Leptospirosis, a zoonotic disease caused by a gram-negative bacterial spirochete,1 adversely affects reproductive efficiency in dairy cattle by resulting in lower than typical conception rates,2 failure to conceive, and embryonic and fetal death.3 More than 200 serovars of Leptospira have been identified worldwide, many of which are pathogenic to cattle.1 Most infections in cattle are attributed to serovar Hardjo,1,4–6 of which 2 strains are known to exist: Leptospira interrogans serovar Hardjo strain hardjoprajitno and Leptospira borgpetersenii serovar Hardjo strain hardjobovis.
The 2 strains of serovar Hardjo are serologically and clinically similar yet genetically and antigenically distinct.5,7–10 Strain hardjoprajitno, known as the US reference strain, was believed to be the primary leptospiral pathogen of North American cattle prior to 1990; however, advances in diagnostic techniques led to the discernment of genetic differences within serovar Hardjo and its subsequent classification into 2 species and strains.5 Strain hardjobovis has been isolated from cattle populations worldwide, including those in the United States, whereas strain hardjoprajitno appears to be restricted primarily to the United Kingdom.1,10,11 Prevalence estimates of strains hardjoprajitno and hardjobovis in the United Kingdom are not readily available; however, strain hardjoprajitno is reportedly endemic in UK cattle.2,12 Strain hardjobovis is believed to be the most common Leptospira strain in cattle in Australia, New Zealand, the United Kingdom, and North America,13,14 with 42% of beef herds15 and 57% of dairy herds16 in the United States having positive test results in 2007 and 2003, respectively.
Cattle serve as maintenance hosts for strains hardjoprajitno and hardjobovis; therefore, the rate of transmission and incidence of infection with these particular strains is high.10 Leptospires penetrate mucous membranes1 via fluids such as contaminated drinking water and excretions from infected animals such as urine, milk, and placental fluids.10,17 The organisms circulate in the blood and can invade all major organs. Colonization is most common in the kidneys and reproductive tract, allowing shedding of leptospires into urine, which increases the risk of transmission to others. Spirochete replication in the reproductive tract increases the potential for venereal and transplacental transmission,17 whereas persistent infection of the reproductive tract can lead to early embryonic death, abortion, stillbirth, and birth of weak calves.3,17,18
Persistent infection of the reproductive tract in female cattle is the most economically important manifestation of serovar Hardjo.1,17 Although the mechanisms are not fully understood, researchers presume that after bacteremia and bacteriuria, spirochetes can persist in the oviduct and uterus, interfering with embryo implantation and other early pregnancy events.1 This causes infertility by failure to maintain the conceptus and is associated with an increase in number of services per conception and a prolonged calving interval.1 Estimated costs of $2.55/d to $4.68/d are incurred when a cow fails to conceive during the optimal calving interval.19,20 In a dairy herd, the cumulative results of failure to conceive or maintain the conceptus are a decrease in the dairy's profitability.
Annual vaccination and antimicrobial administration have been used for prevention and treatment of leptospirosis in cattle.17 Because the 2 strains of serovar Hardjo are serologically similar and results of serologic testing have been classically used to determine the organisms used for vaccine development, the US reference strain hardjoprajitno has been used in available pentavalent vaccines for North American cattle. The commercially available pentavalent vaccines, which target strain hardjoprajitno, have not protected against strain hardjobovis because the vaccines do not contain hardjobovis antigens and cross-protection between strains does not develop.5,21
Research into the efficacy of a commercially available monovalent strain hardjobovis vaccine has been conducted; however, findings were inconclusive.3,22,23 In a small study22 of vaccinated beef heifers experimentally challenged with strain hardjobovis, renal and uterine tissue colonization with leptospires was prevented and urinary shedding of leptospires was eliminated. Although it is feasible that the heifers would also have a lower likelihood of reproductive tract colonization with the organism than would cows, reproductive performance was not assessed. A larger study23 of naturally infected dairy cattle that received the same monovalent hardjobovis vaccine resulted in an increase in conception rates at first service; however, the specific strain of Leptospira was not identified and the reproductive efficiency variables evaluated were of limited scope, compared with those evaluated in similar studies. A third study3 of beef cattle that received the monovalent vaccine coupled with long-acting oxytetracycline revealed no significant improvement in reproductive performance in naturally infected herds.
The effects of antimicrobial use for treatment of infection with Leptospira strain hardjobovis were also evaluated in experimentally infected cattle.17 That study showed that several antimicrobials were suitable for eliminating renal colonization and urinary shedding of leptospires, including injection once with oxytetracycline, although cattle were only monitored for 4 to 6 weeks and resistance to reinfection was not evaluated. The mixed results and various trial protocols used to evaluate the efficacy of a monovalent strain hardjobovis vaccine in US dairy cattle clearly indicate the need for additional research. The objective of the study reported here was to determine whether a monovalent vaccine against L borgpetersenii serovar Hardjo strain hardjobovis would improve reproductive efficiency in Holstein cattle in a commercial dairy setting.
Materials and Methods
Animals—Cattle used in the study consisted of a herd (n = 3,600) at a commercial dairy farm in Lodi, Calif, that was considered representative of a typical production dairy environment in the western United States. The herd included 1,900 milking and nonlactating cows. The study protocol and all procedures and animal handling methods were approved by the University of California-Davis Institutional Animal Care and Use Committee. In addition, consent for use of all cattle enrolled in the study was obtained from the dairy herd owner.
