4th National Symposium on Biosafety

Research with Small Animals

James Fox, DVM
Professor and Director
Division of Comparative Medicine
Massachusetts Institute of Technology
77 Massachusetts Ave., 45-106
Cambridge, MA 02139
617-253-1757

It is a pleasure to be here in the mecca of infectious diseases, where containment, and issues of safety in the research laboratory have had such a high profile for so many years. We all know that Atlanta is the home of CDC; indeed the CDC has provided to all of us guidance over the last several decades in terms of the biosafety issues, many of which we're dealing with today and certainly it is fitting to be here under the umbrella of that organization as we work through new and emerging issues that we are facing daily in terms of bio-containment and biohazard control.

I'm going to focus today on small laboratory animals, primarily rodents for the purpose of this presentation. It's always beneficial to have a historical perspective in terms of assessing where we are today. Pioneers in this area had the foresight to tabulate for us on a longitudinal basis infections acquired in the laboratory. These data were primarily published by Pike and Sulken. They initiated a series of papers in 1949 and followed the tabulation of cases through 1978. There were a wide array of infections recorded in the laboratory environment; close to 4,000 at the time of the final tabulation which consisted of almost 30 years of tracking the incidence of infections. Bacterial infections, one of which in the veterinary profession we're very cognizant of, that is Brucella species. Also infections were acquired from other etiological agents that were handled frequently in the laboratories at that time. One take home message of this chronoogy of acquired infections is something that you should always be aware of: less than 20% of all the 4,000 cases reported could be traced to a known accident in the laboratory.

Since 1978 the order of frequency of infectious agents acquired in the laboratory are bacterial, viral, fungal and rickettesae. Certainly Brucella ranks high in its ability to infect laboratory personnel. With viral infections VEE is listed among the top offenders. Fungal infections were also acquired in the laboratory, particularly Coccidioides immitis; and one which many of us have had personal experience with is dermatomycosis, a zoonotic disease in laboratory animals which is well documented in the literature. The development of rickettseal infections also occurs; Q fever is a notable example especially in personnel working with sheep and goats. Dr. Dan Liebermann compiled in his new second edition of "Biohazards Management Handbook," some 375 infections from 1978 to 1991. Again these were categorized by agent, i.e. bacterial, viral and rickettseal. Salmonella sp. and Brucella sp. rank high on the list. In the viral area, we see new agents such as the virus Herpes simiae and also again the preponderance of rickettseal infections, being Q fever.

It is also important when you're developing a risk assessment plan or a biohazard control program within your animal facilities to focus on at the individuals at risk. I've presented for you two studies that looked at the distribution of infections according to occupation. As would be expected, the principal investigator and the research technician are at high risk of exposure. Even though these individuals supposedly know they're working with an infectious agent and are trained in proper techniques, they still are at the highest risk of contracting the disease from working with an infectious agent. Also its important for us to realize in terms of our training strategies, that animal technicians are also at risk of being infected by these various etiological agents. Unfortunately, innocent bystanders, particularly physical plant personnel are at risk of becoming infected due to transient exposure to a variety of infectious agents.

Knowing the infectious dose of a microorganism is also important when one considers the risk of a particular infectious agent being handled in a bio-containment mode. You should ask what the infectious dose is for man and particularly what is the infectious dose for the laboratory animal that is being experimentally infected. Additionally, upon dosing, how will the agent be shed and potentially further expose individuals or the environment. For example, a high frequency of Q fever and tularemia are being documented in individuals working in the laboratory. A large component of the risk of contracting these diseases is the small number of organisms that are actually required to infecta given individual.

The theme of the decade, as we've heard from the last speaker, Dr. Peters, is new and re-emerging infectious diseases. These certainly are playing a pivotal role in how we assess biohazards in animal facilities today because in many cases these agents have never been worked with in a laboratory animal setting. In addition, because they are newly identified, we don't know all of the particulars about transmission. Certainly high on the list is HIV, Ebola virus and the new strain of zoonotic Haantaan virus recently discovered in wild rodents the Southwestern part of the United States.

Also there is a resurgence of interest in tuberculosis; indeed, numerous laboratories are refocusing on the pathogenesis of tuberculosis and mechanisms of acquired drug resistance. Certainly an organism that our laboratory has been studying, Helicobacter pylori is being investigated in many laboratories internationally. This bacteria is now known to cause not only peptic ulcer disease in humans but is also directly linked to gastric cancer. If you inquire in a biohazard text regarding this agent's risk, the name Helicobacter pylori does not appear. It does not apper because it was not discovered until 1982 and did not receive considerable importance and prominence until the last 3 or 4 years.

