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

Estimates of the prevalence of PI animals in the general cattle population range between 0.13% and 2.0% (Bolin et al., 1985; Houe et al., 1995; Howard et al., 1986; Wittum et al., 2001). Differences in reported prevalence may be due to the population tested, the country/continent where the population was located, management system in effect and/or the diagnostic tests utilized. Persistent infection has a clustered distribution, which means a few herds may contain several cattle but most herds contain only non-PI cattle (Bolin, 1990). Clustering of multiple PI animals in a herd is primarily due to exposure of numerous susceptible dams to a PI or transiently infected source of noncytopathic BVDV prior to day 125 of gestation, but can also be due to surviving PI offspring of a PI dam.

DIAGNOSTIC TESTS FOR THE DETECTION OF PI ANIMALS

Because the PI animal is an important reservoir and transmitter of BVDV, control programs must first identify and remove these animals from the breeding herd. Because of vertical transmission of the virus from viremic dams to their fetuses, PI animals should be removed prior to the start of the breeding season in beef herds with a controlled breeding season. In dairy herds, the PI animals must be removed as soon as possible from direct or indirect contact with the breeding herd. In order to find and remove PI beef cattle prior to the start of the breeding season, all calves, all replacement heifers, all bulls, and all nonpregnant dams without calves must be tested for PI status (Kelling et al., 2000). Nonpregnant cows may lack calves prior to the start of the breeding season due to not becoming pregnant, aborting, or calf mortality. Any female that is still pregnant at the time the herd is tested should be isolated from the breeding herd and kept isolated until her calf is tested and found to be negative. Calves persistently infected with BVDV can be identified by testing strategies that utilize virus isolation from whole blood (buffy coats) or serum, immunohistochemistry (IHC) staining of viral antigen in skin biopsies, antigen-capture enzyme-linked immunosorbent assay (ELISA) from skin biopsies, and polymerase chain reaction (PCR) methods from whole blood or serum (Dubovi, 1996)

VIRUS ISOLATION

Persistently infected animals produce an exceptionally large number of BVDV particles that can be isolated from virtually any tissue sample, including blood, serum, spleen, liver, and lymphoid tissue

(Ellis et al., 1995; Straver et al., 1983). Virus isolation is considered to be very specific for BVDV infection; however, colostral antibodies may temporarily reduce the amount of free virus in the serum of young calves, thus making it less sensitive to isolate virus from the blood or serum of young calves (Kelling, et al., 1990; Palfi et al., 1993; Brock et al., 1998). However, when the maternal antibodies have disappeared, BVDV can be isolated easily from PI calves. BVDV could be isolated by 6 weeks of age in colostrum-fed PI calves in one study (Brock et al., 1998), and by 8 weeks of age in colostrum-fed PI calves in another study (Palfi et al., 1993). Although it is rare, some PI calves develop neutralizing antibody and clear the persistent BVDV strain from their serum. However, in these animals it is still possible to isolate virus from leukocytes (Brock et al., 1998). Virus isolation methods are labor-intensive and take several days to complete. An additional shortcoming is that virus isolation may not differentiate between transiently infected animals and PI animals, unless positive cattle are re tested and remain positive at a later date (i.e., 3 weeks later).

IMMUNOHISTOCHEMISTRY

An IHC test for BVDV infection using skin biopsy samples, such as ear notches, is available that differentiates between PI animals and transient BVDV infections (Njaa et al., 2000). Transiently infected animals may have internal organ tissue samples that are IHC-positive. However, when skin samples were evaluated, transiently infected animals either had no staining, or staining was confined to the epidermal keratinocytes and follicular ostia, in contrast to PI cattle with antigen-positive staining cells in all layers of the epidermis, all levels of hair follicles, and the hair bulb (Njaa et al., 2000). This test is suitable for herd screening because samples can be taken from cattle of any age, sample collection is simple, the samples are stable for transport and handling, it is not affected by the presence of passive antibodies and the test is both sensitive and specific for PI cattle (Ellis et al., 1995; Baszler et al., 1995; Njaa et al., 2000). In addition, the use of modified live BVDV vaccine does not appear to produce false positive IHC reactions on skin biopsies when testing for PI animals (DuBoise et al., 2000).

