Response to Modified Live and Killed Multivalent Viral Vaccine in Regularly Vaccinated, Fresh Dairy Cows*

E. J. Dubovi, Y. T. Gröhn, M. A. Brunner, and J. A. Hertl Response to Modified Live and Killed Multivalent Viral Vaccine in Regularly Vaccinated, Fresh Dairy Cows* Edward J. Dubovi, PhD a Yrjo T. Gröhn, DVM, PhD b Michael A. Brunner, DVM, PhD a Julia A. Hertl, BS, MS b Department of Population Medicine and Diagnostic Sciences a Diagnostic Laboratory b Section of Epidemiology New York State College of Veterinary Medicine Cornell University Ithaca, NY 14853 ABSTRACT Vaccination programs for viral pathogens in the dairy industry span the full spectrum of possibilities even though few of these have been evaluated in field situations. One such program is the vaccination of fresh cows 30 to 60 days postpartum with modified live viral (MLV) vaccines. The purpose of this study was to evaluate the antibody response to booster vaccinations during this period. The impact of vaccinations on milk production and reproductive performance was also examined. The response of cattle boosted with MLV bovine viral diarrhea virus (BVDV) was greatly enhanced compared with the saline controls and the killed vaccine test group. Similar increases were not seen with the MLV infectious bovine rhinotracheitis virus (IBR) and bovine respiratory syncytial virus (BRSV). Changes in milk production were not detected. There was a *This study was sponsored by Fort Dodge Animal Health. positive effect on the rate of conception with the MLV group even though there was no evidence of the presence of the three viruses in the herd at the time of study. Although this was a single field trial, and thus limited in scope and repeatability, the results indicate that the vaccines used had a positive effect. INTRODUCTION Most cattle vaccine programs include multivalent bovine viral diarrhea virus (BVDV), infectious bovine rhinotracheitis virus (IBR), and bovine respiratory syncytial virus (BRSV) vaccines. The viral components in many of these vaccines are killed, modified live viral (MLV), or a combination of both killed and MLV. Veterinarians and producers use one or more of these vaccines in a variety of programs, often without real knowledge of the disease protection that their use achieves. Killed BVDV and IBR vaccines are often used because they can be administered to the 49

Veterinary Therapeutics Vol. 1, No. 1, Winter 2000 entire herd at one time without the risk of causing abortion. 1 A disadvantage of killed BVDV and IBR vaccines is that their duration of fetal protective immunity is limited; thus revaccination is usually recommended in 4 to 6 months. 1 3 A temporary drop in milk production is also sometimes reported following the use of killed multivalent vaccines. 4 Modified live vaccines have gained widespread use because they provide longer and often a greater degree of protective immunity than killed vaccines. 5,6 In addition, a single dose of vaccine can immunize whereas two doses of killed vaccine are needed. 1 A negative aspect of MLV vaccines is that they may cause abortion, which limits their use to only nonpregnant animals. 1 To take advantage of the greater protection afforded by MLV vaccines, many cattle vaccine programs now include a dose of MLV vaccine at about 35 days in milk (DIM). Quite often this is in addition to a previously administered annual or semiannual vaccination with killed components. To evaluate the efficacy of this protocol, a study was designed to measure the immune response to both killed and MLV multivalent vaccines administered at 35 DIM to cows vaccinated semiannually. Other objectives of the study were to estimate the short-term effects of these vaccine treatments on milk production in early lactation and to study the effects on fertility. MATERIALS AND METHODS Serum Neutralization Assays Serum (virus) neutralization (SN) assays followed standard microplate formats with serial twofold dilutions done in duplicate wells. All sera were titrated to endpoints for the three viruses in the study: BVDV, IBR, and BRSV. The virus content for all three agents was approximately 100 to 300 tissue culture infectious dose 50 per well. The