The Risks of Using Live Vaccines

Things you should know about the risks of live vaccines, not only to the animal who has been vaccinated, but also to anyone coming in contact with the animal, written by Ian Tizard, BVMS, PhD.

JAVMA, Vol. 196, No. 11, June 1, 1990

Review Article Pgs 1851-1858

From the Department of Veterinary Microbiology and Parasitology, College of Veterinary Medicine, Texas A&M University. College Station, TX 77843

Although the veterinary medical profession has available a wide range of safe, effective vaccines, it is often necessary to choose between modified-live and inactivated products. The outcome of this choice, commonly based on economic considerations as well as on efficacy, has resulted in the widespread and deliberate use of modified-live vaccines in animals. Unfortunately, modified-live vaccines as a class, present considerable hazards, not only to the vaccinated animal, but also to its contacts. These hazards take 2 forms: disease attributable to residual virulence and disease attributable to contamination.

The process of attenuation is designed to develop an organism that can transiently establish itself in a vaccinated animal, while at the same time, not cause disease. The vaccine manufacturer, therefore, seeks to achieve minimal virulence while retaining maximal immunogenicity. This balance may be readily achievable in clinically normal, healthy animals, but may be unattainable in animals with even minor defects in immune competence. Marginally immunodeficient state may develop simply as a result of normal biological variation. Thus, the range of immunologic competence in a population of clinically normal animals will be such that some will inevitably be naturally susceptible to the otherwise avirulent organisms found in vaccines. Minor suppression of immune competence may result from many forms of stress, including transportation and surgery. More importantly, common viral infections, such as feline leukemia, or canine distemper, are associated with appreciable immunosuppression. Even subclinical infection may make an animal susceptible to the pathogenic effects of attenuated organisms from vaccines. Thus, residual virulence and the morbidity resulting from it are inescapable consequences of the use of modified-live vaccines.

A related problem is the vaccine,s ability to cause disease if given by an inappropriate route or to an animal of the wrong species. Vaccine organisms are usually attenuated for a specific species or for a specific route of administration. If given inappropriately, residual virulence may cause disease.

The second major problem with modified-live vaccines is associated with the need to use them as cultures of living organisms. They can, therefore, contain no preservatives. As a result, their potential for contamination with other infective agents is great. The inadvertent use of a contaminated vaccine may serve to disseminate other infective agents through an animal population.

Described herein are cases in which considerable morbidity and mortality have been caused by modified-live-vaccines, either as a result of residual virulence, or of inadvertent contamination with other infective agents.

Residual Virulence

Veterinarians still use vaccines that contain known virulent organisms. Examples of these include the vaccines against contagious ecthyma (orf), piroplasmosis, bluetongue, and Brucella abortus infection. Strain 19 of B abortus is a highly immunogenic organism that induces lifelong immunity in vaccinated cows and successfully prevents abortion. Therefore, strain 19 may be regarded as an effective vaccine. Strain 19 can, however, also cause severe systemic reactions in vaccinated animals; local swelling, high fever, anorexia, listlessness, and decrease in milk yield have all been reported. Strain 19 may also cause abortion if given to pregnant cattle and can cause severe orchitis in bulls, as well as undulant fever in human beings. In deciding to use strain 19, society has decided that the benefits of its use outweigh its risks, although strenuous efforts have been made to find a safer replacement.

Fetal susceptibility – When modified-live vaccines are administered to healthy adult animals, their residual virulence is controlled by an effective immune system. If a modified-live vaccine is given to a pregnant animal and these organisms gain access to the fetus, the immature fetal immune system may be unable to control the resulting infection. Thus, some live vaccines, although safe for use in nonpregnant animals, may cause fetal infection resulting in malformations or abortion. Examples of such vaccines include bluetongue virus vaccine in lambs, rubella virus vaccines in people, and some Erysipelothrix rhusiopathiae vaccines in swine.

Modified-live bluetongue virus vaccines cause congenital anomalies in fetal lambs. The severity of these lesions is determined by the time of exposure to the vaccine. Bluetongue virus vaccine given to pregnant ewes between 50 and 100 days after conception causes hydranencephaly and retinal dysplasia in fetal lambs. The vaccine virus can be readily isolated from the lamb tissues.

