Effect of peripartal administration of mycobacterium cell wall fraction on health and fertility of Holstein cows under organic-certified management

Abstract

The main objective of this study was to evaluate the effect of peripartal administration of a commercially available nonspecific immune stimulant (mycobacterium cell wall fraction; MCWF [Amplimune, NovaVive Inc., Napanee, ON, Canada]) on the incidence of disease during early lactation and subsequent fertility of dairy cows. A second objective was to characterize the dynamics of circulating white blood cells (WBC) and metabolic markers following treatment administration. Cows in an United States Department of Agriculture (USDA) organic-certified dairy herd were blocked by parity and, based on sequential calving dates, randomly assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as a control (CON; n = 71) or MCWF (n = 65) at enrollment (7 d before expected calving) and within 24 h after calving. Blood samples were collected from a subsample of the study population (MCWF = 16; CON = 18) for WBC count at enrollment, at day 2 post enrollment, and at days 1, 3, 7, and 14 after calving. Serum fatty acids, beta-hydroxybutyrate, and Ca concentrations were determined at days 1 and 7 postpartum (MCWF = 21; CON = 21). Main outcome variables included incidence risk of peripartal and early lactation health disorders and pregnancy at first artificial insemination (AI), at 100, and at 150 days in milk (DIM). In addition, the average daily milk yield up to 90 DIM and death and live culling before 305 DIM were compared. Treatment effects were assessed using multivariable logistic regression, time-to-event analyses, and repeated measures analysis of variance (ANOVA). A treatment effect on the incidence risk of some of the health disorders in the study was established. Incidence risk of metritis and clinical mastitis <28 DIM was smaller in MCWF than in CON cows (36.9% vs. 50.7% and 6.3% vs. 19.7%, respectively). On the contrary, the incidence risk of respiratory disease <28 DIM was smaller in CON (0%) than in MCWF (7.7%). Reproductive performance of multiparous cows was affected by MCWF administration: pregnancy at first AI and pregnancy at 100 and 150 DIM were greater in MCWF than in CON (35.6% vs. 19.2%; 51.1% vs. 25.0%; and 64.4% vs. 40.4%, respectively). Overall, median intervals from calving to pregnancy were 90 vs. 121 d in MCWF and CON cows, respectively. No treatment effects on the dynamics of circulating WBC or in postpartum metabolic status were established. No differences for milk yield or for the proportion of cows that survived up to 305 DIM were determined, although cows in MCWF left the herd earlier than cows in CON. In conclusion, incidence risks of metritis and mastitis in early lactation were smaller in cows receiving MCWF, whereas the incidence risk of respiratory disease was smaller in CON. Fertility significantly improved in MCWF compared with CON cows. As this study was performed in an organic-certified dairy, specific health and reproductive management practices may affect the external validity of the current findings.

Introduction

The peripartum period is a critical time for cow health and survival and for the profitability of the starting lactation (Grummer et al., 2004). Reduced feed intake, inflammation, and immunosuppression are typical conditions during this stage that increase the risk of metabolic and infectious disorders (Bell et al., 1995; Esposito et al., 2014). Notably, disease prevalence in high-producing cows ranges from 30% to 50% during early lactation (LeBlanc et al., 2006; LeBlanc, 2010).

The systemic immunosuppression occurring during the peripartum period includes dysfunction of both polymorphonuclear and mononuclear cells (Kehrli and Goff, 1989; Kimura et al., 1999; Burton et al., 2005). In addition, hormonal and metabolic changes are thought to play a role in these alterations (Cai et al., 1994; Kimura et al., 2002). The concurrence of negative energy balance and immune suppression leaves cows susceptible to infectious diseases and metabolic disorders (LeBlanc, 2010; Esposito et al., 2014). Moreover, because a strong and well-regulated postpartum uterine inflammatory response is necessary to clear the uterine environment from pathogens (Cheong et al., 2017; Bogado Pascottini et al., 2020), an impaired immune function results in increased susceptibility to reproductive tract infections (Kehrli and Goff, 1989; Gilbert et al., 1993; Mallard et al., 1998).