strain of viruses used in the assays were: BVDV-Singer, IBR-Cooper, and BRSV-A4. The cells used in the assays were low pass bovine testicular cells for BVDV and BRSV SN tests, and RK-13 cells for IBR SN tests. Subjects The study group consisted of 163 Holstein cows in a herd housed at the Cornell Dairy Teaching and Research Center in Dryden, NY. The cows were calved between December 1997 and May 1998 and were followed between 35 and 71 DIM after calving. Primiparous cows and multiparous cows were randomized separately into three treatment groups: saline (control group), killed multivalent vaccine (BVDV, IBR, BRSV, bovine parainfluenza type 3 virus [PI 3 ]; Triangle 4, Fort Dodge), and MLV multivalent vaccine (BVDV, IBR, BRSV, PI 3 ; Pyramid MLV4, Fort Dodge). Cows were vaccinated by intramuscular injection according to label instructions on day 35 of lactation. Controls were administered saline in the same manner. A prevaccination serum sample was collected on day 35, and postvaccination serum samples were collected on days 49 and 70 of lactation. In addition, all cows in this study received a scheduled semi-annual herd vaccination with a multivalent vaccine (mixed live and killed; IBR, BVDV, BRSV, PI 3, and Leptospira spp.; Cattlemaster, Pfizer) on November 16, 1997. The earliest study vaccination would have occurred approximately 31 days from the last semi-annual herd vaccination. Effect of Vaccines on Antibody Titers The titer values for BVDV serum neutralization were analyzed with a log-transformation, which normalized the data. Parities (1, and 2 and higher) were analyzed separately. General linear models procedure (PROC GLM) in statistical analysis system (SAS) 7 was used to estimate the effect of vaccine treatments on the titer values at each of the three times (analyzed 50

E. J. Dubovi, Y. T. Gröhn, M. A. Brunner, and J. A. Hertl separately). This statistical procedure fits general linear regression models. Because the data consisted of repeated measurements (titer values recorded at three times) taken within a cow, the SAS procedure linear regression model with fixed and random components (PROC MIXED) 8 was also used to estimate the effect of vaccine treatments on titer values at each of the three times. This statistical procedure is used to analyze data with several sources of variation (i.e., in measurements taken over time, as well as in factors [e.g., vaccine treatment] explicitly controlled for in the analysis). An autoregressive covariance structure, in which a measurement is correlated with the preceding measurement, was used to account for the fact that the titer values are not independent from one another within the same cow. The results from PROC GLM and PROC MIXED were similar; only those from PROC GLM are reported here. The above analyses were repeated on two additional data sets, for BRSV SN titer values and for IBR SN titer values. For all three data sets it was also of interest to determine whether days since the whole herd vaccination on November 16, 1997, had any effect on subsequent titer levels. PROC GLM, 7 with days since vaccination as a covariate, was used to determine this possible effect. It was also of interest to determine if the starting value of the titer (i.e., before vaccination) had any effect on the two subsequent titers. Therefore PROC GLM models were run with titer values at 49 and 70 days as the (separate) outcomes and the titer value at 35 days as a covariate. The models also contained the vaccine treatments and were run separately for each parity group. Of the 163 cows in the serological and milk yield analysis, there were 54 cows in each of the saline and killed vaccine treatment groups and 55 cows in the MLV treatment group. Sixty-five of the cows were primiparous and 98 were multiparous. Effect of Vaccines on Milk Production The data consisted of the same cows described above. Milk measurements were recorded daily, from days 28 to 71 after calving. As stated above, cows were vaccinated on day 35 of lactation. Because the study consisted of repeated measures (daily milk weights) taken within a cow, similar statistical procedures were used as in the serological analysis. Data on primiparous