Serological Evaluation of Precolostral Serum Samples to Detect Bovine Viral Diarrhea Virus Infections in Large Commercial Dairy Herds

Bovine viral diarrhea virus (BVDV; family Flaviviridae, genus Pestivirus) continues to be an important pathogen that affects ruminants worldwide. Despite the widespread use of modified live and killed vaccines, BVDV persists as a pathogen that causes a wide variety of subclinical and clinical infections manifested by respiratory disease, immunosuppression, and decreased reproductive performance.¹

Serological screening strategies to detect herds with endemic BVDV infections are limited, especially in herds that routinely vaccinate against BVDV. One approach for detecting endemic BVDV infection involves demonstrating seroconversion in a subset of animals by testing paired sera (acute and convalescent). The interpretation of this approach is confounded by the widespread use of BVDV vaccines. Also, seroconversion in a subset of animals is difficult to demonstrate due to subclinical infections and to the slow intraherd spread of BVDV. Slow intraherd transmission rates in 2 California dry-lot dairies with similar management practices ranged from 0.5%/day to >1.3%/day, and the proportion of calves infected with BVDV by age 9 months was 67% and 36%, respectively.¹⁵ The presence of subclinical infections and slow intraherd transmission creates a diagnostic challenge when attempting to select the acute and convalescent sampling dates.

Serological evaluation of nonvaccinated sentinel calves has been attempted in both dairy and beef herds and has been marginally successful.^13,17 The presence of nonvaccinated seropositive calves serves as indirect evidence of virus exposure and the likelihood of a BVDV persistently infected (PI) calf within the herd. In Michigan dairy herds containing fewer than 200 lactating dairy cattle, screening nonvaccinated heifers (6–12 months old) had a herd sensitivity of 66% and a herd specificity of 100% for detecting herds with endemic BVDV infections and BVDV PI cattle. The limited sensitivity of sentinel calf screening in dairy and beef herds may be a result of low stocking density and the limited exposure to BVD PI cattle. In addition, large cattle farms have numerous, variable-sized cattle groups that could reduce exposure of BVD PI cattle to sentinels. Commingling or fence-line contact of cattle to positive adjacent herds can result in false-positive results and further confound the use of sentinel calves. Nevertheless, sentinel calves continue to be useful to screen for other pathogens, including Infectious bovine rhinotracheitis virus (IBRV) and Leptospira. Simple, easily interpretable strategies to detect endemic BVDV infections in vaccinated herds with a wide variety of management practices are necessary to provide herd-level information essential for a BVDV control program.

Antigen and nucleic acid detection tests for BVDV, such as immunohistochemistry on skin, antigen-capture enzyme-linked immunosorbent assay on skin, and reverse transcription polymerase chain reaction (RT-PCR) on various tissues and serum, are routinely used for the detection of BVDV PI cattle among vaccinated and nonvaccinated cattle.⁵ All 3 tests have high sensitivity and specificity for detecting BVDV PI cattle.^3,4,6,12 However, a large number of animals must be tested to obtain a high level of confidence that the herd is free of BVDV PI cattle because such animals often represent less than 1% of animals within the herd. Pooling strategies to detect BVDV PI cattle using bulk tank milk¹⁴ or pooled ear notch supernatants⁸ are popular alternatives because of reduced testing fees. When sampled correctly, the bulk tank milk-screening test is reliable at detecting BVDV infections, yet screening bulk milk for BVDV does not capture infections in nonlactating cows and heifers. Pooling saline from soaked skin samples (ear notches) is another method to screen many animals and to reduce testing fees; however, the sensitivity of RT-PCR on pooled ear notch fluids has not been extensively validated, and pooling can result in decreased sensitivity.

The authors propose that newborn calves can serve as a useful sentinel animal for BVDV infection in a herd of pregnant cattle. After 125 days' gestation, the developing bovine fetus is immunocompetent and has completed organogenesis. Fetal infection with BVDV after 125 days of gestation (after the development of a competent immune system) usually leads to the birth of normal calves with precolostral BVDV antibodies.^2,7 Therefore, detecting fetal infection gives a clear indication that BVDV is circulating within the herd, crossing the placenta, and causing fetal infections.

The purpose of the present study was to estimate the percentage of calves with BVDV precolostral serum antibody in large dairy herds with endemic BVDV infections, year-round breeding schedules, and routine vaccination for BVDV. If the percentage of calves with precolostral BVDV serum antibodies is significantly higher than that of BVDV PI cattle, then a precolostral BVDV antibody screening method would have a higher sensitivity than a BVDV PI screening program in detecting endemic infections.