Management—Cows were housed in free-stall barns with access to open dry lots, fed a total mixed ration, and milked 2 times/d. For the study, the farm's herd health protocol was not modified. Vaccines in the herd health plan, including those against common abortifacient agents, continued to be administered as was routine. Specifically, at the 30th to 50th day of lactation (30 to 50 DIM), cows received modified live virus vaccinesa SC for protection from infectious bovine rhinotracheitis, bovine viral diarrhea, bovine parainfluenza-3, bovine respiratory syncytial virus, and leptospirosis. The Leptospira vaccine included 5 serovars (L interrogans serovars Canicola, Grippotyphosa, Hardjo strain hardjoprajitno, Icterohemorrhagica, and Pomona). Killed virus vaccinesb against the agents of these same diseases were administered SC at confirmation of pregnancy and the cessation of lactation. In addition, Escherichia coli bacterinc; Clostridium chauvoei, Clostridium septicum, Clostridium haemolyticum, Clostridium novyi, Clostridium sordellii, and Clostridium perfringens types C and D bacterind; and Fusobacterium necrophorum bacterine were administered SC at cessation of lactation (ie, 8 weeks before parturition) and at 3 weeks before parturition and E coli bacterin was readministered at parturition.
Specific vaccines were also routinely administered as part of the heifers' herd health plan. At birth, heifer calves were given intranasal vaccine for protection against infectious bovine rhinotracheitis and parainfluenza type 3 viruses.f At approximately 4 and 5 months of age, heifers were administered SC the same modified-live vaccinea as the cows received as well as C chauvoei, C septicum, C haemolyticum, C novyi, C sordellii, and C perfringens types C and D bacterind and a trivalent vaccine consisting of a Moraxella bovis bacterin.g In addition, at 4 months of age, a strain RB51 vaccine against Brucella abortush was administered SC.
Determination of leptospirosis herd prevalence—The herd prevalence of leptospirosis due to strain hardjobovis within the herd was determined at 2 points: before and after the vaccine trial. The sample size was determined with the assumption that 20% or none of the herd would be infected, and 42 cattle would be needed to determine herd prevalence with a 99.9% CI in a population of 2,000 cattle. Herd positivity was defined such that if 1 animal was identified as infected, the herd would be considered endemically infected. Initially, a subgroup (n = 46) of cows (23) and heifers (23) was selected as a convenience sample from 3 age groups of cattle within the dairy facility: nulliparous heifers, primiparous cows, and pluriparous cows were used. After the trial concluded (at least 82 days after vaccination and antimicrobial treatment or placebo administration), a subgroup (n = 40) of cattle was randomly selected from the vaccinated group (24) and the control group (16) for prevalence testing. This second sample consisted of heifers (n = 12), first and second lactation cows (18), and third or later lactation cows (10).
Two types of samples were obtained for disease detection at both time points. Blood samples were collected from the coccygeal vein. To obtain urine samples, furosemidei was injected into the coccygeal vein (1.0 mg/kg), and urine was collected midstream by free catch. Urine and blood samples were submitted to a commercial diagnostic laboratoryj for Leptospira testing. The presence of leptospires in the urine samples was determined by fluorescence antibody staining as implemented and described elsewhere.10,15,24 When a positive test result was obtained, the presence of antileptospiral antibodies in serum harvested from the blood samples was analyzed by MAT to identify the Leptospira serovars.10,24
Titers of antibodies against L interrogans serovars Bratislava, Canicola, Grippotyphosa, Hardjo, Icterohemorrhagica, and Pomona were measured via MAT by use of serum harvested from the blood samples. When cattle had leptospires identified in their urine and antibody titers for all tested serovars except Hardjo were ≤ 1:100 and for Hardjo were > 100, the cattle were considered to be infected with L borgpetersenii serovar Hardjo strain hardjobovis.10 In other words, infected cattle as defined for the study were those that were leptospiruric, without high antibody titers for serovars of L interrogans but with an antibody response to Hardjo greater than or equal to that of the other L interrogans species and ≥ 100.
Experimental design—Holstein cows and heifers were selected from the herd for participation in the vaccine trial from June 2009 through June 2011. To be considered for inclusion, all cattle were required to have completed both treatments (vaccinated and antibiotic or placebo, at least 3 weeks apart) at least 2 weeks before breeding. Uniparous or multiparous cows were required to have stayed in the herd and have been present at least 1 day after the voluntary waiting period for breeding (48 days after calving or more), and nulliparous heifers were required to have been bred once. Unbred nulliparous heifers were excluded because they might have had anatomic or physiologic anomalies preventing conception. These criteria yielded 1,894 eligible cattle for random assignment via coin toss to either the vaccinated or control group. Cows were enrolled at cessation of lactation and nulliparous heifers at the time of first breeding.
Cattle in the vaccinated group were inoculated SC with 2 mL of a monovalent Leptospira vaccinek containing chemically inactivated whole cultures of L borgpetersenii serovar Hardjo strain hardjobovis. In addition, an SC injection of oxytetracycline (20 mg/kg) was administered as recommended for control of hardjobovis.l Vaccinated cattle were revaccinated between 28 and 35 days after the first vaccination, as per label instructions. The second vaccine was administered at least 2 weeks before breeding, as recommended by the manufacturer. Cattle in the control group were given 2 doses (2 mL each) of lactated Ringer's solutionm (placebo) SC at the same timing as in the vaccinated group.
Both groups were housed together in pens on the basis of their lactation and stage of lactation within free-stall barns and received additional standard herd health care. All dairy personnel, including the breeder and veterinarian, were unaware of study group assignment; no cattle markings or records were used on the farm to identify study vaccination status.
Lactating cattle were eligible for artificial insemination 44 days after parturition (ie, 44 DIM), and heifers were eligible after introduction to the heifer breeding pen. The observation period for all cows (primiparous and pluriparous) began once they received their second Leptospira vaccination or placebo injection and were eligible for insemination. The observation period for heifers (nulliparous females) began at their initial breeding date. This method was chosen to give the nulliparous heifers sufficient time to begin first estrus (puberty) and reach sufficient body frame size.25,26
The observation period continued for a maximum of 1 year after treatment. Cows were visually monitored for signs of estrus by tail chalk removal in accordance with industry standard. Ovulation in cows in which estrus was not detected by 60 days after parturition or cows deemed nonpregnant at the time of pregnancy diagnosis was synchronized with a standard protocol.27 All cows and heifers were artificially inseminated when estrus was observed. Pregnancy diagnosis was performed as usual by manual and ultrasonic reproductive examination per rectum by the herd veterinarian at 32 to 40 days after insemination. Females declared to be nonpregnant at a reproductive examination subsequent to a positive pregnancy test result were considered to have aborted. All reproductive records were maintained on dairy computer softwaren and used as the data source.