and multiparous cows were analyzed separately. The SAS procedure PROC GLM 7 was used to estimate the effect of vaccine treatments on milk production, on each of the 10 days following vaccination (day 35). In addition, the SAS procedure PROC MIXED, 8 with an autoregressive covariance structure, was used to account for the fact that the milk measurements were not independent from one another within the same cow. As for the serological analysis, the results from both procedures were similar; only those from PROC GLM are reported here. Effect of Vaccines on Reproductive Parameters Data were available on 150 of the 163 cows described previously; information on the reproductive parameters of the remaining 13 cows was unavailable due to premature culling. Days to first service, number of times bred, days open, and whether the cow conceived or remained open were recorded. PROC GLM was used to analyze the effect of vaccines on days to first service and on number of times bred. Survival analysis, using the Cox Proportional Hazards Model in SAS proportional hazards regression model procedure (PROC PHREG), 8 was used to analyze the effect of vaccines on days open. Output from PROC PHREG is in the form of hazard rate ratios. These can be used to compare the rate at which an event (in this study, conception) occurs in different groups 51

Veterinary Therapeutics Vol. 1, No. 1, Winter 2000 of subjects. In this study, for any particular factor, a hazard rate ratio less than 1.0 implies that cows with that factor conceived at a lower rate than cows in the reference group, while a hazard rate ratio greater than 1.0 indicates that cows with the factor conceived at a higher rate than cows in the reference group. Of the 150 cows in the fertility analysis, there were 47 cows in the saline treatment group, 50 cows in the killed vaccine treatment group, and 53 cows in the MLV treatment group. Sixty-five of the cows were primiparous and 85 cows were multiparous. ANALYSIS BVDV Within the saline group BVDV SN titer values did not change significantly over time (Table 1). For primiparous cows receiving either the killed or MLV vaccine, the BVDV SN titer increased significantly between 35 and 49 DIM, as well as between 35 and 70 DIM. Although this response to both vaccines was significant, the response to the MLV vaccine was considerably greater than the response to the killed vaccine. This occurred even though the MLV vaccine group started at a higher TABLE 1. BVDV SN Titer Values a Titer 1 b (35 d) Titer 2 (49 d) Titer 3 (70 d) Log Mean (Mean) c SE d Log Mean (Mean) SE Log Mean (Mean) SE PRIMIPAROUS COWS Saline 4.7 (176.4) 0.2 (40.0) 4.6 (151.6) 0.2 (40.0) 4.6 (152.4) 0.2 (40.0) Killed 4.6 f (153.5) 0.2 (85.0) 5.7 g (517.1) 0.2 (85.0) 5.2 g (289.1) 0.2 (85.0) MLV e 5.2 f (304.0) 0.2 (81.8) 8.4 g (6155.6) 0.2 (81.8) 7.9 g (3072.0) 0.2 (81.8) MULTIPAROUS COWS Saline 5.5 (436.4) 0.2 (68.3) 5.4 (381.7) 0.2 (68.3) 5.3 (349.2) 0.2 (68.3) Killed 5.3 (401.6) 0.2 (69.8) 5.9 (520.0) 0.2 (69.8) 5.6 (400.0) 0.2 (69.8) MLV 5.6 f (550.8) 0.2 (242.5) 7.7 g (2878.1) 0.2 (242.5) 7.4 g (2040.2) 0.2 (242.5) a Log and actual by parity at days 35, 49, and 70 of lactation in 163 dairy cows. Cows were vaccinated on day 35 with either saline, killed, or modified live multivalent (BVDV, IBR, BRSV, PI 3 ) vaccine. Values are logarithms of the original titers; actual mean titers are in parentheses. Both sets of values are presented because the statistical analysis was done on the logarithmic scale to account for the non-normality of the original data. b Titer 1 was measured immediately before vaccination. c Values with different superscripts differ significantly (α = 0.05) across time ( f,g, ). d Standard errors of log means; standard errors of actual means are in parentheses. e Modified live. 52

