Материал: Bovine Viral Diarrhea Virus Diagnosis, Management, and Control

described BVD as a separate syndrome from MD and stated that BVD occurs in three forms: severe acute BVD, mild acute BVD, and chronic BVD (now recognized as a form of MD). Severe acute BVD had been prominent a decade earlier, but by 1963, a very mild clinical form of BVD was most commonly observed. Pritchard (1963) noted that one of the common early clinical signs of the severe acute form of BVD was a harsh, dry cough in many of the animals. In some herds, lameness attributed to laminitis was also prominent. In the chronic form, affected cattle failed to grow at a normal rate or lost weight. Many of these animals became emaciated and developed continuous or intermittent diarrhea.

Although BVD could be reproduced experimentally, MD could not. Only fever or a milder form of BVD could be produced with the virus obtained from cases of MD (Pritchard, 1963). An early study of field cases of BVD and MD indicated that there was an immune component to MD (Thomson and Savan, 1963). In seven cases, the cattle that became diseased and died were serologically negative to the virus. One animal, which was clinically ill for 2 months, remained serologically negative immediately prior to death when viable virus was still isolated from its blood. Based on these observations, Thomson and Savan (1963) suggested that cattle that did not recover from their illness may have been incapable of producing an immune response.

By the end of the 1960s, it had become evident that cattle with MD had persistent viremia and often failed to produce neutralizing antibodies to the virus (Malmquist, 1968). It was also observed that fetal bovine serum frequently contained BVDV (Malmquist 1968). This finding and the finding of BVDV infections in newborn and 1-day-old calves were indicative of intrauterine infections (Bürki and Germann, 1964; Romváry, 1965; Malmquist, 1968). Thus, one mechanism proposed to explain persistent infection, and the failure of cattle with MD to produce antibodies to BVDV was the development of immune tolerance in these cattle during intrauterine infection (Malmquist, 1968). Exposure to BVDV before the age of immune competence could lead to failure of the fetus to recognize the virus as foreign and, thus, a failure to produce antibodies to the virus. This would result in the eventual birth of calves infected with BVDV but without the capability to clear the infection. An alternative mechanism proposed by Malmquist (1968) for the lack of neutralizing antibodies in cattle with MD was the destruction of immunologically competent cells by the

virus. It was noted at the time that, if the operating mechanism was immune tolerance, the lack of immune response would be expected to be BVDVspecific, whereas if immune cell destruction was operating, there would also be a depression of the immune response to unrelated antigens (Segre, 1968). The hypothesis of immune tolerance was eventually proven in the 1980s, a decade in which the precipitating event in MD was also identified and in which MD was experimentally reproduced.

Fetal abortions were associated with BVDV infection from the time of the first outbreaks of BVD (Olafson et al., 1946). Evidence of a causative role in abortion included the isolation of noncytopathic BVDV from aborted fetuses and the occurrence of abortions following experimental exposure and after natural outbreaks (Olafson et al., 1946; Baker et al., 1954; Gillespie et al., 1967). Cerebellar hypoplasia and ocular defects (cataracts, retinal degeneration, and optic neuritis) in newborn calves were also observed after BVDV infection of pregnant cows. These BVDV-induced congenital defects were first reported by Ward et al. (1969). For the dams of three calves with cerebellar hypoplasia, Kahrs et al. (1970) reported that infection occurred at an estimated 134–183 days of gestation. Experimentally, cerebellar and ocular defects were produced in calves by inoculation of their dams with a noncytopathic BVDV at 79–150 days of gestation (Scott et al., 1973).

Early in the 1960s, it became established that BVDV was antigenically related to hog cholera virus, now more commonly known as classical swine fever virus (CSFV) (Darbyshire, 1962). Soon after, live BVDV vaccines were proposed for immunization of swine against CSFV in the U.S., but this was later abandoned in 1969 when the Department of Agriculture issued a notice against their use for this purpose (Fernelius et al., 1973). Swine had originally been thought to be dead-end hosts for BVDV, making the virus attractive for immunization of swine. However, this was proven to be false when BVDV was isolated from naturally infected swine (Fernelius et al., 1973). Later, serological evidence indicated that the agent causing border disease in sheep was also related to BVDV and CSFV (Plant et al., 1973).

DIAGNOSIS AND CONTROL

The serum neutralization test developed after the discovery of cytopathic strains of BVDV became an important diagnostic assay and is still widely used today. Early studies in the 1960s using this test proved that BVDV was present worldwide at a high

seroprevalence (usually about 60%) in most adult cattle populations and that the majority of BVDV infections were subclinical in nature.