Live E rhusiopathiae vaccine has been reported to cause abortion in pregnant sows. In one reported case, the vaccine was administered to sows 30 days before the expected farrowing date. Many of the sows aborted, giving birth to dead and dying pigs. The organism, probably derived from the vaccine strain, was isolated from the aborted pigs.

Immunosuppression - Immunosuppression is a common consequence of virus infections in animals. In people, acquired immunodeficiency syndrome is a similar example, and there are many such examples in domestic animals, including canine distemper (CD) and feline lentivirus, feline leukemia, feline panleukopenia, and possibly, canine parvovirus (CPV) infections. Although immunosuppression is usually associated with infection by virulent virus, some attenuated vaccine strains of virus may also be able to provoke mild immunosuppression. Immunosuppression may also be induced, not only by individual live virus vaccines, but also by vaccine combinations. Thus, a combination of CD virus (CDV) and canine adenovirus type-1 or type-2 vaccines may induce suppression of blood lymphocyte counts in dogs, as well as lymphocyte responsiveness to mitogens, although the individual vaccines are not detectably immunosuppressive when given alone.

Consequences of use of some modified-live virus vaccines

Canine parvovirus - There is a disagreement about the immunosuppressive properties of CPV. Some investigators have suggested that CPV is immunosuppressive and that modified-live CPV vaccines (ML-CPV) may induce immunosuppression in dogs. For example, ML-CPV consistently caused transient depression of the response of canine lymphocytes to concanavalin A. The depressed response lasted for 2 to 5 weeks before returning to normal. A mixture of ML-CPV and ML-CDV depressed the response to concanavalin A and phytohemagglutinin (PHA) in 3 of 8 dogs, although the response to PHA was erratic. The antibody response to CDV was lower in dogs given the vaccine and seemed to correlate with the effects on lymphocytes. Some investigators claim that ML-CPV causes lymphocytopenia in dogs 3 days after vaccination, although others have failed to document this.

Canine distemper virus – Modified-live CD vaccine virus multiplies in dogs, causing immunosuppression and thrombocytopenia. Thrombocytopenia develops 1 week after vaccination. Although frank bleeding has not been observed, it may be appropriate to avoid performing elective surgery on dogs within a week after ML-CDV administration.

In Australia, in 1968 and 1970, the use of certain batches of ML-CDV in dogs was associated with the development of clinical distemper. Dogs were administered a combination modified-live attenuated CDV/canine adenovirus type-1 vaccine produced in canine kidney tissue culture. Eight to 18 days later, the dogs developed anorexia, listlessness, fever, ocular discharge, and diarrhea. The vaccine was especially virulent for Greyhounds, and up to 50% of vaccinated Greyhounds became sick. Development of nervous system lesions was reflected by observation of signs of aggression, which eventually progressed to incoordination, prostration, convulsions, and death. Dogs survived from 11 to 24 days after onset of signs. At necropsy, eosinophilic intranuclear and cytoplasmic inclusions were found in large neurons, along with nonsuppurative encephalitis and malacia in the ventral pontine gray matter. Electron microscopy revealed nucleocapsids indistinguishable from those of CDV lesions. A single similar case was reported in West Germany. Although it is most probable that the ML-CDV caused these encephalitides, the possibility exists, in view of the failure to isolate CDV, that these infections were attributable to action of a separate agent provoked by vaccination.

The commonly used Snyder Hill vaccine strain of CDV causes disease in exotic animals. For example, ML-CDV given to gray foxes, resulted in all the foxes dying of the vaccine-induced distemper. The disease was also transmitted from vaccinated to unvaccinated foxes. The virus isolated from these foxes provoked a mild febrile illness in pups and could be detected in conjunctival smears by use of immunofluorescence. Interestingly, the original vaccine virus did not cause fever and could not be detected in conjunctival smears, suggesting that single passage through the foxes had caused considerable reversion to virulence.