E. J. Dubovi, Y. T. Gröhn, M. A. Brunner, and J. A. Hertl titer level than the killed vaccine group. The reason for this difference in starting titer is unknown. Among multiparous cows, the BVDV SN titer increased significantly between 35 and 49 DIM, and between 35 and 70 DIM, only for those receiving the MLV vaccine. Time since the whole herd vaccination did not have a meaningful impact on BVDV titer levels, except in two of the subgroups. Among primiparous cows receiving saline, time since vaccination had a negative effect on titer values. Among multiparous cows receiving the killed vaccine, time since vaccination had a positive effect on titer values. For both the killed and MLV vaccines, there were significant differences in mean titer values between parities at days 35, 49, and 70. The titer values for the multiparous cows were consistently higher than those for the primiparous cows. The only exception was at day 49 for cows receiving the killed vaccine; there was no difference in mean titer values between the parity groups. To determine whether pre-vaccination BVDV SN titer values at 35 DIM affected titer values at 49 and 70 DIM, linear regression models using PROC GLM 7 were run with the prevaccination titer value at 35 DIM as an explanatory variable and the titer value at 49 and 70 DIM as the outcomes (in separate runs). Among primiparous cows in the killed vaccine group, for every unit increase in the prevaccination log titer value at 35 DIM, the value at 49 DIM increased by 1.6 units (314.3 units when analyzed on the actual scale; P = 0.0001). Similarly, it increased by 1.1 units (134.5 units; P = 0.0062) by 70 DIM. For primiparous cows in the MLV vaccine group, a unit increase in the prevaccination log titer at 35 DIM led to a 3.9 unit increase (6274.9 units; P = 0.0001) at 49 DIM. At 70 DIM, this increase was 3.4 units (2931.3 units; P = 0.0001). Among multiparous cows in the killed vaccine group, a unit increase in the prevaccination log titer at 35 DIM led to a 1.0 unit increase (137.4 units; P = 0.0068) at 49 DIM. By 70 DIM, it had led to a 0.9 unit increase (56.4 units; P = 0.0372). For multiparous cows in the MLV vaccine group, a unit increase in the prevaccination log titer at 35 DIM led to a 2.7 unit increase (2499.5 units; P = 0.0001) at 49 DIM. At 70 DIM, this increase was 2.4 units (1673.4 units; P = 0.0001). As can be seen at both times (49 and 70 DIM), the multiparous cows, who had a higher mean prevaccination titer, had a lower immune response than did the primiparous cows. BRSV For BRSV SN titer values there were no significant increases across time associated with treatment in either parity group (Table 2). Time since the whole herd vaccination had a negative effect on SN titer values in primiparous cows in the saline and MLV groups. Similar to BVDV, it was determined whether prevaccination BRSV SN titer values at 35 DIM affected titer values at 49 and 70 DIM. Among primiparous cows, for every unit increase in the prevaccination log titer value at 35 DIM, those in the killed vaccine group showed a 1.1 unit increase (26.6 units; P = 0.048) at 49 DIM. For every unit increase in the prevaccination log titer value at 35 DIM, primiparous cows in the MLV vaccine group showed a 1.3 unit increase (103.5 units; P = 0.0007) at 49 DIM. There was no significant increase at 70 DIM in either vaccine group. Among multiparous cows, for every unit increase in the prevaccination log titer value at 35 DIM, those in the MLV vaccine group showed a 1.2 unit increase (130.4 units; P = 0.001) at 49 DIM. No other differences were significant. As with BVDV, primiparous cows showed a consistently higher immune response than did multiparous cows. 53

Veterinary Therapeutics Vol. 1, No. 1, Winter 2000 TABLE 2. BRSV SN Titer Values a Titer 1 b (35 d) Titer 2 (49 d) Titer 3 (70 d) Log Mean (Mean) c SE d Log Mean (Mean) SE Log Mean (Mean) SE PRIMIPAROUS COWS Saline 4.3 (86.5) 0.1 (12.9) 4.3 (85.0) 0.1 (12.9) 4.2 (83.8) 0.1 (12.9) Killed 4.4 (118.9) 0.2 (24.5) 4.6 (135.3) 0.2 (24.5) 4.4 (119.5) 0.2 (24.5) MLV d 4.6 (169.1) 0.2 (50.9) 5.1 (250.2) 0.2 (50.9) 4.8 (199.3) 0.2 (50.9) MULTIPAROUS COWS Saline 5.2 (257.9) 0.1 (36.8) 5.1 (222.1) 0.1 (36.8) 5.2 (240.5) 0.1 (36.8) Killed 5.1 (243.3) 0.2 (41.5) 5.2 (259.8) 0.2 (41.5) 5.2 (280.3) 0.2 (41.5) MLV 5.3 (269.1) 0.1 (58.9) 5.6 (362.7) 0.1 (58.9) 5.5 (338.4) 0.1 (58.9) a Log and actual by parity at days 35, 49, and 70 of lactation in 163 dairy cows. Cows were vaccinated on day 35 with either saline, killed, or modified live multivalent (BVDV, IBR, BRSV, PI 3 ) vaccine. Values