Statistical analysis—Dairy reproductive records for enrolled animals were used for statistical analysis with the aid of statistical software.o Survival analysis, a regression technique for which the time to an event is the measured outcome,28,29 was used to measure the effect of the vaccine protocol on interval to conception. When a cow or heifer failed to conceive, that animal was censored at the time of death or culling or at study termination. Cox proportional hazards regression30 was used to analyze the effect of the vaccine protocol on interval to conception while adjusting for potentially explanatory and confounding factors.31 The fixed effects were treatment, lactation (as a surrogate for age and parity), DIM at first breeding, and DIM at conception or risk of conception. The model developed was extended to consider the potential impact of sire and the possibility of interval to conception being influenced through heredity by the animal's sire and dam as well as the sire of conception.32 Two software functions, Cox mixed-effects and Cox proportional hazards, were used. The Kaplan-Meier (product limit) estimator was used to estimate the interval from parturition to conception for the survival analysis.
Nonsignificant variables and interaction terms identified during the initial modeling process were excluded from the final model. Thus, variables and interactions included in the final model were treatment, lactation, DIM at first breeding, and DIM at conception. Dependent variables consisted of days to conception or days to censoring and conception status. The explanatory variables treatment, lactation status, and DIM at conception were forcibly retained in all preliminary models and the final model. All models had the following general form:
The hazard function, hi(t), was the probability of the ith observation having a positive conception test result at t days after 44 DIM (initiation of risk period). The baseline hazard function, h0(t), was the likelihood of conception at time t, when all the independent (explanatory) variables were at their mean values. Unknown coefficients (βj) were estimated for the Xj independent variables. Initial models used to examine the effect of the vaccine protocol on interval to conception included the variables for treatment, lactation number, DIM at first breeding, number of times bred, DIM at conception, season of first breeding, conception season, and season of previous calving, with seasons defined as January to February, March to May, June to October, and November to December.
For evaluation of interval to conception, 4 categories of cattle were developed for time to event or censoring and coded separately as follows: inseminated but never conceived, inseminated and confirmed pregnant via manual or ultrasonic reproductive examinations, culled or died for unknown causes during the observation period and therefore did not receive a first insemination, or met all criteria throughout the observation period yet were never inseminated and therefore never conceived. To analyze the effect of lactation, cattle were grouped by lactation status as follows: heifer, first and second lactation, and third or greater lactation. Days in milk at first breeding was coded as 40 to 49, 50 to 59, 60 to 69, 70 to 79, or 80 to 128 days beginning at 44 DIM, which was recognized as the first day after parturition a female might be exposed to insemination. Because heifers had not yet given birth, a proxy measure for DIM at first breeding was assigned as 0 DIM (approx 11 to 13 months of age). Season of parturition, season of first breeding, and season of conception were coded as previously defined. Cattle that did not have a prior calving date (ie, heifers), first breeding date, or conception date were categorized collectively within their respective groups.
To estimate the conception rate, the cumulative proportion of pregnant cows and heifers was used. Logistic regression was used to estimate the proportion of heifers, first and second lactation cows, and third or later lactation cows within each treatment group that conceived during the observation period. Generalized linear modeling was used to determine the effect of the vaccine protocol and lactation status on the proportion conceiving.
The survival analysis technique used for the conception data was used to determine the interval from conception to pregnancy loss and to evaluate the effect of the vaccine protocol on failure to maintain a viable fetus. Logistic regression was used to estimate the proportion of heifers, first and second lactation cows, and third or later lactation cows within each treatment group that had a pregnancy loss. The proportion of pregnancy losses was used as a proxy measure for rate of pregnancy loss in each treatment and lactation group. A generalized linear model for binomial data was used to determine the effect of vaccine protocol and lactation status on the rate of pregnancy loss.
The Pearson χ2 test with Yates continuity correction was used to compare the initial herd prevalence (pretrial) of Leptospira infection with the posttrial prevalence, seroprevalence by lactation status at the initial and posttrial assessments, and initial seroprevalence with the posttrial seroprevalence in each treatment group. Values of P < 0.05 were considered significant for all analyses.
Results
Animals—Nine hundred eighty-six cattle comprised the vaccinated group, which included 325 (33.0%) heifers, 369 (37.4%) cows in their first or second lactations, and 292 (29.6%) cows in their third or greater lactation; 878 (89.0%) of vaccinates became pregnant. The control group was comprised of 908 cattle, which included 291 (32.0%) heifers, 330 (36.3%) cows in the first or second lactation, and 287 (31.6%) cows in their third or greater lactation; 804 (88.5%) of control cattle became pregnant.
Leptospirosis herd prevalence—Before the trial of the vaccine against L borgpetersenii serovar Hardjo strain hardjobovis began, 6 of the 46 (13%) cattle had positive results of testing for infection with the strain. Infected cattle included 2 of 23 (9%) nulliparous heifers, 3 of 21 (14%) first or second lactation cows, and 1 of 2 third or greater lactation cows. The prevalence of infection did not differ significantly (P ≥ 0.38) among these lactation groups.