During the 1960s, the use of cell lines such as Madin-Darby bovine kidney (MDBK) cells simplified the study of the virus and its diagnosis (Marcus and Moll, 1968). Also of importance was the development of the fluorescent antibody technique (FAT) for the detection of BVDV in inoculated cell cultures (Fernelius, 1964). Prior to the development of FAT, identification and titration of noncytopathic BVDV was difficult, often involving calf inoculation, an interference test in which the cytopathic effect of cytopathic BVDV was inhibited by noncytopathic virus (Gillespie et al., 1962; Diderholm and Dinter, 1966), or agar gel-diffusion precipitin tests. With FAT, the detection of noncytopathic BVDV from biological specimens was made much easier, although the interference test was still used experimentally as late as 1985 (McClurkin et al., 1985).

Soon after the discovery of the cytopathic BVDV strain, Oregon C24V, Coggins et al. (1961) reported on its attenuation by serial passages in primary bovine kidney tissue culture. This led to its mass production as a modified-live vaccine for BVD in 1964 (Peter et al., 1967). However, soon after its use, reports began to appear of a few sick animals in some herds following immunization. It became apparent that the vaccine produced an MD-like disease in a small number of animals in a few herds (Fuller, 1965; Peter et al., 1967; McKercher et al., 1968). The animals that became sick with MD-like symptoms usually died. The discussion of possible causes of postvaccinal MD (pvMD), as it became known, included insufficient attenuation of the Oregon C24V strain and contamination of the vaccine with a virulent BVDV field strain. Further, since pvMD was first observed after the introduction of multivalent vaccines against both BVD and infectious bovine rhinotracheitis (IBR), it was considered possible that a synergistic effect between the two viruses may have been responsible for the postvaccinal syndrome. The possible role of the vaccine as a stressor in cattle already in the acute stage of infection with a BVDV field strain was also suggested (McKercher et al., 1968). Peter et al. (1967), however, noted that animals that died from pvMD failed to develop antibodies to BVDV but could produce antibodies to IBR virus. Thus, they suggested that cattle succumbing to MD or pvMD were uniquely susceptible to BVDV and that this susceptibility involved a failure of the immune system.

BVDV was recognized as a highly contagious disease that was easily spread from herd to herd (Pritchard, 1963). The original outbreaks in New York were described as explosive in character, in which practically all the animals in a herd came down with the disease within a few days (Olafson et al., 1946). Thus, before a vaccine became available, control procedures were limited to protecting the noninfected herd by prevention of direct and indirect contact with infected animals.

By the midand late 1960s, modified-live vaccines comprising the attenuated Oregon C24V and NADL strains were in use (Bittle, 1968; Gutekunst, 1968). The latter strain was isolated in 1962 at the National Animal Disease Laboratory and attenuated in porcine kidney cell culture. Although modifiedlive vaccines were considered efficacious in preventing acute BVD, their use was thought to be somewhat risky due to the occurrence of pvMD. The risk of pvMD was somewhat a matter of perception and personal experience. Bittle (1968), describing the Oregon C24V vaccine, stated that the incidence of postvaccinal problems reported to the USDA compared to the number of doses used had been extremely small (less than 1 in 10,000) and he encouraged its use. Likewise, Gutekunst (1968) stated that, following 8 months of field use of 350,000 doses of the NADL vaccine, no reports of postvaccinal reactions had been received. In contrast, a commentator at the meeting at which Drs. Bittle and Gutekunst spoke remarked that he had personally observed 40 cases of pvMD in a 3-month period in approximately 1,000 vaccinated cattle (Clark, 1968). Subsequently, it was discovered that pvMD occurs only in persistently infected cattle and although these animals comprise a small portion of the total cattle population, they could comprise a significant portion of animals in an individual herd.

The modified-live vaccines were contraindicated for use in pregnant animals because of the possibility of inducing abortions, and thus, emphasis was placed on vaccination of calves and heifers. Kahrs (1971) also advocated protecting naive cattle from contact with potentially infected cattle, particularly those from auction markets or shipping centers and, probably because of the risk of pvMD, considered this the “best control method.” For feedlot cattle, Fuller (1965) stated that it was poor economy to vaccinate already stressed cattle immediately upon arrival to feedlots. Instead, he recommended acclimatizing calves to their new surroundings for about 3 weeks before vaccinating them against BVDV.