POLYMERASE CHAIN REACTION

Polymerase chain reaction testing for BVDV infection is more rapid than virus isolation and can also detect virus in antigen-antibody complexes (Brock et al., 1998). Polymerase chain reaction tests are

sensitive and have been shown to differentiate between BVDV genotypes. However, a single BVDVpositive blood sample tested by PCR does not allow the diagnostician to differentiate between viremia from a postnatal acquired infection and that from being a PI animal. Because PCR tests can identify minute amounts of virus, this test can be used in pooled samples of blood or milk in surveillance programs.

SEROLOGY

Although PI cattle are usually seronegative to BVDV, an immune response can be elicited to a heterologous strain (Brock et a., 1998). Presumably, this response can follow either natural or vaccine exposure. In addition, some cattle in both vaccinated and unvaccinated herds are seronegative, making serology alone unsuitable for identification of PI animals (Bolin et al., 1985; Houe et al., 1995; Wittum et al., 2001).

DIAGNOSTIC TESTING STRATEGIES TO IDENTIFY PI CALVES

If a herd has had confirmed PI calves, or if the history strongly suggests the presence of PI calves, the a priori assessment of PI prevalence is fairly high, making the predictive value of a positive test high enough so one can conclude that the animal is persistently infected and that PI animals are present in the herd. Consequently, a second confirmatory test may not be needed. In contrast, if there is no a priori evidence of PI prevalence greater than 0.3–0.5%, the predictive value of a positive test is low and a different confirmatory test may be advisable before making conclusions about the individual animal or the herd (Table 14.2). When a calf is identified as PI, it

should be euthanized or removed for slaughter and the dam should be tested. Most dams of PI calves are not PI themselves, and if confirmed as non-PI, can reenter the breeding herd because naturally acquired immunity is considered to prevent future fetal infections (Orban et al., 1983). Dams identified as PI, however, should be sold to slaughter immediately.

In most whole-herd testing situations, IHC testing of skin samples is currently the test of choice because it can be accurately performed on animals of any age and a single sample is all that is usually sufficient. The use of virus isolation or PCR to identify BVDV-infected cattle requires a second test 3 weeks following any positive results to differentiate between transient viremia and PI with BVDV.

IDENTIFICATION OF DAMS CARRYING A PI FETUS

Purchase of non-PI dams pregnant with PI fetuses (PI-carriers) is a potential source of BVDV spread between herds. Any pregnant heifer or cow that is purchased should not come in contact with other pregnant cattle unless they have been tested and confirmed not to be PI. In addition, because the status of the fetus is unknown, the calf must be tested soon after birth and prior to the breeding season to prevent contact with pregnant cattle in the first 210 days of gestation. Because it would be advantageous to identify PI fetuses prior to purchase or entry of the dam into a herd, other strategies to determine the PI status of the fetus are being investigated.

Unvaccinated non-PI dams carrying PI fetuses are seropositive and virus-negative, but they have markedly higher titers of antibodies to BVDV compared to dams carrying healthy fetuses (Lindberg et al., 2001). Optical density (OD) of an indirect ELISA technique has a strong positive correlation to

virus neutralization titer, and for unvaccinated seropositive dams carrying normal calves, the average OD value stays the same irrespective of when in pregnancy the sample is taken (Lindberg et al., 2001). In contrast, unvaccinated dams that carry PI fetuses have an increasing antibody titer from early pregnancy (when they become infected) through the last month of gestation, resulting in significantly higher OD by the 7th month of gestation compared to dams carrying normal calves (Lindberg, 2001). The sensitivity of this method to identify dams carrying PI calves ranged from 0.79–0.96 and the specificity ranged from 0.37–0.70, depending on the test cutoff (Lindberg et al., 2001). Because of the relatively high sensitivity, relatively low specificity, and low prevalence of dams carrying PI fetuses, the negative predictive value will be high and the positive predictive value will be poor. Therefore, a negative test should be considered good evidence for the absence of a PI fetus, but a positive test would not be a particularly strong indication for the presence of a PI fetus in tested unvaccinated dams in herds with endemic BVDV.