The current study was conducted on 4 large commercial dairy farms of at least 1,000 lactating cows each; 2 herds were in California (herds A and B), and 2 were in Minnesota (herds C and D). The data collected from the California herds were previously published in an article examining the health impact of natural congenital BVDV infection.¹¹ Serum samples were collected from 961 newborn calves before colostrum ingestion. After blood collection (approximately 6 ml per calf), the calf was fed colostrum and managed according to farm protocols.

A single-tube TaqMan RT-PCR was performed on all precolostral serum samples from the Minnesota herds to detect in utero BVDV infection.¹⁰ Briefly, RNA was extracted and purified from 200 μl of serum using a commercially available kit.^a The viral RNA was subjected to reverse transcription and amplification using a TaqMan dual-labeled fluorescent probe. Endpoint analysis of the amplified products was performed on an ABI Prism 7000 Sequence Detection System.^b Calves that tested positive for BVDV by RT-PCR prior to colostrum feeding were retested approximately 2 weeks after the initial result to confirm BVDV PI status. Retesting included RT-PCR, immunohistochemistry on formalin-fixed ear notches, and virus isolation. There was complete agreement between RT-PCR, immunohistochemistry, and virus isolation on all retested samples.

A serum neutralization test was used to detect antibody for BVDV-1 and -2 in the California herds, and a commercially available enzyme-linked immunosorbent assay kit^c was used to determine BVDV serum antibody in the Minnesota herds. The serum neutralization test was performed by the use of serial 2-fold dilutions of heat-inactivated serum, a 100- to 500-tissue culture infective dose (TCID₅₀) of BVDV-NADL (National Animal Disease Laboratory; type 1) or BVDV-125c (type 2), bovine fetal testicle cells, and 96-hr incubation in 5% CO₂ at 37°C. The antibody titer was reported as the highest serum dilution that caused complete inhibition of BVDV-induced cytopathic effects. Titers of ≥1:4 were considered evidence of specific antibody against BVDV, and endpoint dilutions were reported as ≥1:4,096. In the Minnesota herds, antibodies against BVDV-1 and -2 were evaluated with a commercially available indirect enzyme-linked immunoassay. Briefly, 25 μl of serum was added to 100 μl of sample diluent, mixed gently by tapping the plate, and incubated for 90 min. The wells were washed 5 times, the liquid contents were aspirated, and 100 μl of horseradish peroxidase conjugate was added. After a 30-min incubation and wash cycle, 100 μl of tetramethylbenzidine substrate was added; then 100 μl of stop solution was added to the well, and the absorbance was measured at 450 nm. Sample-to-positive ratios were calculated against negative controls.

Serum samples that tested positive for BVDV antibody were further analyzed for total immunoglobulin to detect unintentional colostrum feeding or suckling prior to blood collection. If the serum sample had <400 mg/dl of antibody as detected by a qualitative assay (zinc sulfate turbidity), the calf was considered to be congenitally infected with BVDV with in utero antibody seroconversion.

In the 4 herds examined, the percentage of BVDV RT-PCR-positive calves at birth was 2.6% (25/961), compared with 0.8% (8/961) of calves later determined to be BVDV PI. The data from all herds reemphasize that calves testing positive by RT-PCR should be isolated and retested to rule out transient or acute BVDV infections. In 1 Minnesota herd (herd D), the unknown introduction of a PI heifer to a group of late-gestation cows likely resulted in more calves being viremic at birth. Bulk milk screening in the Minnesota herds C and D was not performed. Minnesota herd D documented BVDV PI lactating cattle prior to the initiation of precolostral screening, and this herd was enrolled to document fetal infections in a herd with known endemic BVDV infections. Minnesota herd D managers did not attempt to detect the lactating BVDV PI cattle because they concluded that additional testing was cost prohibitive. All cattle in Minnesota herd C were tested for BVDV PI by RT-PCR, and 2 heifers were detected in the young stock population. There were no lactating BVD PI cattle detected in Minnesota herd C at the time of testing; therefore, bulk milk testing in Minnesota herd C would likely have been negative at the time of test and removal. Bulk milk screening was not attempted in either of the California herds. No PI calves were born during the sampling period in California herd B.

The detection of BVDV antibody in precolostral serum samples correlated with the birth of BVDV PI calves in 3 of 4 herds examined (Table 1). Bovine viral diarrhea virus serum antibody was detected in 6.8% (range 5.3–8.1%) of live newborn calves. The ratio of BVDV antibody positive to BVDV PI in newborn calves across all 4 herds indicated that for every PI calf born, there are approximately 8 calves congenitally infected and seropositive for BVDV antibody prior to colostrum feeding.