In a similar situation, ML-DCV was administered to young kinkajous (Potos flavus), causing them to develop diarrhea and CNS disease. They developed interstitial pneumonia and encephalitis characterized by the presence of lymphoid cells and many CDV intranuclear inclusions. Fatal disease was also induced in lesser pandas (Ailurus fulgens) 2 weeks after they were given a dose of ML-CDV. The pandas developed giant cell pneumonia with CDV inclusion bodies in the pulmonary and digestive tract epithelia. Vaccine strain CDV could be isolated from these pandas. Four endangered black-footed ferrets (Mustela nigripes) died 21 days after vaccination with ML-CDV. All clinical signs of disease and necropsy findings were compatible with a diagnosis of distemper.

The virulence of CDV appears to be directly related to its ability to replicate in the dog macrophages. Attenuated CDV adapted to grow in canine kidney cells is unable to grow in canine lung macrophage-cultures. This is however, an unstable characteristic, and an attenuated vaccine strain of CDV (Rockborn strain) required its ability to grow in canine lung macrophages after as few as 3 serial passages in dogs. In those dogs, weight loss was observed after 4 passages, reduction in lymphocyte blasto-genesis was seen after 5 passages, viral antigen was detected in the conjunctive after 6 passages. Thus after approximately 6 sequential passages, the CD vaccine virus was no longer avirulent. A similar phenomenon was seen in the ferrets; after 14 sequential passages, the attenuated CDV strain regained virulence and, by 22 passes, it was lethal for them.

Foot-and-mouth disease – Modified-live foot-and-mouth disease virus vaccines are used in several countries. They are effectively attenuated for clinically normal cattle, but have been reported to induce clinical disease in swine or cattle under stress. For example, a group of high-producing Friesian cows developed vesicles on teats after administration of an otherwise acceptably attenuated foot-and-mouth disease virus vaccine.

Bovine viral diarrhea virus – Modified-live vaccine strains of bovine viral diarrhea virus (BVDV) may cause appreciable reactions. Some strains of modified-live BVD vaccine viruses are immunosuppressive in calves, so that vaccination with attenuated vaccine strains may potentiate intercurrent infections. These vaccine strains cause a decrease in circulation lymphocyte and neutrophil numbers, suppression of lymphocyte responses to mitogens such as PHA and pokeweed mitogen, and suppression of neutrophil iodination and antibody-dependent cellular cytotoxicity. Thus, these vaccine strains of BVDV have retained their ability to induce prolonged suppression of hose defense mechanisms.

Bovine viral diarrhea bears a unique relationship to mucosal disease of cattle. Clinical mucosal disease develops only in cattle that are immunotolerant to and persistently infected with BVDV. These cattle may be clinically normal when born, but develop fatal mucosal disease after superinfection with another pestivirus. If not so infected, they remain clinically normal. As a result, administration of ML-BVDV may induce clinical mucosal disease. Postvaccinal mucosal disease develops 7 to 20 days after inoculation. Affected cattle are depressed, and anorectic, with oculonasal discharge, lameness, fever, leucopenia, and eventually, water diarrhea. Death occurs 10 to 14 days after the first signs of illness appear. Morbidity is usually around 0.2% of vaccinated cattle. In some severely affected herds, however, the incidence of mucosal disease may reach as high as 5%.

Although the use of some combination BVDV bovine herpesvirus vaccines has been followed in some cattle by a severe condition resembling BVD,it is not necessarily because of contamination. It may be attributable to the incubations period for BVD overlapping vaccination.

Rabies virus – The progressive decrease in the use of modified-live rabies virus vaccines testifies to the problems encountered in their use. The most important of these problems is residual virulence. Thus, postvaccinal rabies has been reported in dogs, and cats, as well as in people. Some vaccine strains of rabies virus such as the low-egg passage Flury strain can regain their virulence after a few as 5 serial passes in mouse or human neuroblastoma cell cultures.

In dogs, postvaccinal rabies has been reported after use of modified-live chicken embryo low-egg passage Flury vaccine, the CVS-Kissling vaccine, and the high-egg passage Flury vaccine. Affected animals developed ascending paralysis starting in the inoculated hind limb 12 to 14 days after vaccination. This paralysis spreads to involve the other hind limb, then the forelimbs. The condition is not invariably lethal, because in 1 series, 2 of 6 affected dogs recovered. One of the dogs that died also had distemper, and it is reasonable to assume that it may have been immunosuppressed. Monoclonal antibodies have confirmed that the virus isolated from the brain was indeed the vaccine strain. The incidence of vaccine-induced rabies in dogs was 0.4 cases/million doses of low-egg passage Flury vaccine and 0.01 cases/million doses of other modified-live virus rabies vaccines. In California, the rates were 3 and 0/million doses, respectively.