THE DECADE OF THE 1970s

PERSISTENT INFECTIONS IN SICK AND

APPARENTLY HEALTHY CATTLE

By the 1970s, it was established that calves with congenital BVDV infections were unthrifty and usually died within a few months, and that surviving calves often suffered from chronic disease, were persistently infected with the virus, and were deficient in serum neutralizing antibodies against BVDV (Bürki and Germann, 1964; Malmquist 1968; Johnson and Muscoplat, 1973). These observations were made in sick and unthrifty animals. In 1978, Coria and McClurkin (1978) reported on the persistent infection and immune tolerance of an apparently healthy bull. The bull was continuously viremic and a noncytopathic BVDV was isolated from its blood leukocytes repeatedly from birth to 2.5 years of age. The virus was also isolated repeatedly from the bull’s semen. The bull remained seronegative for BVDV antibodies during this time. When challenged with the equivalent of three doses of a killed vaccine, it failed to produce a significant immune response. It was suggested that the bull had acquired the infection early in gestation before the development of its immune system and had thus acquired a specific immune tolerance to BVDV.

A year later, McClurkin et al. (1979) described the reproductive performance of four healthy cattle persistently infected with BVDV: the bull previously identified and three newly identified pregnant cows. As an aid to the identification of apparently healthy, persistently infected cattle in a herd, they noted that cattle that remain seronegative while in contact with seropositive cattle were likely to be immune-tolerant and persistently infected. They further noted that, after three vaccinations of the herd, all cattle except these three cows had serum neutralizing antibody titers of 1:16 or greater against BVDV. As in the persistently infected bull, a noncytopathic BVDV was consistently isolated from the three persistently infected cows, and these cows gave birth to calves that became ill: Two died within the first week postpartum and one calf was euthanized at a few weeks of age. A noncytopathic BVDV was isolated from the blood leukocytes of all three calves indicating maternal transmission.

When seropositive cows were bred by the persistently infected bull, normal calves were born but services per conception averaged a high 2.3. For five seronegative heifers bred by the bull, services per conception averaged 2.0. All seronegative heifers seroconverted to BVDV and had high antibody titers

( 1:128) 6 weeks after breeding. Four heifers gave birth to normal calves and one heifer aborted at 6 months of gestation but no BVDV was isolated from the fetus. None of the calves produced by either group showed evidence of intrauterine infection. However, the study indicated that BVDV may play a role in repeat breeding problems. It was suggested that losses due to repeat breeding and from neonatal disease might be prevented by ensuring that all cattle have high antibody titers to BVDV before breeding (McClurkin et al., 1979).

Also of interest in this study was the observation that BVDV antibody titers in the herd (38% of the animals had titers >1:256) were generally much higher than expected (usually 1:16 to 1:64) for cattle vaccinated with the killed vaccine. It was surmised that virus shedding by the persistently infected animals had constantly challenged the other cattle in the herd and boosted their antibody titers. The idea that the level of BVDV antibodies in a herd could be used to predict which herds contained persistently infected animals was later developed in the 1990s and recently evaluated for herds of unknown BVDV status (Pillars and Grooms, 2002), as will be discussed later in this chapter.

Lesions in the persistently infected, healthy cattle were found to be microscopic, primarily in the brain and kidney (Cutlip et al., 1980). However, immunofluorescence staining of tissues demonstrated widespread distribution of viral antigen in brain and spinal cord neurons, renal glomeruli, renal tubules, lymph nodes, spleen, small intestine crypts, testicular tubules, and endothelial cells.

DIAGNOSIS AND CONTROL

The observation that fetal bovine sera frequently contained BVDV and neutralizing antibodies against BVDV (Kniazeff et al., 1967; Malmquist, 1968) was an increasing concern in the 1970s for cell culture work because contaminated cultures could have undesirable consequences for research and vaccine production. Tamoglia (1968) found that 8% of licensed live IBR vaccines were contaminated with BVDV, raising concerns that such vaccines might give rise to fetal abortions. Experimental studies, like those examining the effects of cytopathic BVDV, could be compromised if cell cultures used to propagate the cytopathic virus were contaminated with noncytopathic BVDV. Commercial fetal bovine sera were contaminated with noncytopathic BVDV as a result of pooling sera from infected and noninfected fetuses. Commonly, sera of 500 fetuses

were pooled and, although the incidence of fetal infection was unknown at the time, Nuttall et al. (1977) calculated that contamination of a batch of fetal bovine serum required only a 0.2% incidence of infection. At the time, the interference test (Gillespie et al., 1962), which relied on the visible inhibition of cytopathic BVDV, was used to screen fetal bovine sera and bovine cells for noncytopathic BVDV. Nuttal et al. (1977) believed that the interference test was not sensitive enough to detect lowlevel contamination of fetal bovine serum with noncytopathic BVDV and advocated the use of FAT for regular screening of sera and cells for noncytopathic virus.