Another method that has been investigated in an effort to identify dams carrying a PI fetus is the collection of fetal fluids and their evaluation for the presence of BVDV (Callan et al., 2002). Callan et al. (2002) were able to use abdominal ultrasound to guide a spinal needle introduced through a surgically prepared area of the right flank into the pregnant uterus to collect a sample of fetal fluid. Virus isolation performed on the fetal fluid samples correctly identified 1/1 BVDV-infected and 168/168 noninfected fetuses. In addition to the technical expertise required, this method may have limited application in most herds because 8% (14/169) of tested dams aborted or delivered premature calves within 3 weeks of fetal fluid collection that may have been related to the sample collection (Callan et al., 2002).

MONITORING HERDS FOR BVDV PI RISK

The cost of initiating a BVDV PI whole-herd screening protocol on a farm or ranch is significant. Because of the low prevalence of herds with at least one PI animal, veterinary practitioners may not be economically justified to initiate whole-herd screening protocols to find PI BVDV beef cattle for herds at low risk for the presence of PI cattle or herds that cannot gain significant market price advantage for selling groups of cattle that have been tested and determined to be free of PI individuals (Wittum et al.,

2001; Larson et al., 2002). However, if ranch history raises a suspicion of BVDV PI cattle being present in the herd, or if significant marketing advantages exist, a protocol to screen the herd can be defended based on its likelihood to improve or protect economic return (Larson et al., 2002). Several strategies can be employed to monitor herds for their risk of having PI cattle present. The interpretation of results from these strategies would be different if the goal were to monitor for the presence of BVDV rather than PI animals. If complete eradication of BVDV is desired, the effort and cost of monitoring is much greater than for monitoring for the presence of PI cattle.

USE OF PRODUCTION RECORDS AND

LABORATORY EVALUATION OF MORIBUND

AND DEAD CALVES

The minimal level of surveillance for every herd should include monitoring of herd fertility (early breeding season pregnancy proportion, pregnancy per insemination proportion, and total pregnancy proportion), neonatal calf morbidity and mortality proportions, and weaning proportions. Because of the negative effect of the presence of PI calves in a breeding herd on measures of reproductive efficiency, the presence of physical abnormalities at birth, and calf survivability to weaning, an unacceptable level of these symptoms increases the risk that BVDV is a problem in the herd and increases the likelihood that whole-herd screening for PI cattle will be economically rewarding (Houe and Meyling, 1991). Although, in many situations, pregnancy rate drops significantly at the time of conception of the oldest PI animal, and about 6 months later calf mortality increases; using production records alone lacks sensitivity for identifying herds with PI animals because the clinical indications of PI presence may be less noticeable in some outbreaks (Houe and Meyling, 1991). The clinical signs and time sequence following introduction of BVDV infection into different herds varies considerably due to the different proportions of seronegative animals in the critical period of pregnancy and different virulence among BVDV strains (Houe, 1995).

In addition to monitoring production records, minimal surveillance should include the necropsy examination of as many aborted fetuses, stillborn calves, and calves that die preweaning as possible, with whole blood submitted for determination of BVD viremia and serum submitted for serologic evidence of infection. In addition, moribund calves from clusters of pneumonia, neonatal scours, or sep-

ticemia outbreaks that are not easily explained by sanitation or other problems should also be tested for BVDV exposure and PI status. If most perinatal and preweaning mortalities are examined for BVDV antigen via IHC and found to be negative, it is not likely that PI animals are present in the herd. The presence of PI animals in the herd will be established by a single confirmed IHC-positive skin sample. The presence of PI animals is not ruled out and may be considered likely if few moribund or dead cattle are tested and found to be IHC-negative, but other tests indicate the presence of viremia or serology indicates recent BVDV infection and the possibility of PI animals being in contact with the moribund or dead sample animals (Table 14.3).