After the trial concluded, leptospires were detected in the urine samples of 6 of 40 (15%) cattle (vaccinated group, 3/24 [13%]; control group, 3/16 [18.8%]). Speciation by MAT confirmed the presence of antibody against L borgpetersenii serovar Hardjo strain hardjobovis in all cattle with a positive urine test result. Comparison of the initial herd prevalence with the posttrial prevalence revealed no difference (P ≥ 0.93), nor was a difference identified in prevalence when the initial prevalence was compared with that of the vaccinated (P = 0.75) and control (P = 0.94) groups after treatment. After treatment, 1 of 12 heifers, 4 of 18 first or second lactation cows, and 1 of 10 third or greater lactation cows were confirmed to be infected with strain hardjobovis, and the difference among these lactation groups was nonsignificant (P ≥ 0.61). An equal proportion of vaccinated and control cattle had positive results for strain hardjobovis (P ≥ 0.98), signifying no difference approximately 1 year after the trial commenced. A power calculation based on the assumptions of a 15% herd prevalence of Leptospira strain hardjobovis infection at the start of the trial and a vaccination protocol that would eliminate that infection revealed that the power to detect a difference between treatment groups was 20%.
Effect of the vaccine protocol on pregnancy—The probability of conceiving was not statistically different between the vaccinated and control groups, as indicated by a CI of the estimated HR of conceiving that included 1.0 (95% CI, 0.88 to 1.07; HR, 0.97 [Table 1; Figure 1]). No interactions among main effects were identified as significant. The mean number of days from the day a heifer or cow became eligible for insemination until conception was not significantly different between treatment groups. Eleven percent (108/986) of cattle in the vaccinated group failed to conceive, compared with 11.5% (104/908) in the control group.
Kaplan-Meier curves for interval to conception in Holstein dairy cattle that were (n = 986) or were not (control; 908) vaccinated against Leptospira borgpetersenii serovar Hardjo strain hardjobovis. Intervals at risk for conception began when heifers were enrolled after being bred once and for cows, after 48 days following parturition. Hatch marks signify conception. The cumulative proportion of cattle conceiving at any 1 point was not significantly different between the treatment groups.
Citation: Journal of the American Veterinary Medical Association 242, 11; 10.2460/javma.242.11.1564
Kaplan-Meier curves for interval to conception in Holstein dairy cattle that were (n = 986) or were not (control; 908) vaccinated against Leptospira borgpetersenii serovar Hardjo strain hardjobovis. Intervals at risk for conception began when heifers were enrolled after being bred once and for cows, after 48 days following parturition. Hatch marks signify conception. The cumulative proportion of cattle conceiving at any 1 point was not significantly different between the treatment groups.
Citation: Journal of the American Veterinary Medical Association 242, 11; 10.2460/javma.242.11.1564
Kaplan-Meier curves for interval to conception in Holstein dairy cattle that were (n = 986) or were not (control; 908) vaccinated against Leptospira borgpetersenii serovar Hardjo strain hardjobovis. Intervals at risk for conception began when heifers were enrolled after being bred once and for cows, after 48 days following parturition. Hatch marks signify conception. The cumulative proportion of cattle conceiving at any 1 point was not significantly different between the treatment groups.
Citation: Journal of the American Veterinary Medical Association 242, 11; 10.2460/javma.242.11.1564
Results of Cox proportional hazards analysis of the effect of injection with oxytetracycline and a monovalent vaccine against Leptospira borgpetersenii serovar Hardjo strain hardjobovis, lactation number, and DIM on the probability of conception in Holstein dairy cattle (n = 1,894).
| Variable | HR | 95% CI | P value |
|---|---|---|---|
| Vaccination (vs no vaccination) | 0.97 | 0.88–1.07 | 0.557 |
| Lactation 1 and 2 (vs 0) | 0.35 | 0.31–0.39 | < 0.001 |
| Lactation ≥ 3 (vs 0) | 0.35 | 0.31–0.40 | < 0.001 |
| DIM at conception | 0.99 | 0.99–0.99 | < 0.001 |
A value of P < 0.05 was considered significant.
The risk of heifers conceiving over the period at risk for conceiving, regardless of vaccination, was approximately 2.8 times as high as that for primiparous and pluriparous cows. During the study observation period, a similar cumulative proportion of vaccinated (96.6%) and control (96.5%) heifers became pregnant (P > 0.05). However, the cumulative proportion of pregnant heifers was greater than that of cows; the proportion of cows in their first or second lactation was similar between the vaccinated (86.7%) and control groups (86.3%), and these proportions were not significantly (P > 0.05) different from each other or from the proportion of cows in their third or later lactation (83.4% for both groups).
Effect of DIM on conception—Mean ± SE DIM at first insemination for primiparous and pluriparous cows was 61.7 ± 0.6 days and 61.2 ± 0.6 days for the vaccinated and control groups, respectively suggesting that the groups had estrus at exactly the same time (P > 0.05). The proportion that conceived by 65 DIM or from first insemination (nulliparous heifers) was similar for vaccinated (43.0%) versus control (45.8%) cattle. The same was true for those that conceived by 128 DIM (76.4% vs 76.6% for vaccinated vs control cattle, respectively).
The mean DIM at conception was 105.5 ± 1.6 days for the vaccinated group and 101.4 ± 1.6 days, and these values were not significantly different. When the nonsignificant variable of vaccination status was removed from the regression model, the variable DIM at time of conception was significant (P < 0.001), suggesting cows were more likely to conceive at an earlier DIM than at a greater DIM (HR, 0.99 [Table 1; Figures 1 and 2]).
Kaplan-Meier curves for interval to conception in Holstein dairy cattle that were (n = 986) or were not (control; 908) vaccinated against L borgpetersenii serovar Hardjo strain hardjobovis, stratified by lactation (L) number. Intervals at risk for conception began when heifers were enrolled after being bred once and for cows, after 48 days following parturition. Hatch marks signify conception. The cumulative proportion of cattle conceiving at any 1 point was not significantly different between the treatment groups or lactation classifications, with the exception of nulliparous (L = 0) heifers versus the other lactation groups. Interval to conception was significantly (P < 0.05) faster for nulliparous heifers but not different among treatment groups.
