Q. Zhang, and B. Shen. 2005.Biosafety level 3 laboratory for autopsies of patients with severe acuterespiratory syndrome: principles, practices, and prospects. Clin. Infect. Dis.41:815–821.

80. Lim, P. L., A. Kurup, G. Gopalakrishna, K. P. Chan, C. W. Wong, L. C. Ng,S. Y. Se-Thoe, L. Oon, X. Bai, L. W. Stanton, Y. Ruan, L. D. Miller, V. B.Vega, L. James, P. L. Ooi, C. S. Kai, S. J. Olsen, B. Ang, and Y. S. Leo. 2004.Laboratory-acquired severe acute respiratory syndrome. N. Engl. J. Med.350:1740–1745.

81. Lim, W., and D. N. C. Tsang. 2006. Laboratory containment of SARS virus.Ann. Acad. Med. Singapore 35:354–360.

82. Lisieux, T., M. Coimbra, E. S. Nassar, M. N. Burattini, L. T. de Souza, I.Ferreira, I. M. Rocco, A. P. da Rosa, P. F. Vasconcelos, F. P. Pinheiro, et al.1994. New arenavirus isolated in Brazil. Lancet 343:391–392.

83. Loeb, M., I. Zando, M. C. Orvidas, A. Bialachowski, D. Groves, and J.Mahoney. 2003. Laboratory-acquired vaccinia infection. Can. Commun.Dis. Rep. 29:134–136.

84. Low, K. B., M. Ittensohn, T. Le, J. Platt, S. Sodi, M. Amoss, O. Ash, E.Carmichael, A. Chakraborty, J. Fischer, S. L. Lin, X. Luo, S. I. MillerI, L.Zheng, I. King, J. M. Pawelek, and D. Bermudes. 1999. Lipid A mutantSalmonella with suppressed virulence and TNFalpha induction retain tu-mor-targeting in vivo. Nat. Biotechnol. 17:37–41.

85. Lucero, N. E., and F. Sineriz. 2005. The Argentine experience in enhancing

biosafety through good laboratory practices. Asian Biotechnol. Dev. Rev.8:99–120.

86. Macher, J. M., and M. W. First. 1984. Effects of airflow rates and operatoractivity on containment of bacterial aerosols in a class II safety cabinet.Appl. Environ. Microbiol. 48:481–485.

87. Marshall, B. M., H. Shin-Kim, D. Perlov, and S. B. Levy. 1999. Release ofbacteria during the purge cycles of steam-jacketed sterilizers. Br. J. Biomed.Sci. 56:247–252.

88. Memish, Z. A., and M. W. Mah. 2001. Brucellosis in laboratory workers ata Saudi Arabian hospital. Am. J. Infect. Control 29:48–52.

89. Mempel, M., G. Isa, N. Klugbauer, H. Meyer, G. Wildi, J. Ring, F. Hof-mann, and H. Hofmann. 2003. Laboratory acquired infection with recom-binant vaccinia virus containing an immunomodulating construct. J. Inves-tig. Dermatol. 120:356–358.

90. Michael, A., R. Stratford, S. Khan, A. Dalgleish, and H. Pandha. 2004.Novel strains of Salmonella typhimurium as potential vectors for gene de-livery. FEMS Microbiol. Lett. 238:345–351.

91. Miller, C. D., J. R. Songer, and J. F. Sullivan. 1987. A twenty-five yearreview of laboratory-acquired human infections at the National AnimalDisease Center. Am. Ind. Hyg. Assoc. J. 48:271–275.

92. Mortland, K. K., and D. B. Mortland. 2003. Biosafety considerations inlaboratory design. Clinical leadership and management. 17:175–177.

93. Moss, B. 1996. Genetically engineered poxviruses for recombinant geneexpression, vaccination, and safety. Poc. Natl. Acad. Sci. USA 93:11341–11348.

94. Moussatche, N., M. Tuyama, S. E. Kato, A. P. Castro, B. Njaine, R. H.Peralta, J. M. Peralta, C. R. Damaso, and P. F. Barroso. 2003. Accidentalinfection of laboratory worker with vaccinia virus. Emerg. Infect. Dis.9:724–726.

95. Munro, K., J. Lanser, and R. Flower. 1999. A comparative study of methodsto validate formaldehyde decontamination of biological safety cabinets.Appl. Environ. Microbiol. 65:873–876.

96. Nahrstedt, H., J. Waldeck, M. Grone, R. Eichstadt, J. Feesche, and F.Meinhardt. 2005. Strain development in Bacillus licheniformis: constructionof biologically contained mutants deficient in sporulation and DNA repair.J. Biotechnol. 119:245–254.