The advantage of utilizing production measures and necropsies to determine whether herds have either a high or low risk for the presence of BVDV PI animals is that minimal expense is involved and these management tactics are inclusive for the monitoring of other disease and production problems. This level of monitoring is probably appropriate in herds with no evidence for the presence of PI animals and that are at low risk of PI introduction (Larson et al., 2002). The disadvantage is that at least one PI animal is allowed into the herd before production losses are identified, and production losses will continue for at least 1 year after intervention is initiated.

USE OF POOLED SAMPLES OF WHOLE

BLOOD FOR PCR TESTING

Herd monitoring for the introduction of PI animals can also be accomplished with pooled whole blood samples for PCR testing. By pooling samples, the expense of screening herds with a low prevalence of PI animals is minimized. Polymerase chain reaction is well suited to pooled-sample testing for the presence of BVDV PI animals because it is sensitive enough to detect minute amounts of virus. A single PI animal was detectable in pools of 200–250 negative samples (Muñoz-Zanzi et al., 2000). Animals contributing to negative samples are all assumed to be non-PI, whereas positive pools may contain samples from PI animals or transiently viremic animals. If the initial pool is PCR-positive, it must be split and retested to differentiate viremic and nonviremic animals. After the viremic animals are identified, they must be classified as transiently infected or PI with either a subsequent PCR or virus isolation test in 3 weeks or via IHC of a skin sample.

The best size of the initial pool is determined by the balance between the cost savings of having large

Table 14.3. Diagnostic test results of clinically ill animals with BVDV infection.

aImmunohistochemistry. bPolymerase chain reaction. cVirus isolation.

numbers of individuals represented in negative pools and few individuals represented in positive pools that require further diagnostics. If pool size is too large, there is an increased chance that any single pool will test positive, requiring additional testing to identify the few truly viremic individuals in the pool. If the samples are grouped in unnecessarily small pools, the cost benefit of pooling samples is lost to the large number of negative pools tested for each positive pool identified (Muñoz-Zanzi et al., 2000). MuñozZanzi et. al. (2000) developed a simulation model to determine that the economically optimum sample size depends on prevalence of true positives in the population. For a PI prevalence of 0.5–1.0%, the optimum number of samples in an initial pool is 20–30, using a described repooling strategy for test-positive initial pools (Muñoz-Zanzi et al., 2000). As prevalence increases, the least-cost initial pool size decreases (Muñoz-Zanzi et al., 2000).

If whole blood samples are collected for pooled PCR from all suckling calves prior to the start of the breeding season, PI cattle can be identified and removed before contact with pregnant females, thereby eliminating the opportunity for a PI animal within the herd to cause reproductive failure and to create more PI animals in the next calf crop. Screening for PI animals at a later time, such as weaning, is discouraged. If samples are taken at weaning, although PI cattle can be removed from the herd, those continuously viremic animals were in contact with pregnant females throughout much of gestation and can cause reproduction and production losses, including the creation of PI cattle in the next calf crop.

USE OF SEROLOGIC EVALUATION OF

SENTINEL ANIMALS

Herd surveillance of dairy herds has been described using sentinel animals. Pillars and Grooms (2002)

found that serologic evaluation of unvaccinated 6- to 12-month–old heifers for the presence of a high serum neutralizing titer had a sensitivity of 66% and specificity of 100% for detecting herds that have PI cattle present when reference strains from both genotypes 1 and 2 were used. Herds that were identified as containing PI animals could then utilize other diagnostic tests to identify individual PI animals for removal. In countries that have BVDV control programs that do not allow the use of vaccination, a similar strategy to identify herds with PI cattle has been investigated. This strategy is based on the fact that in unvaccinated herds, there are significantly more antibody-positive animals, especially in young stock, in herds with PI animals than in herds without PI animals (Houe, 1992).