Citation: Journal of the American Veterinary Medical Association 242, 11; 10.2460/javma.242.11.1564
Kaplan-Meier curves for interval to conception in Holstein dairy cattle that were (n = 986) or were not (control; 908) vaccinated against L borgpetersenii serovar Hardjo strain hardjobovis, stratified by lactation (L) number. Intervals at risk for conception began when heifers were enrolled after being bred once and for cows, after 48 days following parturition. Hatch marks signify conception. The cumulative proportion of cattle conceiving at any 1 point was not significantly different between the treatment groups or lactation classifications, with the exception of nulliparous (L = 0) heifers versus the other lactation groups. Interval to conception was significantly (P < 0.05) faster for nulliparous heifers but not different among treatment groups.
Citation: Journal of the American Veterinary Medical Association 242, 11; 10.2460/javma.242.11.1564
Kaplan-Meier curves for interval to conception in Holstein dairy cattle that were (n = 986) or were not (control; 908) vaccinated against L borgpetersenii serovar Hardjo strain hardjobovis, stratified by lactation (L) number. Intervals at risk for conception began when heifers were enrolled after being bred once and for cows, after 48 days following parturition. Hatch marks signify conception. The cumulative proportion of cattle conceiving at any 1 point was not significantly different between the treatment groups or lactation classifications, with the exception of nulliparous (L = 0) heifers versus the other lactation groups. Interval to conception was significantly (P < 0.05) faster for nulliparous heifers but not different among treatment groups.
Citation: Journal of the American Veterinary Medical Association 242, 11; 10.2460/javma.242.11.1564
Effect of lactation status on conception—The mean number of lactations per cow in each group was not significantly different (vaccinated group, 1.8 ± 0.1 lactations; control group, 1.9 ± 0.1). In addition, the distribution of cows within the treatment groups was similar (HR, 0.95; 95% CI, 0.86 to 1.04; P = 0.26). The interval from parturition to conception did not differ between first and second lactation cows and third or later lactation cows, nor did interactions of treatment by lactation number differ. Interval to conception was significantly different between heifers and cows in their first and second lactation or third and later lactation. Lactating cows had a longer interval from first breeding to conception than heifers and had a near equal risk of conception (HR, 0.35 for first or second lactation cows and 0.35 for third or later lactation cows; Table 1; Figure 2).
Pregnancy loss—One hundred forty-three cattle were recorded as having pregnancy loss during the study period. Of these, 74 were in the vaccinated group, representing 7.5% of all vaccinates and 8.4% of all vaccinates that became pregnant (n = 878), with a distribution of loss as follows: heifers, 2 of 314 (0.6%); first or second lactation cows, 36 of 320 (11.3%); and third lactation or greater cows, 36 of 244 (14.8%). The remaining 69 cattle with an aborted pregnancy in the control group (7.6% of all control cattle or 8.6% of all control cattle that became pregnant [804]) were distributed as follows: heifers, 0 of 281 (0%); first or second lactation cows, 37 of 285 (13.0%); and third lactation or greater cows, 32 of 238 (13.4%).
Logistic regression revealed that the proportion of primiparous and pluriparous cattle that failed to maintain a conceptus was approximately equal between treatment groups. From the survival analysis, time from conception to pregnancy loss was not influenced by treatment (HR, 0.99; 95% CI, 0.70 to 1.38; P = 0.93). Primiparous cows and cows with > 1 lactation were 3 times as likely to have pregnancy loss as were nulliparous heifers, but the HR was not significant (lactations 1 and 2, HR = 3.59 [95% CI, 0.57 to 22.60]; lactation 3 or greater, HR = 3.42 [95% CI, 0.55 to 21.50]). Only lactating cows had pregnancy loss after day 30 of gestation, with half of the losses (50.4%) detected in cattle in their first or second lactation, regardless of vaccination status.
Discussion
Infection with L borgpetersenii serovar Hardjo strain hardjobovis may adversely affect the reproductive efficiency of North American dairy cattle, and an effective vaccine would be of economic importance. In the present study, a commercially available monovalent vaccine against L borgpetersenii serovar Hardjo strain hardjobovis was administered in conjunction with long-acting oxytetracycline to Holstein dairy cattle naturally infected with strain hardjobovis as is recommended to control infection with that strain. Use of this vaccine protocol failed to significantly improve reproductive efficiency during the trial period. Interval from parturition to conception and conception rate were not significantly different between vaccinated and control cattle, indicating the vaccine protocol had no effect on fertility. The results are congruent with those obtained from another study3 in which effects were evaluated of the same vaccine protocol on reproductive performance in naturally infected beef cattle. In that study3 as in ours, the vaccine protocol did not significantly improve pregnancy and calving rates.
Similar to the protocol in the other study,3 oxytetracycline was administered (20 mg/kg) to eliminate any existing or persistent Leptospira infections and to decrease transmission of leptospirosis among cattle in the herd.17 Use of antimicrobials in conjunction with initial vaccination was intended to initiate a primary antibody response against the pathogen and allow cattle to remain uninfected until the time of second vaccination between 28 and 35 days after first vaccination. In a trial23 similar to ours, although concurrent administration of oxytetracycline was not included in the vaccine protocol, a significantly increased rate of conception at first service (27% for control cattle and 36% for vaccinates) was achieved when 2 doses of the monovalent strain hardjobovis vaccine were administered. In the study reported here, the proportions of cows that conceived by 65 DIM were 36% for control cattle and 32% for vaccinates and were not significantly different; this was considered a proxy measure for first service conception rates. Although investigators in the other study23 identified an advantage of vaccination with respect to first service conception rate, overall pregnancy rate was not different between the vaccinated and control groups in their study, supporting our findings.