are logarithms of the original titers; actual mean titers are in parentheses. Both sets of values are presented because the statistical analysis was done on the logarithmic scale to account for the non-normality of the original data. b Titer 1 was measured immediately before vaccination. c Standard errors of log means; standard errors of actual means are in parentheses. d Modified live. IBR IBR SN titer values increased significantly following vaccination in both the killed and MLV groups (Table 3). Among primiparous cows, titer values increased significantly between 35 and 49 DIM for these two treatment groups. Among multiparous cows, the most marked increase occurred in cows in the killed vaccine treatment group. Time since whole herd vaccination had a negative effect in primiparous cows in the saline group, but not for any others (results not shown). Similar to BVDV and BRSV, we determined whether prevaccination IBR SN titer values at 35 DIM affected titer values at 49 and 70 DIM. Analysis of predictive titer value data on IBR indicated that among primiparous cows receiving the killed vaccine, for every unit increase in the prevaccination log titer at 35 DIM, there was an increase of 1.7 units (72.8 units; P = 0.0001) at 49 DIM. There was an increase of 1.2 units (24 units; P = 0.0035) at 70 DIM. Among primiparous cows receiving the MLV vaccine, for every unit increase in the prevaccination log titer at 35 DIM, there was an increase of 1.4 units (54.6 units; P = 0.0001) at 49 DIM. There was an increase of 1.1 units (30.7 units; P = 0.0043) at 70 DIM. Among multiparous cows receiving the killed vaccine, for every unit increase in the prevaccination log titer at 35 DIM, there was an in- 54

E. J. Dubovi, Y. T. Gröhn, M. A. Brunner, and J. A. Hertl TABLE 3. IBR SN Titer Values a Titer 1 b (35 d) Titer 2 (49 d) Titer 3 (70 d) Log Mean (Mean) c SE d Log Mean (Mean) SE Log Mean (Mean) SE PRIMIPAROUS COWS Saline 3.6 (52.3) 0.2 (9.2) 3.6 (50.3) 0.2 (9.2) 3.7 (53.9) 0.2 (9.2) Killed 3.3 f (40.6) 0.2 (13.9) 4.5 g (109.1) 0.2 (13.9) 3.9 h (67.1) 0.2 (13.9) MLV e 3.8 f (60.4) 0.2 (14.9) 4.5 g (112.7) 0.2 (14.9) 4.3 f,g (90.5) 0.2 (14.9) MULTIPAROUS COWS Saline 4.3 (103.5) 0.1 (15.2) 4.1 (86.8) 0.1 (15.2) 4.2 (95.2) 0.1 (15.2) Killed 4.4 f (104.4) 0.1 (13.1) 5.1 g (176.0) 0.1 (13.1) 4.8 h (132.5) 0.1 (13.1) MLV 4.2 f (102.0) 0.1 (13.1) 4.6 g (126.3) 0.1 (13.1) 4.4 f,g (114.2) 0.1 (13.1) a Log and actual by parity at days 35, 49, and 70 of lactation in 163 dairy cows. Cows were vaccinated on day 35 with either saline, killed, or modified live multivalent (BVDV, IBR, BRSV, PI 3 ) vaccine. Values are logarithms of the original titers; actual mean titers are in parentheses. Both sets of values are presented because the statistical analysis was done on the logarithmic scale, to account for the non-normality of the original data. b Titer 1 was measured immediately before vaccination. c Values with different superscripts differ significantly (α=0.05) across time ( f,g,h ). d Standard errors of log means; standard errors of actual means are in parentheses. e Modified live. crease of 1.5 units (89.3 units; P = 0.0001) at 49 DIM. There was an increase of 1.1 units (33.4 units; P = 0.0003) at 70 DIM. Among multiparous cows receiving the MLV vaccine, for every unit increase in the pre-vaccination log titer at 35 DIM, there was an increase of 1.2 units (41.1 units; P = 0.0001) at 49 DIM. There was no significant increase at 70 DIM for the MLV vaccine group. Milk Yield Within each parity group there were no significant differences in milk yield among the three vaccines at any of the days after vaccination (Table 4). The variation among cows was fairly constant, with standard errors ranging from 0.8 to 1.3. Within each vaccine group, milk yield tended to increase slightly over time, generally following the shape of the lactation curve. No significant differences were detected because of the large variation in milk yield in this herd. Among primiparous cows there were no significant differences in milk yield from day to day in any of the three treatments. Among multiparous cows the daily milk yield differed significantly on days 30 and 41 compared with the milk yield on day 45 of lactation for the saline group (results not shown). There were no differences from day to day for either the killed or MLV vaccine. Reproductive Parameters The vaccines had no effect on days to first service or on number of times bred. However, cows re- 55