However, because of the variable percentage of antibody-positive animals in herds without PI animals, it was not possible to predict the presence of PI animals in dairy herds in The Netherlands using this method (Zimmer et al., 2002). Because this monitoring occurs when potential PI animals are at least 6 months old, a positive herd test would indicate that the herd has been exposed to at least one PI animal for at least 6 months, reproductive losses and the presence of PI calves in the next calf crop would be expected. When utilizing serology strategies for BVDV monitoring or diagnostics, it is important to use reference strains from both genotypes 1 and 2 because it is not unusual for BVDV-exposed cattle to have a low titer against one genotype and a higher titer against the other.

USE OF ANNUAL WHOLE-HERD TESTING

Certain high biosecurity herds, such as herds selling or developing replacement breeding animals, may elect to undergo a high level of surveillance even in the absence of evidence that PI animals are present. This high level of biosecurity may be important to their marketing plan or may indicate a high value placed on avoiding the small, but real, risk of introducing BVDV virus into the herd with subsequent negative reproductive, health, and marketing consequences. The first year that a beef herd adopts this strategy, all suckling calves, all females that were bred that failed to present a calf on test day, all replacement heifers, and all bulls should be tested. If any calf is confirmed as a PI animal, his dam should be tested as well. In subsequent years, only suckling calves and any purchased animals need to be tested. If pregnant animals are purchased, the dam should be tested prior to or at arrival and the calf should be tested immediately after birth. In beef herds with a

confined breeding season, this testing must occur before the start of the breeding season to ensure that no PI animals are in contact with pregnant females during gestation. Heifer development operations should test every heifer prior to or at arrival at the facility. Dairy operations should test every calf that stays on the same farm as the breeding herd or is cared for by breeding herd employees, or if any equipment used on the calf-rearing farm is used on the breeding herd farm.

Following the identification and removal of PI animals from a herd, testing of all suckling calves should be done for one or more breeding seasons to ensure the complete accounting for PI animals. Because no test or testing strategy is perfectly sensitive, and because risk factors involved in the initial introduction may still be present, a vigilant monitoring system is wise until a high confidence for PI-free status is achieved.

OTHER POTENTIAL SOURCES OF BVDV

Male PI calves will occasionally be selected for use as breeding bulls. The amount of BVDV excreted in the semen of persistently infected bulls is very high (104–106 TCID50/ml) (Revell et al., 1988). When BVDV is infused into the uterus at the time of breeding, seronegative cattle exhibit a significant reduction in conception rate, but seropositive animals may not be adversely affected (Whitmore et al., 1981). BVDV-contaminated semen is an efficient horizontal transmitter of disease from bull to cow (Paton, et al., 1989). If PI bulls are used for natural service, the cows may conceive when immunity has developed, resulting in the birth of normal (non-PI) calves (Barber et al., 1985; McClurkin et al., 1979). If PI bulls are used for AI, all or most seronegative females bred with the semen will become infected with BVDV although most will not produce a PI calf (Meyling and Jensen, 1988).

Although some evidence exists for BVDV to cause latent infections, particularly in gonads and accessory sex glands (Kirkland, et al., 1991; Voges et al., 1998), recrudescence and excretion in immunecompetent animals has not been shown to date to be involved in the epidemiology of the disease (Brownlie, 1990). And although PI bulls will shed BVDV in semen for prolonged periods of time, virus excretion in semen from transiently infected bulls was confined to days 10–14 post–experimental- infection, and the virus titer in semen of transiently infected bulls was much lower than for PI bulls (Paton et al., 1989).