Whereas no significant effect of the vaccine protocol on fertility in dairy cattle was detected, lactation number and DIM at conception did affect calving interval, as was expected. A longer interval to conception and lower cumulative proportion of cattle conceiving were identified for lactating cows versus heifers. Another study33 also demonstrated lower reproductive performance of cows in their third or later lactation, compared with that in cows in their first or second lactation. Cows that have had a greater number of lactations are expected to have poorer reproductive performance because of the inverse relationship between milk production and reproduction.34 When dairy heifers, low- to moderate-producing dairy cows, and high-producing dairy cows are compared, conception rate or embryo survival rate is lower in high-producing dairy cows. Thus, milk production history rather than age or parity may be responsible for poorer reproductive performance in older cows, as opposed to vaccination status.35
In the present study, cattle inseminated later in their lactation period (at a greater number of DIM) were less likely to conceive than other cattle. One explanation for failure to conceive at an earlier DIM includes the possibility that these cows were infertile and subject to culling. The infertility may have been due to reproductive tract diseases that prevented implantation or caused embryonic death or the homeorrhectic changes of increasing milk production. Cows that failed to conceive at a greater DIM or after many services may have early embryonic loss or conceptions that were not detected. Research36,37 has shown the greatest risk of reproductive loss is during early gestation, with most conception losses occurring prior to day 15 of gestation. Classically, pathogenic bovine serovars of Leptospira were believed to cause abortions in the later stages of gestation, specifically late in the second trimester or early in the third trimester.38 However, Leptospira serovar Hardjo (including strain hardjobovis) has the ability to persist in the bovine reproductive tract39 and infection is linked to infertility and early embryonic death.40 Previous research2 has shown seropositive test results for detection of Hardjo infection is directly correlated to reduced fertility. Although the herd in the present study had a 15% prevalence of Hardjo infection, there was no observed effect of treatment on fertility.
Pregnancy loss analysis revealed no apparent effect of the vaccine protocol on failure to maintain a pregnancy. However, the study cattle had lower risk of losing a pregnancy at a greater DIM than at an earlier DIM. Although infection with serovar Hardjo was implicated in 12% of bovine abortions in 1 UK study,41 the limited number of cows that failed to maintain a pregnancy (n = 143) in the present study suggested that if strain hardjobovis negatively affects maintenance of pregnancy, that impact may be at earlier stages of gestation (early embryonic), rather than later stages, although our findings did not support this possibility. No heifers had a pregnancy loss after approximately day 30 of gestation, but this did not eliminate the possibility of earlier embryonic death. The finding that only lactating cows had pregnancy loss after day 30 of gestation, with half of the losses occurring in cattle in their first or second lactation, is similar to that reported for similar cattle management systems.42
The routine use of leptospiral vaccines other than the one used in our study complicates the interpretation of serologic testing in cattle.10 The usual herd health program was maintained throughout the trial period, including 2 pentavalent leptospiral vaccines. Therefore, the development of low titers of antibody (1:100 to 1:400) against serovars in the pentavalent vaccine (L interrogans serovars Canicola, Grippotyphosa, Icterohemorrhagica, and Pomona), excluding Hardjo, was expected, and such titers reportedly persist for up to 3 months after vaccination.10 Our serologic test results for these serovars were consistent with the expected titers, and this information was taken into account when determining herd prevalence of Leptospira serovar Hardjo strain hardjobovis. The resulting titers were also expected because of the cross-reactivity of Leptospira serovars. An animal infected with 1 serovar of Leptospira will often produce antibodies against > 1 serovar as detected via MAT,10 which explains why some cattle that had positive results for strain hardjobovis also had measurable titers for other serovars. No cattle had positive results for any Leptospira serovar other than serovar Hardjo (strain hardjobovis).
Managing leptospirosis in dairy herds requires careful analysis of the problem and practical intervention protocols. An evaluation of cost of treatment and return on investment should be assessed prior to attempting any intervention. Antimicrobial treatment is believed to decrease the probability of renal and reproductive tract lesions indicative of persistent colonization by serovar Hardjo.3 The recommended protocol for the monovalent vaccine is an initial vaccination, in conjunction with administration of long-acting antimicrobials such as oxytetracycline, followed by a booster vaccination 4 to 6 weeks later.22 We estimated the cost of implementing the vaccine protocol to be $7.20/cow/y ($4 for 2 doses of vaccine and $3.20 for 1 dose of oxytetracycline). An annual vaccine booster is also recommended. Cumulatively, the expense of the suggested Leptospira vaccine protocol for a 2,000–milk cow dairy, with 1,000 heifers entering the herd each year, would be approximately $20,000 for the first year.
Our study had limitations, and possible reasons existed to explain why the vaccine protocol did not lead to successful improvement of reproductive performance in the study herd. Throughout the study, vaccinated cattle were housed with control cattle, which might have resulted in a herd-protective immunization effect from the vaccinates. We believe this to have been possible because use of the vaccine can result in a decrease in urinary shedding of spirochetes by infected cattle22 and antimicrobial administration can also stop urinary shedding.17 However, the prevalence of leptospiruria was equivalent between groups before and after treatment, suggesting that treatment with oxytetracycline had no effect. Determination of precise herd prevalence was not a primary goal of the study, but the lack of any reproductive effect coupled with no change in prevalence suggests that the vaccine protocol was not causally associated with disease prevention in the herd.
Because the control cattle did not receive antimicrobials, a decrease in Leptospira prevalence would be expected in the vaccinated animals had the antimicrobial worked successfully. In addition, because of the possibility of vertical transmission from the dam to the fetus, diagnostic testing of fetal serum, kidney tissue, and fluids at the time of pregnancy loss might have yielded insight as to the cause of pregnancy loss.43
The lack of treatment effectiveness, whether ascribed to vaccination or antimicrobial use, suggested that current methods of controlling serovar Hardjo strain hardjobovis in Holstein dairy cattle in conditions typical for commercial milk production in California is ineffective. We acknowledge that although there was no significant difference between the study treatment groups, the sample size was too small to discern minor differences in prevalence. However, had the vaccine and antimicrobial protocol been effective, the treated group would have had no infected cattle and there would have been marked differences in interval to conception in the treated versus control group, which was not observed.