97. Nalca, A., A. W. Rimoin, S. Bavari, and C. A. Whitehouse. 2005. Reemer-gence of monkeypox: prevalence, diagnostics, and countermeasures. Clin.Infect. Dis. 41:1765–1771.

98. Nelson, C. H. 2001. Risk perception, behavior, and consumer response togenetically modified organisms. Toward understanding American and Eu-ropean public reaction. Am. Behav. Sci. 44:1371–1388.

99. Nikiforov, V. V., I. U. I. Turovskiı, P. P. Kalinin, L. A. Akinfeeva, L. R.Katkova, V. S. Barmin, E. I. Riabchikova, N. I. Popkova, A. M. Shestopalov,V. P. Nazarov, et al. 1994. A case of a laboratory infection with Marburgfever. Zh. Mikrobiol. Epidemiol. Immunobiol. 3:104–106. (In Russian.)

100. Normile, D. 2003. SARS experts want labs to improve safety practices.Science 302:31.

101. Normile, D. 2004. Infectious diseases. Second lab accident fuels fears aboutSARS. Science 303:26.

102. Normile, D. 2004. Severe acute respiratory syndrome: lab accidents promptcalls for new containment program. Science 304:1223–1225.

103. Noviello, S., R. Gallo, M. Kelly, R. J. Limberger, K. DeAngelis, L. Cain, B.Wallace, and N. Dumas. 2004. Laboratory-acquired brucellosis. Emerg.Infect. Dis. 10:1848–1850.

104. Nulens, E., and A. Voss. 2002. Laboratory diagnosis and biosafety issues ofbiological warfare agents. Clin. Microbiol. Infect. 8:455–466.

105. Openshaw, P. J., W. H. Alwan, A. H. Cherrie, and F. M. Record. 1991.Accidental infection of laboratory worker with recombinant vaccinia virus.Lancet 338:459.

106. O’Reilly, D. R., L. K. Mileer, and V. A. Luckow. 1994. Baculovirus expres-sion vectors: a laboratory manual. Oxford University Press, New York, NY.

107. Osborne, R., T. Durkin, H. Shannon, E. Dornan, and C. Hughes. 1999.Performance of open fronted microbiological safety cabinets: the value ofoperator protection tests during routine servicing. J. Appl. Microbiol. 86:962–970.

108. Paoletti, E. 1996. Applications of poxvirus vectors to vaccination: an up-date. Proc. Natl. Acad. Sci. USA 93:349–353.

109. Peeters, B., A. Bouma, T. de Bruin, R. Moormann, A. Gielkens, and T.Kimman. 1994. Non-transmissible pseudorabies virus gp50 mutants: a newgeneration of safe live vaccines. Vaccine 12:375–380.

110. Perfetto, S. P., D. R. Ambrozak, R. A. Koup, and M. Roederer. 2003.Measuring containment of viable infectious cell sorting in high-velocity cellsorters. Cytometry 52A:122–130.

111. Perfetto, S. P., D. R. Ambrozak, M. Roederer, and R. A. Koup. 2004. Viableinfectious cell sorting in a BSL-3 facility. Methods Mol. Biol. 263:419–424.

112. Phillips, G. B. 1965. Microbiological hazards in the laboratory. I. Control.J. Chem. Educ. 42A:43–48.

113. Phillips, G. B., and R. S. Runkle. 1967. Laboratory design for microbiolog-ical safety. Appl. Microbiol. 15:378–389.

114. Pike, R. M. 1976. Laboratory-associated infections: summary and analysisof 3921 cases. Health Lab. Sci. 13:105–114.

424 KIMMAN ET AL. CLIN. MICROBIOL. REV.

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from

115. Pike, R. M. 1978. Past and present hazards of working with infectiousagents. Arch. Pathol. Lab. Med. 102:333–336.

116. Pike, R. M. 1979. Laboratory-associated infections: incidence, fatalities,causes, and prevention. Annu. Rev. Microbiol. 33:41–66.

117. Richmond, J. Y., R. C. Knudsen, and R. C. Good. 1996. Biosafety in theclinical mycobacteriology laboratory. Clin. Lab. Med. 3:527–550.

118. Romano, G., C. Pacilio, and A. Giordano. 1999. Gene transfer technologyin therapy: current applications and future goals. Stem Cells 17:191–202.