In cats with postvaccinal rabies, IM administration provoked lameness in the inoculated limb, followed by ascending paralysis starting in the inoculated hind limb 2 to 5 weeks after vaccination. This progressed through self-mutilation, seizures, and rigid posterior paralysis to total quadriplegia and convulsions. Vaccine strain rabies virus (most commonly, the SAD/ERA strain) was isolated from the brain tissue of these cats, In many instances, cats had been vaccinated previously against rabies so that prior vaccination did not appear to be protective. In cats that had vaccine induce rabies, 60% were positive for FeLV antigen, suggesting that they may have been immunosuppressed. As a result of exposure to a cat with vaccine-induced rabies, 43 people in Michigan in 1980 were obliged to undergo rabies prophylaxis. Likewise, in a cat that was administered live canine rabies vaccine accidentally 20 persons had to undergo postexposure rabies prophylaxis.

Strains of modified-live rabies virus vaccines used in dogs may retain virulence for nondomestic species. Thus, the low-egg passage Flury strain, designed for use in dogs, has been reported to cause rabies in skunks and raccoons.

Herpesviruses – Herpesviruses have the ability to cause latent infections in animals that become carriers (without clinical signs of disease) as a result. Ill-defined immunosuppressive factors such as stress acting on the carrier animals may provoke virus shedding as well as disease recrudescence. Administration of corticosteroids can also result in the shedding of herpesvirus from such carrier animals.

Bovine herpesvirus 1 – Latent infections have been shown to develop in cattle vaccinated intranasally with a temperature-sensitive mutant of bovine herpesvirus 1 (BHV-1), the cause of infectious bovine rhinopneumonitis virus. When vaccinated cattle were treated twice with dexamethasone, temperature sensitive-BHV-1 could be isolated from 7 of 8 vaccinated cattle for 1 to 8 days. The reactivated virus was temperature sensitive and had a characteristic restriction endonuclease cleavage pattern. Other intranasal BHV-1 vaccine strains may also be shed and so be transmitted to contact control calves. Infected calves have signs of upper respiratory tract infection and may she the virus heavily for up to 8 postvaccination days. It is also of interest that even parenterally vaccinated calves may have transient virus shedding.

Evidence also suggests that ML-BHV-1 given to cattle within 48 hours of arriving at a feedlot may result in higher mortality than that in unvaccinated cattle. For example, when BHV-1 vaccine was administered to cattle 10 to 24 hours after arrival in a feedlot, vaccinated and nonvaccinated cattle began to have signs of respirator tract disease 10-14 days later, and overall morbidity was 2.5 to 50%. Morbidity and mortality were greater in vaccinated than in unvaccinated cattle; 5% of vaccinated cattle died, and 3.2% of nonvaccinated cattle died. Infectious bovine rhinotracheitis was diagnosed at necropsy as the cause of death.

Feline rhinotracheitis virus – Cats vaccinated with modified-live feline calicivirus or rhinotracheitis virus (FRV) could not be induced to shed vaccine virus after treatment with dexamethasone and prednisolone. An important side effect of intranasally administered vaccines is mild clinical disease including sneezing, ocular and nasal discharge, and sometimes, mouth and nasal ulceration. Although it might be considered that these side effects are an inescapable result of successful intranasal vaccination, they have a considerable negative effect on owner''s attitudes toward the vaccination process. In addition, evidence suggests that a vaccine containing modified-live FRV may induce disease after vaccination of cat colonies where the disease is enzootic. If modified-live FRV, designed to be given via the systemic route, is inadvertently given intranasally (for example, if an aerosol is generated with the syringe or the cat licks the injection site), the cats may develop signs of severe respiratory tract disease.