Another limitation is few apparent clinical consequences of Leptospira infection were observed in the study herd; some common clinical features not observed include pyrexia, septicemia, and jaundice.44 The literature regarding clinical infection due to serovar Hardjo in dairy cattle provides conflicting information. The clinical signs of serovar Hardjo infection can range from absent to mild, but infected dairy cattle typically have a decrease in milk production and reproductive efficiency.10 Few clinical signs of infection were expected because of the ability of host-adapted serovars (ie, serovar Hardjo) to maintain persistent infection in the host, whereas other serovars cause sporadic infections43 with more prominent clinical signs. Clinical evidence of disease was not reported for other Leptospira vaccine efficacy studies3,23 despite bacterial and serologic evidence of infection.
The observation that breeding efficiency in a dairy herd with endemic leptospirosis was not improved after inoculation with the monovalent vaccine against Leptospira borgpetersenii serovar Hardjo strain hardjobovis provides evidence that this protocol was not effective when administered to naturally infected dairy cattle. By comparing the results of the previous studies3,23 with ours, we concluded that administration of this monovalent vaccine with oxytetracycline was unable to improve reproductive efficiency in naturally infected Holstein dairy cows and heifers. Our findings conflict with those of another study22 in which administration of the same vaccine effectively eliminated urinary shedding. However, the results were also congruent with those of other research involving beef cattle,3 suggesting the vaccine protocol was not successful when administered as a treatment and prevention technique in 2 types of cattle production systems. We recommend that cattle producers and bovine practitioners carefully evaluate the potential costs and benefits prior to implementing a protocol that involves the evaluated monovalent vaccine.
ABBREVIATIONS
| CI | Confidence interval |
| DIM | Days in milk |
| HR | Hazard ratio |
| MAT | Microscopic agglutination test |
Titanium L5, Agrilabs Ltd, St Joseph, Mo.
Triangle 9 plus BVD type 2, Boehringer Ingelheim Vetmedica Inc, St Joseph, Mo.
Upjohn J5 Escherichia coli Bacterin, Pfizer Animal Health, Exton, Pa.
Ultrabac 8, Pfizer Animal Health, Exton, Pa.
Fusoguard, Novartis Animal Health US Inc, Basel, Switzerland.
Nasalgen, Intervet/Schering-Plough Animal Health, Summit, NJ.
Piliguard Pinkeye-1, Intervet/Schering-Plough Animal Health, Summit, NJ.
Brucella abortus, RB 51, Professional Biological Co/Colorado Serum Co, Denver, Colo.
Furosemide Injection 5%, Teva Animal Health Inc, St Joseph, Mo.
Diagnostic Center for Population and Animal Health, Michigan State University, East Lansing, Mich.
Spirovac, Pfizer Animal Health, Exton, Pa.
Vetrimycin 200, VetOne, MWI Veterinary Supply Inc, Meridian, Idaho.
Lactated Ringer's injection, VetOne, MWI Veterinary Supply Inc, Meridian, Idaho.
DairyCOMP 305, Valley Ag Software, Tulare, Calif.
R, version 2.15.1, R Foundation for Statistical Computing, Vienna, Austria. Available at: www.r-project.org/. Accessed Jun 1, 2010.
References
1. Grooms DL. Reproductive losses caused by bovine viral diarrhea virus and leptospirosis. Theriogenology 2006; 66: 624–628.
2. Dhaliwal GS, Murray RD, Dobson H, et al. Reduced conception rates in dairy cattle associated with serological evidence of Leptospira interrogans serovar hardjo infection. Vet Rec 1996; 139: 110–114.
3. Kasimanickam R, Whittier WD, Collins JC, et al. A field study of the effects of a monovalent Leptospira borgpetersenii serovar Hardjo strain hardjobovis vaccine administered with oxytetracycline on reproductive performance in beef cattle. J Am Vet Med Assoc 2007; 231: 1709–1714.
4. Kingscote BF. Diagnosis of Leptospira serovar hardjo infection in cattle in Canada. Can Vet J 1985; 26: 270–274.
5. LeFebvre RB, Thiermann AB, Foley J. Genetic and antigenic differences of serologically indistinguishable leptospires of serovar hardjo. J Clin Microbiol 1987; 25: 2094–2097.
6. Alves D, McEwen B, Hazlett M, et al. Trends in bovine abortions submitted to the Ontario Ministry of Agriculture, Food and Rural Affairs, 1993–1995. Can Vet J 1996; 37: 287–288.
7. Ellis WA, O'Brien JJ, Neill SD, et al. Bovine leptospirosis: serological findings from an aborted bovine fetus. Vet Rec 1982; 99: 458–459.
8. Ellis WA, O'Brien JJ, Bryson DG, et al. Bovine leptospirosis: some clinical features of serovar Hardjo infection. Vet Rec 1985; 117: 101–104.
9. Zuerner RL, Ellis WA, Bolin CA, et al. Restriction fragment length polymorphisms distinguish Leptospira borgpetersenii serovar hardjo type hardjo-bovis isolates from different geographical locations. J Clin Microbiol 1993; 31: 578–583.
10. Bolin CA. Diagnosis and control of bovine leptospirosis, in Proceedings. 6th Western Dairy Manage Conf 2003;155–159.
11. Naiman BM, Alt D, Bolin CA, et al. Protective killed Leptospira borgpetersenii vaccine induces Th1 immunity comprising responses by CD4 and gamma-delta T lymphocytes. Infect Immun 2001; 69: 7550–7558.
12. Little TWA, Hathaway SC, Broughton ES, et al. Control of Leptospira hardjo infection in beef cattle by whole-herd vaccination. Vet Rec 1992; 131: 90–92.