One of the biggest challenges in my mind in terms of infectious disease studies in the laboratory animal environment is the advent of new genetically defined laboratory mice being used in animal facilities worldwide. As an example, at MIT some 6 years ago we had about 10,000 mice on a daily basis. Today our census of mice approaches 50,000 and of those, 95% fall in the category of being genetically manipulated. Because these mice are in many cases immunocompromised, one has to rethink the issues of bio-containment and spread of infectious agents that are administered into these genetically manipulated mice. Of course in many laboratories, HIV is being studied. Because SCID mice with combined immunodeficiency are being used for HIV studies, the central question is, as it should be with all infectious agents, what hazards does this agent pose when the agent is injected into a defined animal system.

In a publication in the ASM News in 1990 the authors made the statement in the title of their paper that the hazards of using HIV are no greater than those of human or in vitro studies. They went on to defend this premise by asking two central questions about HIV replication in SCID mice and whether the interactions between endogenous viruses of mice and HIV might produce pseudotypes. The latter question was addressed effectively showing that this possibility was not likely in the SCID model. For assessment of actual tissue culture infectious doses the authors compared the level of viremia in these SCID infected mice with what one would expect from a patient infected with HIV and infectious particles that would be expected in the patient's plasma. They also assessed the risk of aerosol transmission. We've heard that theme recurring throughout this morning's presentations about the risk of aerosol transmission and how that amplifies potential exposure in a laboratory setting. Thus the investigators recommended that BSL2 facilities, employing BSL3 practices, be used with this particular agent in this mouse model system. This exercise should be undertaken for each of the biohazard applications that are ongoing in your facility. It's important to remember the variables with each animal host system and the agent employed.

If one does a literature search they can certainly appreciate how the use of transgenic mice has exploded within the last several years. As expected, transgenic mouse model systems are being used for infectious disease research as exemplified by a paper published where they used transgenic to study polio virus pathogeneses. The reason the experiment was achievable was they coded the human polio virus receptor gene into the mouse genome; therefore; experimental infection does take place with these transgenic animals. This recurring theme exists throughout transgenic literature where researchers have been able to insert particular gene fragments into the mouse genome allowing receptor studies or other genetic manipulations within thereby making the mouse receptive to a number of infectious diseases that heretofore were not a concern within the biological laboratory because of their inability to be manipulated in vivo in rodents.

Another example of modern molecular biology is the use of recombinant vectors; one of the favorites is vaccinia virus research which has been genetically engineered to contain and express foreign DNA.

Unfortunately vaccinia virus has zoonotic potential and as early as 1980 an adhoc Immunization Committee recommended the use of vaccinia virus vaccines to protect laboratory workers. In 1988, the Committee recommended that personnel working with laboratory animals and these recombinant vaccinia vectors be vaccinated, the rationale being that laboratory infections have been documented with this particular construct. Therefore it is recommended that investigators working with recombinant vaccinia vectors in your laboratory should seek medical avice regarding possible vaccinia vaccination. Of course the infectious disease physician should take primary responsibility for decisions regarding vaccination protocols.

This table (referring to slide) is taken from page 130 of the 3rd edition of the CDC biohazard guideline. The class 3 agents recommended for BSL3 containment are listed. Lymphocytic choriomeningitis and Haantaan virus are on that list. I want to concentrate as a case study today on the issue of zoonotic risk of lymphocytic choriomeningitis infection in laboratory rodents.

Fortunately, many of us have not personally experienced outbreaks of LCM in mice or hamsters in our animal facilities. However, if you examine the literature, it is readily apparent that this viral agent has caused numerous outbreaks both in the laboratory as well as in the public domain. There have been at least three distinct outbreaks in humans in research institutions working with infected tumor cell lines in hamsters, documented over the last some 20 to 25 years. I want to stress that LCM infections in rodents are still occurring, as indicated by the latest publication appearing in 1992. This paper details an outbreak of LCM in laboratory workers manipulating nude mice. As you know, immunocompromised animals often cannot mount a sufficient immune response to clear viral infections. Therefore, nude mice are persistently infected with LCM. The index case in this outbreak was a laboratory animal technician handling LCM infected nude mice. This individual presented with non-specific, signs but the symptoms were compatible with a viral infection; i.e. fever, malaise, severe headache, chest pain, photophobia, coughing, and vomiting. The physicians who authored this paper, recommended that anyone with a diagnosis of severe viral infection or aseptic meningitis, especially those working with laboratory rodents, should have LCM included in the differential diagnosis.

In reviewing the risk and exposure of LCM in this outbreak, it's clear from the material derived of the questionnaires and personal interviews conducted with the laboratory technicians, that the individuals at highest risk were those individuals working directly with the animals. Included were those personnel cleaning cages, animal bedding or changing the animals' water. Statistically, of the 90 people interviewed, those who had seroconverted and had clinical signs of LCM were in this occupational risk category.