Equine herpesvirus 1 – the earliest live equine herpesvirus (EHV-1), the cause of rhinopneumonitis) vaccines consisted of a virulent, hamster-adapted strain that could cause abortion and respiratory tract disease in susceptible horses. As vaccine production technology improved, less virulent vaccine strains were developed. These were generally considered to be safe, although some were found to be capable of causing abortion in pregnant mares after direct intrauterine inoculation.

A subsequent modified-live EHV-1 vaccine was associated with posterior paralysis in a large number of horses – 486 of 60,000 vaccine recipients. In 3 horses in which clinical features of the condition were described, the disease developed 8 to 11 days after IM administration of ML-EHV-1 vaccine. After euthanasia, lesions consisted of perivascular mononuclear cell infiltration, focal malacia, and vasculitis with axonal swelling closely resembling those seen in horses with naturally acquired rhinopneumonitis. Because of the presumption that these lesions might have been attributable to vaccination, the offending ML-EHV-1 vaccine was withdrawn from the market. Subsequently, it was reported that a virus isolated from a case of vaccine-induced posterior paresis had been characterized as a herpesvirus, but was not neutralized by known equine herpesvirus antisera in virus-neutralization tests.

Adenoviruses – Canine adenovirus 1 vaccine strains, although satisfactory from a protective point of view, can multiply in the renal tubular cells of vaccinated dogs and cause subclinical kidney infections. As a result, virus may be shed in the urine of vaccinated dogs and may spread to other animals. In addition, attenuated canine adenovirus 1 may be isolated from the brain of vaccinated dogs. In one situation, this vaccine strain, together with a DCV strain was isolated from the brain of a dog with encephalitis, about 10 day after vaccination with modified-live virus vaccines. It was hypothesized that the 2 viruses acted synergistically to cause the disease.

Vaccine Contamination

Although morbidity as a result of residual virulence tends to draw the attention of the practicing veterinarian and the animal owner, of greater importance is the possibility that modified-live vaccines may be contaminated and so spread other, unrelated infective agents. Because vaccines are widely exported and may be used in many countries, the potential to cause worldwide epizootics exists.

Canine parvovirus – Prior to 1978, few parvoviruses had been isolated from dogs, and none were considered pathogenic. Attempts to transmit the major cat parvovirus, feline panleukopenia virus (FPLV) to dogs had been uniformly unsuccessful. Nevertheless, in the spring of 1978, severe disease attributable to a parvovirus was recognized almost simultaneously in dogs in the United States, Canada, Australia, New Zealand, South Africa, and Europe. The origin of CPV is unknown, however, it is distinguishable from FPLV and the related mink enteritis virus on the basis of its restriction enzyme patterns and by its host range. Molecular analysis indicates that “only a few differences within the capsid protein gene can endow FPLV with the host range, antigenic type, and hemagglutinating properties of CPV.” These genetic differences are a result of several different mutations in the parvovirus genome. Mutations do, of course, occur spontaneously, and new viruses develop through natural processes. Nevertheless, several mutations infrequently occur spontaneously in a virus over a short period, unless the virus is under severe selective pressure. One hypothesis regarding the origin of CPV suggests that this selective pressure and subsequent mutation more likely occurred in a tissue culture environment than through dog-to-dog passage. Thus, it is speculated the CPV might have arisen through rapid accidental passage of FPV or mink enteritis virus in a canine cell culture line. Cells infected with the new mutant may then have been transmitted globally through a contaminated biological product such as a vaccine.

The development of CPV as a suspected host range variant of FPLV or mink enteritis virus caused widespread mortality in dogs. It is interesting to speculate on the consequences of this host range extension had it involved human beings rather than dogs.

Reticuloendotheliosis – Episodes of reticuloendotheliosis in chickens in Japan and the spread of this disease to the Australian poultry industry have been traced to contaminated batches of Marek disease virus vaccines (Turkey herpesvirus).

Egg drop syndrome 76 virus – an avian adenovirus (EDS 76) was first identified in 1976. It has been suggested that it spread widely as a result of the use of contaminated vaccines. Although speculative, this was considered the most plausible explanation for the sudden, widespread dissemination of this virus in poultry flocks.