Veterinary Therapeutics Vol. 1, No. 1, Winter 2000 TABLE 4. Least Squares Means and Standard Errors (SE; kg) of Daily Milk Yield Saline Killed MLV a Mean SE Mean SE Mean SE PRIMIPAROUS COWS Day 1 b 31.7 34.8 36.2 Day 0 c 33.5 c 1.0 34.0 0.9 33.8 1.0 Day 1 35.0 1.0 34.8 0.9 33.8 1.0 Day 2 35.7 1.1 34.3 1.0 33.5 1.1 Day 3 35.2 1.0 36.3 1.0 33.6 1.0 Day 4 36.1 1.1 34.9 1.0 34.0 1.1 Day 5 36.8 1.2 34.9 1.2 35.0 1.1 Day 6 36.0 1.0 34.6 1.1 34.1 1.1 Day 7 35.6 1.0 34.9 1.0 34.3 1.0 Day 8 36.0 1.0 35.5 0.9 34.4 0.9 Day 9 36.1 1.0 35.7 1.0 35.0 1.0 Day 10 37.5 1.0 37.2 1.0 35.3 1.0 MULTIPAROUS COWS Day 1 47.3 48.6 44.9 Day 0 46.5 1.0 45.2 1.0 48.0 1.0 Day 1 46.6 1.0 47.9 1.0 47.4 1.0 Day 2 46.6 1.0 44.7 1.1 45.0 1.1 Day 3 46.8 0.9 46.8 1.0 46.5 1.0 Day 4 47.1 0.8 46.7 0.9 48.2 0.9 Day 5 47.9 1.1 48.0 1.1 48.7 1.1 Day 6 49.1 1.2 46.7 1.3 49.4 1.2 Day 7 47.4 1.1 47.3 1.0 48.4 1.1 Day 8 47.0 1.4 47.1 1.4 48.4 1.3 Day 9 46.6 1.0 47.7 1.0 48.7 1.0 Day 10 48.5 1.2 47.0 1.2 48.9 1.2 a Modified live. b Indicates mean milk yield on day before vaccination. c Indicates number of days after vaccination (Day 0 = day of vaccination); values for Days 0 10 are adjusted by vaccine and parity. ceiving the MLV vaccine were nearly twice as likely to conceive as cows receiving saline (Table 5). This unexpected finding may be explained by the fact that no information was available on postpartum disease monitoring; disorders such as dystocia, retained placenta, and metritis could have accounted for this difference. There was no significant difference in the rate of conception between cows receiving the killed vaccine and cows receiving saline. Other factors influencing time to conception (days open) were days to first service and number of times bred. The parity group (1, >2) had no effect on time to conception. 56

E. J. Dubovi, Y. T. Gröhn, M. A. Brunner, and J. A. Hertl TABLE 5. Effect of Vaccines on Days Open (Hazard Rate Ratio a ) b Factor Hazard Rate Ratio 95% Confidence Limits Treatment Saline (control group) 1.0 Killed 1.59 0.90, 2.83 MLV c 1.91 d 1.10, 3.31 Days to 1st service 0.96 d 0.94, 0.98 Times bred 1 1.0 2 0.17 d 0.09, 0.30 3 0.07 d 0.03, 0.13 4 0.03 d 0.01, 0.07 >5 0.01 d 0.00, 0.03 Parity 1 1.0 >2 0.85 0.55, 1.32 a Hazard rate ratio: a value less than 1.0 implies that cows with that factor conceived at a lower rate than cows in the reference group, while a value greater than 1.0 indicates that cows with the factor conceived at a higher rate than cows in the reference group. The baseline group always has a hazard rate ratio of 1.0 to which all other groups (i.e., levels) for that variable are to be compared. b Adjusted for days to first service, number of times bred, and parity for 150 cows calving in New York State between December 1997 and May 1998. c Modified live. d P 0.05. DISCUSSION Because SN titer values in the saline control group remained constant, it is unlikely that any of the treatment animals were exposed to a field strain of BVDV, IBR, or BRSV during the course of the experiment. Therefore, the SN titer changes noted in the vaccinated groups are a true response to the treatment vaccines. The marked SN titer response to the BVDV fraction of the modified live vaccine strongly supports the efficacy of vaccinating cows with a modified live BVDV vaccine prior to breeding. This response does not support the hypothesis that cows previously vaccinated on a semiannual basis with a killed BVDV vaccine are likely to neutralize a modified live BVDV vaccine before it can replicate sufficiently to stimulate an immune response. This is further verified by the finding that the magnitude of the SN response was not correlated with the length of time following the semiannual herd vaccination. It is not known whether the adjuvant in the MLV vaccine had an impact on the enhanced response to the BVDV fraction of the vaccine. These data also support the idea that some killed