13. Radostits OM, Gay CC, Hinchcliff KW, et al. Diseases associated with Leptospira spp. In: Veterinary medicine. 10th ed. St Louis: WB Saunders Co, 2007; 1094–1123.
14. Hairgrove TB. Leptospirosis in cattle, in Proceedings. 37th Annu Conv Am Assoc Bovine Pract 2004; 36–39.
15. Wikse SE, Rogers GM, Ramachandran S, et al. Herd prevalence and risk factors of Leptospira infection in beef cow/calf operations in the United States: Leptospira borgpetersenii serovar Hardjo. Bovine Pract 2007; 41(1):15–23.
16. Bolin CA. Bovine leptospirosis prevalence in US dairy herds. Bovine Veterinarian 2003;Feb:14–15.
17. Alt DP, Zuerner RL, Bolin CA. Evaluation of antibiotics for treatment of cattle infected with Leptospira borgpetersenii serovar hardjo. J Am Vet Med Assoc 2001; 219: 636–639.
18. Wikse SE. Update on Leptospira hardjo-bovis control in beef herds, in Proceedings. 39th Annu Conv Am Assoc Bovine Pract 2006; 79–87.
19. MacKay RD. The economics of herd health programs. Vet Clin North Am Large Anim Pract 1981; 3: 347–374.
20. Dohoo IR. Cost of extended open period in dairy cattle. Can Vet J 1982; 23: 229–230.
21. Brown JA, LeFebvre RB, Pan MJ. Protein and antigen profiles of prevalent serovars of Leptospira interrogans. Infect Immun 1991; 59: 1772–1777.
22. Bolin CA, Alt DP. Use of a monovalent leptospiral vaccine to prevent renal colonization and urinary shedding in cattle exposed to Leptospira borgpetersenii serovar hardjo. Am J Vet Res 2001; 62: 995–1000.
23. Guard CL, Nydam DV, Eicker SW. Field trial of vaccination against Leptospira borgpetersenii serovar hardjo bovis in a single New York dairy herd, in Proceedings. 39th Annu Conv Am Assoc Bovine Pract 2006; 160–161.
24. Rajeev S, Berghaus RD, Overton MW, et al. Comparison of fluorescent antibody and microscopic agglutination testing for Leptospira in pregnant and nonpregnant cows. J Vet Diagn Invest 2010; 22: 51–54.
25. Arrayet JL, Oberbauer AM, Famula TR, et al. Growth of Holstein calves from birth to 90 days: the influence of dietary zinc and BLAD status. J Anim Sci 2002; 80: 545–552.
26. Graham TW, Breher JE, Farver TB, et al. Biological markers of neonatal calf performance: the relationship of insulin-like growth factor-I, zinc, and copper to poor neonatal growth. J Anim Sci 2010; 88: 2585–2593.
27. Pursley JR, Kosorok MR, Wiltban MC. Reproductive management of lactating dairy cows using synchronization of ovulation. J Dairy Sci 1997; 80: 301–306.
28. Cox DR. Regression models and life-tables. J R Stat Soc Series B Stat Methodol 1972; 34: 187–220.
29. Lee LA, Ferguson JD, Galligan DT. Effect of disease on days open assessed by survival analysis. J Dairy Sci 1989; 72: 1020–1026.
30. A package for survival analysis in S. R, version 2.35–8. R Foundation for Statistical Computing, Vienna, Austria. Available at: CRAN.R-project.org/web/packages/survival/survival.pdf. Accessed Jun 1, 2010.
31. Graham TW, Thurmond MC, Gershwin ME, et al. Serum zinc and copper concentrations in relation to spontaneous abortion in cows: implications for human fetal loss. J Reprod Fertil 1994; 102: 253–262.
32. Kinship: mixed-effects Cox models, sparse matrices, and modeling data from large pedigrees. R, version 1.1.0–23. R Foundation for Statistical Computing, Vienna, Austria. Available at: CRAN.R-project.org/package=kinship. Accessed Jun 1, 2010.
33. Hillers JK, Senger PL, Darlington RL, et al. Effects of production, season, age of cow, days dry, and days in milk on conception to first service in large commercial dairy herds. J Dairy Sci 1984; 67: 861–867.
34. Ray DE, Halbach TJ, Armstrong DV. Season and lactation number effects on milk production and reproduction of dairy cattle in Arizona. J Dairy Sci 1992; 75: 2976–2983.
35. Diskin MG, Morris DG. Embryonic and early foetal losses in cattle and other ruminants. Reprod Domest Anim 2008; 43: 260–267.
36. Ayalon N. A review of embryonic mortality in cattle. J Reprod Fertil 1978; 54: 483–493.
37. Forar AL, Gay JM, Hancock DD. The frequency of endemic fetal loss in dairy cattle: a review. Theriogenology 1995; 43: 989–1000.
38. BonDurant RH. Selected diseases and conditions associated with bovine conceptus loss in the first trimester. Theriogenology 2007; 68: 461–473.
39. Thiermann AB. Experimental leptospiral infections in pregnant cattle with organisms of the Hebdomadis serogroup. Am J Vet Res 1982; 43: 780–784.
40. Smith CR, McGowan MR, McClintock CS, et al. Experimental Leptospira borgpetersenii serovar hardjo infection of pregnant cattle. Aust Vet J 1997; 75: 822–826.
41. Murray RD. A field investigation of causes of abortion in dairy cattle. Vet Rec 1990; 127: 543–547.
42. Thurmond MC, Picanso JP. A surveillance system for bovine abortion. Prev Vet Med 1990; 8: 41–52.
43. Givens MD. A clinical, evidence-based approach to infectious causes of infertility in beef cattle. Theriogenology 2006; 66: 648–654.
44. Bielanski AB, Surujballi O. Leptospira borgpetersenii serovar hardjo type hardjobovis in bovine embryos fertilized in vitro. Can J Vet Res 1998; 62: 234–236.