For serological diagnosis of BVDV infection, Lambert et al. (1974) recommended the use of the serum neutralization assay on paired serum samples collected 2 or 3 weeks apart, with a rising titer indicating active infection. To prevent BVD in neonatal calves, they recommended vaccination of open heifers and cows 30–60 days prior to breeding and the consumption of antibody-rich colostrum by calves.

Because of the concern for pvMD, abortions, and in utero infections, a number of studies were done in the early and mid-1970s on the efficacy of killed vaccines for BVDV. However, these efficacy studies were often limited in scale and involved challenge with homologous virus. Lambert et al. (1971) evaluated a killed BVDV-NADL vaccine and found it efficacious against homologous virus challenge in calves aged 5–11 months. McClurkin et al. (1975) examined the use of killed vaccines using inactivated cytopathic NADL and Singer strains to prevent fetal infection and found these vaccines efficacious. However, these researchers used homologous challenge, and a heterologous or noncytopathic BVDV challenge was not attempted. Challenge with noncytopathic virus would prove to be essential in evaluating fetal protection when it was later discovered that only noncytopathic viruses caused persistent infections.

THE DECADE OF THE 1980s

EXPERIMENTAL PRODUCTION OF

PERSISTENT INFECTION AND MUCOSAL

DISEASE

In 1984 and 1985, a number of important advances were made in BVDV research. McClurkin et al. (1984) described the production of persistently infected, immune-tolerant calves in five experiments involving 44 cows in the first trimester of pregnancy

(42–125 days). Cows or their fetuses were inoculated directly with one of five different BVDV isolates. Four of the isolates were noncytopathic strains and the fifth was the cytopathic NADL strain. For 38 pregnant cows or fetuses inoculated with noncytopathic strains, there were 10 abortions, 1 stillborn calf, 4 weak or unsteady calves, and 23 calves that had a normal and vigorous appearance at birth. All weak calves and 22 of the 23 normal appearing calves were persistently infected and seronegative and were the result of inoculations occurring at days 42–125 of gestation. One calf infected at 125 days of gestation was immune-competent, virus-negative, and seropositive at birth.

Of interest, but not fully appreciated at the time, was the failure by McClurkin et al. (1984) to produce any persistently infected calves with the cytopathic NADL strain. Prior to this study, Done et al. (1980) had infected 15 pregnant cows at 100 days of gestation with a mixture of 10 cytopathic strains of BVDV. Virus was recovered from eight live-born calves, but in each case these viruses were noncytopathic BVDV. These researchers also did not appear to fully recognize the significance of this finding, stating that “the absence of cytopathogenicity in all reisolates of virus is remarkable, but perhaps no more than a reminder of the genetic variablity of viruses in general and of the pestiviruses in particular.” It is almost certain that noncytopathic BVDV contaminated their cytopathic virus stocks used for infection. Later, Brownlie et al. (1989) infected pregnant cattle with a cytopathic virus and could not produce persistent infection. Thus, it became generally accepted that only noncytopathic BVDV could produce persistent infections.

McClurkin et al. (1984) also followed the fate of the persistently infected calves they had produced. All four weak calves either died or were euthanized within 4 months after birth. Of the 22 apparently healthy calves, 6 developed diarrhea and/or pneumonia within 5 months of birth and died, and 1 became unthrifty and remained small. Ten of the apparently healthy calves remained healthy at 6 months of age or as yearlings (most of these calves were then used in other experiments). Three animals remained healthy as 2-year-olds and were bred. Two of the persistently infected cows produced apparently healthy, persistently infected calves, whereas the third cow lost her calf and developed MD at 28 months of age. This study demonstrated several characteristics of persistent infection: that persistently infected calves may be born weak or apparently normal; that some may live to breeding age;

and that persistently infected families can arise by breeding persistently infected cattle.