119. Rusnak, J. M., M. G. Kortepeper, R. J. Hawley, A. O. Anderson, E. Bou-dreau, and E. Eitzen. 2004. Risk of occupationally acquired illnesses frombiological threat agents in unvaccinated laboratory workers. Biosecur. Bio-terror. 2:281–293.

120. Samulski, R. J., M. Sally, and N. Muzyczka. 1999. Adeno-associated viralvectors, p. 131–172. In T. Friedmann (ed.). The development of genetherapy. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

121. Sargent, E. V., and F. Gallo. 2003. Use of personal protective equipment forrespiratory protection. ILAR J. 44:52–56.

122. Sawyer, J., A. Bennet, V. Haines, E. Elton, K. Crago, and S. Speight. 2007.The effect of microbiological containment systems on dexterity. J. Occup.Environ. Hyg. 4:166–173.

123. Schellekens, H. 2001. Veilig werken met microorganismen, parasieten encellen in laboratoria en andere werkruimtes. Nederlandse Vereniging voorMicrobiologie, Bilthoven, The Netherlands.

124. Schmid, I., A. Kunkl, and J. K. Nicholson. 1999. Biosafety considerationsfor flow cytometric analysis of human immunodeficiency virus-infectedsamples. Cytometry 38:195–200.

125. Schmid, I., S. Merlin, and S. P. Perfetto. 2003. Biosafety concerns forshared flow cytometry core facilities. Cytometry A 56:113–119.

126. Schmid, I., C. Lambert, D. Ambrozak, G. E. Marti, D. M. Moss, S. P.Perfetto, and the International Society of Analytical Cytology. 2007. Inter-national Society for Analytical Cytology biosafety standard for sorting ofunfixed cells. Cytometry A 71:414–437.

127. Schou, S., and A. K. Hansen. 2000. Marburg and Ebola virus infections inlaboratory non-human primates: a literature review. Comp. Med. 50:108–123.

128. Schweder, T., I. Schmidt, H. Herrmann, P. Neubauer, M. Hecker, and K.Hofmann. 1992. An expression vector system providing plasmid stabilityand conditional suicide of plasmid-containing cells. Appl. Microbiol. Bio-technol. 38:91–93.

129. Seidler, A., A. Nienhaus, and R. Driel. 2005. Review of epidemiologicalstudies on the occupational risk of tuberculosis in low-incidence areas.Respiration 72:431–446.

130. Sejvar, J. J., D. Johnson, T. Popovic, J. M. Miller, F. Downes, P. Somsel, R.Weyant, D. S. Stephens, B. A. Perkins, and N. E. Rosenstein. 2005. Assess-ing the risk of laboratory-acquired meningococcal disease. J. Clin. Micro-biol. 43:4811–4914.

131. Sewell, D. L. 1995. Laboratory-associated infections and biosafety. Clin.Microbiol. Rev. 8:389–405.

132. Smith, C. E., D. I. Simpson, E. T. Bowen, and I. Zlotnik. 1967. Fatal humandisease from Vervet monkeys. Lancet ii:119–1121.

133. Sørensen, T. U., G. J. Gram, S. D. Nielsen, and J. E. Hansen. 1999. Safesorting of GFP-transduced live cells for subsequent culture using a modifiedFACS vantage. Cytometry 37:284–290.

134. Spina, N., S. Zansky, N. Dumas, and S. Kondracki. 2005. Four laboratory-associated cases of infection with Escherichia coli O157:H7. J. Clin. Micro-biol. 43:2938–2939.

135. Srinivasan, A., C. N. Kraus, D. DeShazer, P. M. Becker, J. D. Dick, L.Spacek, J. G. Bartlett, W. R. Byrne, and D. L. Thomas. 2001. Glanders ina military research microbiologist. N. Engl. J. Med. 345:256–258.

136. Steidler, L., S. Neirynck, N. Huyghebaert, V. Snoeck, A. Vermeire, B.Goddeeris, E. Cox, J. P. Remon, and E. Remaut. 2003. Biological contain-ment of genetically modified Lactococcus lactis for intestinal delivery ofhuman interleukin 10. Nat. Biotechnol. 21:785–789.

137. Steidler, L. 2004. Live genetically modified bacteria as drug delivery tools:

at the doorstep of a new pharmacology? Expert Opin. Biol. Ther. 4:439–441.

138. Sulkin, S. E., and R. M. Pike. 1951. Laboratory infections. Science 114:3.139. Sulkin, S. E., and R. M. Pike. 1951. Survey of laboratory-acquired infec-

tions. Am. J. Public Health Nations Health 41:769–781.140. Swanepoel, R. 2001. Marburg hemorrhagic fever, 1967 to the present.