Interestingly, the laboratory had been working with tumor cell lines that had been historically passed in hamsters. However, during the last several years, a decision had been made to passage the tumors in nude mice. Prior to this time there had been no apparent LCM infections in the laboratory personnel. It wasn't until 1987 through 1989 that the number of nude mice dramatically increased in their research program. In addition, the length of time the tumors were passed in these animals had been increased prior to and during the time the LCM outbreak occurred. This interval corresponded with the increased LCM exposure of the laboratory personnel working with nude mice in the laboratory. Thus, a change in the direction of the research program which mandated using nude mice in much larger numbers and for a longer period of time, translated into greater LCM exposure and subsequent human LCM infections. In all probability, the nude mice were being infected with a tumor cell line contaminated with LCM. This continual exposure to LCM combined with possible LCM horizontal transmission between mice probably established the persistent murine LCM infection within the mouse colony. The quantity of LCM was substantially increased because of rapid expansion of the nude mice colony as well as maintaining the mice longer in the laboratory. The LCM contaminated cell line had probably been introduced into the laboratory some 20 years earlier and had been serially passed in hamsters. We could theorize that there had been spot occurrences of LCM infections in personnel during that interval. However, it wasn't until the increased use of the tumor cell line in the nude mic that a documented outbreak of LCM infections occurred.

As illustrated by this LCM outbreak, the use of transplantable tumors in biomedical research requires routine identification and viral screening of murine cell lines being passaged in laboratory rodents. The other disturbing fact discussed in this LCM outbreak was that sentinel animals had been placed in the animal facilities some 2 to 3 years prior to the LCM outbreak. For unknown reasons there were no LCM seroconversion noted in the sentinel mice. A sentinel animal program is certainly justified and important. But this example of failure to detect LCM points out the necessity of developing, assessing and monitoring sentinel animal programs to ensure their accuracy in detecting intercurrent disease in animals being maintained for biomedical research. Certainly, one of the primary reasons we develop a sentinel animal program is to prevent possible introduction of zoonotic agents into animals being used in our facilities.

This raises a pertinent question - are we performing tasks and instituting policies in our laboratory animal facilities to minimize zoonotic transmission? Certainly there are preventative measures that Emmett Barkley referred to briefly this morning in his presentation. An established quarantine as part of this program should be emphasized. Access to state of the art diagnostic facilities to allow microbiological assessment of animals on a routine basis is also a very important part of your biohazard program. Furthermore, personnel education is a fundamental building block on which a successful biohazard control program is sustained as is access to medical care and personnel health screening of individuals working in your facilities.

The animal care and use program should be defined as part of an infrastructure that includes many disciplines and many areas of expertise. This includes involvement of medical department personnel and as well as a well staffed biohazard office, in addition to other groups that deal with occupational health such as radiation safety and/or industrial hygiene. All of these individuals have to work together to compliment each other's expertise to help ensure a successful biohazard safety program. As part of that process, it's very important to have a very well defined tracking system for biohazard experiments in animals. Investigators that are using hazardous materials at MIT are required to fill out protocols to allow proper assessment of infectious or chemical agents as well as hazard potential, manipulation, etc. Importantly, the protocol has to have the signature of the principal investigator to assure us that he or she has taken direct responsibility for the conduct of a given experiment involving hazardous agents.

Having appropriate literature available to you for evaluation of hazardous agents is an indispensable part of the program. The 3rd edition of the CDC manual which I've referred to earlier is an obvious part of your library. You should also have the 1st and the 2nd editions on your bookshelf. To summarize, it's important that individuals involved in the infrastructure of biosafety and occupational health are conversant in four critical elements of the program:

Let me conclude with this quote of a famous scientist. "The ingenuity, knowledge and organization of humans can alter but cannot cancel humanity's vulnerability to invasion by parasitic forms of life. Infectious disease which antedated the emergence of human kind will last as long as humanity itself and will surely remain as it has heretofore one of the most fundamental parameters and determinants of human history." This prophetic statement has been clearly illustrated by Dr. Peters in the previous presentation. In terms of the laboratory environment however, we can certainly define, in many instances, exposure risk to infectious agents and under what conditions the agents are to be manipulated and thereby hopefully devise strategies to minimize the risk of laboratory acquired infections to personnel.

Office of Health and Safety, Centers for Disease Control and Prevention, 1600 Clifton Road N.E., Mail Stop F05 Atlanta, Georgia 30333, USA Last Modified: 1/2/97
Send us your Comments.

OhASIS HOME | Biosafety Information | Safety Manuals ||| Symposium Contents

Research with Small Animals