The effects of disease on worsened reproductive performance have been well characterized and include impaired resumption of postpartum ovulation, compromised fertilization, and increased pregnancy loss (Santos et al., 2009; Ribeiro et al., 2013; Carvalho et al., 2019). In consequence, potential interventions to elicit nonspecific immune responses in periparturient dairy cows have been under consideration in recent years.

Strategies to enhance the immunity of the transition dairy cows have been explored. In a series of studies, the effects of recombinant bovine granulocyte colony-stimulating factor on polymorphonuclear function and numbers and on partial protection against intramammary challenge with Staphylococcus aureus were documented (Nickerson et al., 1989, 2019; Kimura et al., 2014; Van Schyndel et al., 2018; Zinicola et al., 2018).

A different compound based on mycobacteria’s ability to incite strong humoral and cellular immune responses has also been explored (Vézina and Archambault, 1997; Fumuso et al., 2003, 2007; Rogan et al., 2007; Milovanovic et al., 2018). Different mycobacterial cell wall fractions (MCWF) have been reported to have anticancer effects in humans and animals (Subramanian et al., 2016), and research in dairy calves indicated a reduction in the incidence and severity of diarrhea and pneumonia (Nosky et al., 2017; Romanowski et al., 2017, Omontese et al., 2020). In addition, the administration of a bovine MCWF formulation derived from nonpathogenic Mycobacterium phlei, resulted in reduced death and clinical signs associated with Escherichia coli K99⁺-induced diarrhea in calves (Romanowski et al., 2017). Moreover, a report by Griebel (1999) examining newborn calves showed an increment in the frequency of activated T-lymphocytes in blood and in circulating white blood cells (WBC) at 1 and 2 d post-MCWF administration.

In lactating dairy cows, repeated doses of MCWF were associated with decreased clinical signs related to persistent Mycoplasma bovis infection (Masic et al., 2017). However, the potential effect of MCWF administration on transition dairy cows has not been widely investigated.

Consequently, we hypothesized that subcutaneous administration of MCWF derived from nonpathogenic M. phlei in the proximity of calving would generate a nonspecific immune response capable of reducing the risk of peripartal infectious diseases in dairy cows. Moreover, this improved general and uterine health would translate into enhanced reproductive performance. Therefore, the main objective was to evaluate the effect of peripartal administration of a commercially available nonspecific immune stimulant (MCWF; Amplimune, NovaVive Inc., Napanee, ON, Canada) on the incidence risk of disease during early lactation and subsequent fertility of dairy cows. A second objective was to characterize the dynamics of circulating WBC and metabolic markers following treatment administration.

Materials and Methods

All the procedures conducted as part of this research were approved by the Institutional Animal Care and Use Committee at Colorado State University (protocol ID: 18-7874A). Farm managers also approved all the activities completed in this study.

Study population

The experiment was conducted in a single United States Department of Agriculture (USDA) organic-certified dairy herd located in Northern Colorado, USA, with a rolling herd milk average of 8,909 kg. The overall complex included four units and all the study cows were housed in the same unit, milking 1,800 cows. Cows within 21 d of the expected due date were housed in a maternity barn common to the four units, with wood shavings-bedding and free access to an adjacent dry lot. Fresh study cows were moved to one of the milking facilities and housed in sand-bedding free stalls with free access to an adjacent dry lot. The facilities were designed to allow access to grazing, which was provided for at least 120 d per year. After calving, all cows were milked twice a day in a 60-stall rotary parlor with an average milk yield of 30 kg/cow/d. Cows were fed a mixed ration twice a day to meet or exceed the nutritional requirements for a lactating Holstein cow producing 30 kg/d of milk with 3.5% fat and 3.1% true protein (NRC, 2001). During the grazing season (April to September), cows had access to pasture and grazing provided at least 30% of the dry matter intake of the total ration. The total mixed ration was based on corn silage (5% to 7%); wheat silage (17% to 19%); grain mix containing soybean, soy hulls, corn, wheat, and minerals and vitamins (38% to 41%); sorghum silage (5% to 7%); alfalfa hay (2%); grass hay (0% to 1.5%); and pasture grazing (estimates as 30% to 38%).