ProMED-mail, 7 February 2001. 20010207.0242. http://www.promedmail.org. Accessed 22 September 2007.

141. Tacket, C. O., S. M. Kelly, F. Schodel, G. Losonsky, J. P. Nataro, R.Edelman, M. M. Levine, and R. Curtiss. 1997. Safety and immunogenicityin humans of an attenuated Salmonella typhi vaccine vector strain express-ing plasmid-encoded hepatitis B antigens stabilized by the Asd-balancedlethal vector system. Infect. Immun. 65:3381–3385.

142. Thompson, R. J., H. G. Bouwer, D. A. Portnoy, and F. R. Frankel. 1998.Pathogenicity and immunogenicity of a Listeria monocytogenes strain thatrequires D-alanine for growth. Infect. Immun. 66:3552–3561.

143. Tumpey, T. M., C. F. Basler, P. V. Aguilar, H. Zeng, A. Solorzano, D. E.Swayne, N. J. Cox, J. M. Katz, J. K. Taubenberger, P. Palese, and A.Garcıa-Sastre. 2005. Characterization of the reconstructed 1918 Spanishinfluenza pandemic virus. Science 310:77–80.

144. Vaquero, M., P. Gomez, M. Romero, and M. J. Casal. 2003. Investigation ofbiological risk in mycobacteriology laboratories: a multicentre study. Int. J.Tuberc. Lung Dis. 7:879–885.

145. VROM. 2004. Integrale versie van de Regeling genetisch gemodificeerdeorganismen en het Besluit genetische gemodificeerde organismen. Minis-terie voor Volksgezondheid, Ruimtelijke Ordening en Milieu, Den Haag,The Netherlands.

146. Walker, D., and D. Campbell. 1999. A survey of infections in UnitedKingdom laboratories, 1994–1995. J. Clin. Pathol. 52:415–418.

147. Wedum, A. G. 1953. Bacteriological safety. Am. J. Public Health NationsHealth 43:1428–1437.

148. Wedum, A. G., E. Hanel, G. Briggs Phillips, and O. T Miller. 1956. Labo-ratory design for study of infectious disease. Am. J. Public Health NationsHealth 46:1102–1113.

149. Wedum, A. G. 1961. Control of laboratory airborne infection. Bacteriol.Rev. 25:210–216.

150. Wedum, A. G. 1964. Laboratory safety in research with infectious aerosols.Public Health Rep. 79:619–633.

151. Wedum, A. G. 1964. Airborne infection in the laboratory. Am. J. PublicHealth Nations Health 54:1669–1673.

152. WHO. 2004. Laboratory biosafety manual, 3rd ed. World Health Organi-zation, Geneva, Switzerland.

153. Williams, N. R., and B. M. Cooper. 1993. Counselling of workers handlingvaccinia virus. Occup. Med. (London) 43:125–127.

154. Williams-Smith, H. 1978. Is it safe to use Escherichia coli in recombinantDNA experiments? J. Infect. Dis. 137:655–660.

155. Willshaw, G. A., J. Thirlwell, A. P. Jones, S. Parry, R. L. Salmon, and M.Hickey. 1994. Vero cytotoxin-producing Escherichia coli O157 in beefburg-ers linked to an outbreak of diarrhoea, haemorrhagic colitis and haemolyticuraemic syndrome in Britain. Lett. Appl. Microbiol. 19:304–307.

156. Wilson, D. E., and L. C. Chosewood, ed. 2007. Biosafety in microbiologicaland biomedical laboratories, 5th ed. CDC and NIH, Washington, DC.www.cdc.gov/od/ohs/biosfty/bmbl5/bmbl5toc.htm.

157. Reference deleted.158. Wlodaver, C. G., G. J. Palumbo, and J. L. Waner. 2004. Laboratory-ac-

quired vaccinia infection. J. Clin. Virol. 29:167–170.159. Yagupsky, P., N. Peled, K. Riesenberg, and M. Banai. 2000. Exposure of

hospital personnel to Brucella melitensis and occurrence of laboratory-acquired disease in an endemic area. Scand. J. Infect. Dis. 32:31–35.

160. Yamashita, G. M., T. Misawa, T. Komori, K. Okuzumi, and T. Takahashi.2007. Current biosafety in clinical laboratories in Japan: report of ques-tionnaire data obtained from clinical laboratory personnel in Japan.Kanesnhogaku Zasshi 81:39–44.

VOL. 21, 2008 EVIDENCE-BASED BIOSAFETY 425

on February 3, 2021 by guest

http://cmr.asm

.org/D

ownloaded from