There is a lack of simple and cost-effective strategies to detect endemic BVDV infections in large commercial dairy herds that routinely vaccinate for BVDV. The current “all-antigen” and nucleic acid-detection strategies used to detect BVDV in vaccinated herds have limitations, and eliminating vaccination to determine seroconversion is not appealing to producers and veterinarians. The clinical signs associated with BVDV infection are often subclinical and nondescript. Because of this, livestock producers and veterinarians are often reluctant to invest the time and resources needed to rule in or rule out an endemic BVDV infection based solely on nonspecific clinical impression. Hundreds of cattle in a herd of thousands would need to be tested for BVD PI status to achieve statistical significance that the herd is free of BVDV infections.

In dairy herds with year-round breeding schedules, endemic BVDV infections, and slow intraherd transmission rates, the risk of fetal BVDV infection would be expected to be equal across the entire gestational period. The odds of a fetus being infected during the last half of gestation, a period of approximately 160 days, would be 3 to 4 times greater than the period when fetal infection results in a BVDV PI animal (approximately 50 days). Therefore, screening precolostral serum samples for BVDV antibodies would theoretically yield 3 to 4 times more BVDV antibody-positive calves than BVDV PI calves. The higher percentage of BVDV-seropositive calves than that of BVDV PI calves reduces the number of animals tested to achieve a high confidence of detecting endemic fetal infection. In the 4 herds examined, there were approximately 8 BVDV-seropositive calves for every calf determined to be BVDV PI.

The relatively high number of calves classified as acutely infected at birth in Minnesota herd D was unusual and unexpected. The number of acutely infected calves at birth was twice that of persistently infected calves. Yet this finding may be explained by noting how pregnant cows and heifers are comingled on dairy farms during late gestation. In most dairies, pregnant dry cows (50–60 days prior to calving) and pregnant heifers (50–60 days prior to calving) are first commingled during late gestation. The comingling of these 2 groups of cattle shortly before calving is unique to dairy cattle and not beef cow-calf herds. When these 2 groups are comingled, a PI dairy heifer or cow has the potential to infect naïve, late-gestating cows or heifers in either group. If these 2 groups of cattle were commingled at the beginning of gestation, fewer viremic calves at birth would be expected, but commingling during late gestation would likely result in more viremic calves.

Precolostral screening of newborn calves appears to have many advantages. When newborn calves from first-calf heifers and second lactation and older cows are screened for precolostral BVDV antibodies, both bred heifers and dry cows are screened simultaneously. This approach is advantageous over the young stock sentinel program because it detects virus in pregnant dry cows and first-calf heifers. Nonpregnant cattle can continue to be vaccinated with BVDV vaccines because BVDV-modified live virus vaccine has been shown not to shed and infect nonvaccinated animals.⁹

Although precolostral screening does not appear to be confounded by vaccination, it must be noted that many vaccine companies have received U.S. Department of Agriculture (USDA) exemptions to use modified live Infectious bovine rhinotracheitis virus and BVDV vaccine in pregnant cows. With the USDA exemption, modified live virus vaccines can be used in pregnant cows and heifers provided they were vaccinated, according to label directions, with an approved vaccine prior to breeding. With regard to BVDV, the USDA exemption study parameters included presuckling serum sampling from at least 400 randomly chosen calves from cattle that were administered modified live vaccine during the second and third trimester of pregnancy. All presuckling serum samples were tested and found negative for antibodies to BVDV-1 and -2, thus demonstrating lack of fetal exposure to BVDV modified live vaccine in utero.¹⁶ Therefore, there is no evidence to suggest that fetal seroconversion will occur in cattle vaccinated according to label directions. Improperly administering a BVDV modified live vaccine to pregnant cattle during the last 2 trimesters of pregnancy could result in BVDV fetal infection and fetal seroconversion that could compromise a precolostral surveillance program. Therefore, following label directions when vaccinating against BVDV is critical to maintaining a successful precolostral surveillance program.

Fetal exposure to BVDV and subsequent fetal seroconversion are expected when late-gestation cattle are exposed to BVDV. If BVDV infections are occurring in nonpregnant cattle, congenital infections will not occur and screening for precolostral BVDV antibody in newborn calves will fail to detect infections in nonpregnant cattle and cattle in early gestation. Although precolostral screening needs to be validated across many herds and compared with other screening tests, such as bulk tank milk PCR and calf sentinel programs, preliminary data suggest that screening newborns for BVDV antibody can reliably detect endemic infections in large dairy herds and be used to monitor the progress of BVDV control programs over time.