Soon after the study by McClurkin et al. (1984), Brownlie et al. (1984) and Bolin et al. (1985c) experimentally reproduced MD in persistently infected cattle. Brownlie et al. (1984) noted that while healthy, persistently infected cattle were infected with only noncytopathic BVDV, both noncytopathic and cytopathic BVDV could be isolated from persistently infected cattle that were clinically ill with MD. The latter finding was also observed by McClurkin et al. (1985). From their observations, Brownlie et al. (1984) developed a hypothesis for the induction of MD, which stated that cattle become persistently infected with noncytopathic BVDV after in utero infection and postnatally succumbed to MD when superinfected with a cytopathic BVDV. To test the hypothesis, Brownlie et al. (1984) used a cytopathic isolate from an animal suffering from MD to inoculate two healthy persistently infected herdmates. Both animals came down with MD, supporting the hypothesis. In another study, Bolin et al. (1985c) inoculated persistently infected cattle with noncytopathic or cytopathic BVDV. The cattle inoculated with noncytopathic virus did not develop clinical signs of disease, whereas MD developed in all cattle inoculated with cytopathic BVDV.

Although pvMD was known to be a relatively common phenomenon, failure to consistently induce MD in persistently infected cattle with cytopathic BVDV vaccines (Bolin et al., 1985b) suggested that the induction of MD was somewhat more complicated than simple superinfection with a cytopathic BVDV in an animal persistently infected with noncytopathic BVDV. Subsequent studies indicated that the noncytopathic and cytopathic viruses (called a virus pair) from individual MD cases were antigenically similar. Howard et al. (1987) compared five virus pairs from separate MD outbreaks and found that virus pairs from the same outbreak were antigenically indistinguishable when tested in crossneutralization tests with antisera. Corapi et al. (1988) using a panel of monoclonal antibodies also showed that noncytopathic and cytopathic viral pairs from MD had a high degree of antigenic similarity. This led to the conclusion that in natural outbreaks of MD the likely origin of the cytopathic BVDV was via mutation of the noncytopathic BVDV infecting the persistently infected animal. Thus, the hypothesis for the induction of MD was refined to include antigenic similarity: MD was induced in a persistently infected animal by “superinfection” with a cy-

topathic BVDV with antigenic similarity to the noncytopathic BVDV. Superinfection could occur by mutation in spontaneous MD or, in the case of pvMD, by vaccination with a vaccine virus antigenically similar to the noncytopathic BVDV infecting the persistently infected animal.

The above hypothesis suggested that inoculation of persistently infected animals with a cytopathic BVDV antigenically different (heterologous) from the noncytopathic persisting virus would not result in MD; but rather, the animal would produce antibodies to the cytopathic virus and clear the infection. This apparently was the reason for the failure of superinfection with cytopathic BVDV (e.g., by vaccination) to produce MD consistently. In support of this hypothesis, Moennig et al. (1990) showed that persistently infected animals superinfected with closely related cytopathic BVDV developed MD within 14 days of infection but those that were superinfected with heterologous cytopathic BVDV did not develop MD within a 2–3-week time frame. Instead, these animals developed antibodies to the superinfecting virus. Other studies, however, showed that the inoculated cytopathic BVDV could sometimes be heterologous and yet precipitate MD in a persistently infected animal. Westenbrink et al. (1989) inoculated 14 clinically healthy, persistently infected animals with three heterologous cytopathic viruses. Twelve of these animals developed MD, some within the expected time frame for MD of 2–3 weeks postinoculation, and several others after several months (so called late-onset MD). The actual range when MD began was 17–99 days. Interestingly, neutralizing antibodies were produced against the inoculated cytopathic virus but, as expected, not to the persisting noncytopathic virus. However, in 10 of 12 cases, the neutralizing antibodies did not neutralize the cytopathic virus recovered at necropsy from the intestines. This suggested that mutational changes had occurred and, as a result, cytopathic virus antigenically similar to the persisting noncytopathic virus arose and induced MD.

Shimizu et al. (1989) also produced MD in persistently infected cattle with inoculation of heterologous cytopathic BVDV. Similar to the findings of Westenbrink et al. (1989), they found that the cytopathic viruses recovered from blood early after infection were antigenically similar to the challenge cytopathic virus but that the cytopathic viruses isolated from the carcasses at necropsy were antigenically different from the challenge virus but similar to the noncytopathic persisting virus. Neutralizing antibodies were produced to both the challenge