Pseudorabies virus – After the use of a modified-live pseudorabies virus vaccine in France, pigs from vaccinated sows developed a disease syndrome that resembled classic hog cholera. Pigs were born with eyelid edema, locomotive disorders, arthritis, and diarrhea. Examination of suspect batches of the vaccine indicated that it was contaminated with a pestivirus similar to the Border disease agent. Pigs were chronically infected with the virus and were immunotolerant. The contamination virus was believed to have been derived from lamb kidney cells used for propagation of the pseudorabies virus.

Bovine viral diarrhea virus – To propagate cells in culture economically, it is usually necessary to use fetal bovine serum as a source of trace nutrients and growth factors. This serum is obtained from fetal or colostrums-deprived newborn calves. If the calves used as donors harbor any infective agents, their serum may be contaminated. When 51 lots of fetal bovine serum from 14 suppliers were tested for bovine viruses, > 30% were found to be contaminated with the following: BVDV in 12%; parainfluenza virus 3-type agent in 10%; BHV-1 in 6%; and bovine enterovirus type-4 in 4%. Even in serum lots that had been pretested by suppliers, 25% (5/20) contained endogenous bovine viruses. This contaminated serum, if used in vaccine production, represents a potential cause of disease. In one example, 8% of BHV-1 vaccine serials tested were found to be contaminated with BVDV. Use of contaminated vaccines has resulted in major episodes of BVD in vaccinated calves. Calves that were inoculated with a BVDV-contaminated rotacoronavirus vaccine had severe growth retardation, thymic atrophy, abomasal ulcers and severe diarrhea. Affected dams aborted or had delayed conception.

Mycoplasmas - Mycoplasmal contamination is also a potential problem in tissue culture vaccines that use fetal bovine serum without preservative. In one study of contaminants in fetal bovine serum from commercial sources, bovine mycoplasmas were found in 17 of 139 lots.

Bovine leukemia virus – As early as 1961, Olson pointed out an apparent relationship between the use of a Babesia vaccine and an increase in the number of cattle with leucosis in Sweden. The Babesia vaccine consisted of fresh blood obtained from Babesia-infected animals. At that time, however, it was not possible to confirm this link. More recently, a major episode of vaccine-induced leucosis was observed in cattle in Australia. Many seropositive cattle were found in a herd from which bovine leucosis had been eliminated. Investigation of this episode indicated that the cattle had been recently vaccinated with a “tick fever” vaccine. The vaccine contained blood that was later shown to have been obtained from a BoLV-positive calf; 13,959 doses of contaminated vaccine had been distributed, but 22,627 cattle from 111 herds given the contaminated vaccine had to be tested to validate claims for compensation. For dairy and beef herds, 62 and 51.8%, respectively, of vaccinated cattle were found to be infected, compared with 6.1 and 1.5%, respectively, of nonvaccinated cattle in the same herds. Investigations indicated that the blood-containing vaccine had more than sufficient viable lymphocytes to transmit the virus to recipient cattle.

Other viruses – Canine adenovirus has been isolated from a FPLV vaccine.

Before condemning all modified-live vaccines as totally unsafe, it is essential to point out that the production of any vaccine must proceed with care and rigorous quality control. Thus, several major disease episodes have been associated with the use of inactivated vaccines where, because of inadequate quality control, vaccine organisms have not been effectively inactivated prior to release.

It is also important to point out that association of adverse side effects with prior use of a vaccine does not prove cause and effect. Not all reactions that occur soon after vaccination are attributable to the vaccine, and in many instances, it may be impossible to prove that the vaccine was the cause. It is indeed tempting to ascribe any pathologic condition that occurs within 10 days of vaccination as a response to that vaccine. This is by no means true, and care should be taken to establish a causal relationship through experimental or epidemiologic analysis.

Nevertheless, it is clear that modified-live vaccines are intrinsically more hazardous than inactivated products. The hazards of residual virulence and contamination are obvious. More important, however, is the ability of new infective agents to develop through the use and dissemination of these vaccines. These agents are major threats to animal populations, wild and domesticated. The best way to ensure that such hazards do not develop in the future is to seek alternatives to modified-live products. We must begin to replace modified-live vaccines with inactivated products. Modified-live vaccines have served us well, but their time is past. We can no longer afford them.