vaccines may not induce a complete humoral immune response. The challenge of the animals with a live virus permitted an enhanced response with antibody titers stabilizing at significantly higher values. A vaccination strategy of priming with killed vaccine followed by an MLV booster may achieve titer levels not found with either vaccine type alone. Further studies to assess vaccine strategies will be forthcoming. The SN response to the BVDV fraction of the killed vaccine administered postpartum was much less than the response observed with the modified live vaccine. A possible explanation for this might be that one dose of the killed vaccine was insufficient to stimulate a strong response because the vaccine previously used in 57

Veterinary Therapeutics Vol. 1, No. 1, Winter 2000 the semiannual herd vaccination contained a different adjuvant and strain of BVDV. A more likely explanation is that there was a threshold value of antibody stimulated by killed vaccines, and repeated vaccination with a similar product has minimal effect. Multiparous cows had a lower response than did primiparous cows; this may be due in part to the fact that older cows started out at a higher titer, even before vaccination, than did the younger cows. The apparent lack of a response to the IBR and BRSV fractions of the modified live vaccine administered postpartum may be due to the fact that the semiannual vaccine used in the herd also contained modified live fractions of IBR and BRSV. These fractions may have already stimulated the immunity to these viruses to a maximum threshold above which the additional postpartum booster had no effect. As with BVDV, older cows tended to show less of a response than did younger cows. In all of these discussions regarding vaccine responses, one must keep in mind that no assessments were made of cell-mediated immunity. Many veterinarians report a temporary decrease in milk production following administration of multivalent killed vaccines. In this experiment this was not noted, nor was there a positive response for the lactation in milk production or reproductive performance. For a positive vaccine response in these parameters, a disease problem theoretically would have to exist in a herd at the time of vaccination. In this study, because no field viruses were evident in the herd during the experiment, there was no opportunity for the additional vaccine to improve on these parameters. No challenge was administered to the live study subjects because this was a field situation and introducing a disease into the population was not an option. CONCLUSION Although the results from this field study on one dairy herd in New York State cannot necessarily be generalized to other settings, they do indicate a beneficial immune response in dairy cattle to the products used. ACKNOWLEDGMENT The authors thank Fort Dodge Animal Health for its generous financial support of this study. REFERENCES 1. Bolin SR: Control of bovine viral diarrhea infection by use of vaccination. Vet Clin North Am Food Animal Pract 11(3):615 625, 1995. 2. Bolin SR, Ridpath JF: Range of viral neutralizing activity and molecular specificity of antibodies induced in cattle by inactivated bovine viral diarrhea virus vaccine. Am J Vet Res 51:703 707, 1990. 3. Van Oirschot JT, Bruschke CJM, van Rijn PA: Vaccination of cattle against bovine viral diarrhea. Vet Microbiol 64:169 183, 1999. 4. Schultz RD: Certain factors to consider when designing a bovine vaccination program. Proc Annual Am Assoc Bov Pract 26:19 26, 1994. 5. Cortese VS, Whittaker R, Ellis J, et al: Specificity and duration of neutralizing antibodies induced in healthy cattle after administration of a modified-live-virus vaccine against bovine diarrhea. Am J Vet Res 59:848 850, 1998. 6. Cortese VS, Grooms DL, Ellis J, et al: Protection of pregnant cattle and their fetuses against infection with bovine viral diarrhea virus type 1 by use of a modifiedlive-virus vaccine. Am J Vet Res 59:1409 1413, 1998. 7. SAS User s Guide: Statistics, Version 5 Edition. Cary, NC, SAS Inst., Inc., 1985. 8. SAS Technical Report P-229 SAS/STAT Software: Changes and Enhancements, Release 6.07. Cary, NC, SAS Inst. Inc., 1992. 58