Grazing management consisted of rotational grazing in pastures based on perennial forages, alfalfa, Italian ryegrass, oat ryegrass, and teff grass. Through the study period, prepartum diet was based on corn silage (10.5% to 16.0%, DM basis), wheat silage (10.5% to 16.6%), alfalfa hay (32% to 41%), grass hay (14.2% to 20%), grain mix (12.6% to 19.2%), and mineral and vitamin mix (3.8%). Anionic salts were included in the ration (dietary cation–anion difference = −100 mEq/kg).

Due to prohibitions in the use of hormones in organic-certified systems, no artificial hormones were used to manage reproduction, and cows were bred based on artificial insemination (AI) using visual estrous detection. After 45 days in milk (DIM), all the cows were moved into insemination groups. To assess estrous behavior, the cows’ tailheads were painted with colored chalk daily and checked for chalk removal, which indicated mounting behavior. If estrus was determined, the cows were inseminated in the morning. Cows with >4 inseminations or diagnosed not pregnant at 180 DIM were transferred into pens with bulls for natural service. If cows in the bull pens were diagnosed pregnant, they were moved to pens housing only pregnant cows. Pregnancy diagnosis was performed by the attending veterinarian via ultrasonography on day 32 ± 3 after AI and reconfirmed at day 60 ± 3 of gestation or on a weekly basis for cows in the bull pens. Pregnancy was rechecked by transrectal palpation before dry-off.

Treatment allocation and monitoring procedures

The experimental design was a randomized block design with cows blocked by parity (primiparous and multiparous) and randomly assigned to the treatments, based on sequential calving dates. Cows housed at the maternity facility were considered for inclusion with an attempted distribution of 35% primiparous and 65% multiparous cows, which represents the farm population distribution. A total of 155 Holstein cows were enrolled from August to December 2018, 7 ± 2 d before their expected calving date. Within each block, cows were randomly assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as a control (CON; n = 71) or a mycobacterium cell wall fraction (MCWF; Amplimune, NovaVive Inc., Napanee, ON, Canada; n = 65) 7 d before expected calving and within 24 h after calving. The specific active ingredient consists of fragmented sections of mycobacterium cell wall with nucleic acids conserved onto it. These fragments are adjuvanted in 2% squalane, a fully hydrogenated, digestible oil that is a precursor of cholesterols. The final product is an emulsion of MCWF in squalane in phosphate-buffered saline. Randomization of cows was performed using Microsoft Office Excel (Excel RAND function) based on weekly lists from PCDART (Dairy Records Management Systems, Raleigh, NC), including cows with an expected calving date within 2 wk.

Cows that had calving-related problems (severe dystocia, defined as a cow unable to deliver the calf within 90 min post amniotic sac presentation and assisted by two people, cesarean section, fetotomy, twins, or uterine prolapse) or cows calving ≤2 or ≥14 d from the initial treatment administration were excluded from the study. In addition, to improve the accuracy of the estimated calving date, only cows that conceived after AI were considered for inclusion in the study.

After calving, the study cows were housed in a fresh pen for 28 d and then moved to different groups across the same milking unit. Cows were evaluated for body condition (Ferguson et al., 1994) on days 1 and 7 after calving. Trained personnel supervised by farm veterinarians performed daily general health inspections, including clinical mastitis diagnosis within 60 DIM in the milking parlor. In addition, a milk sample was collected at day 14 ± 3 after calving for somatic cell count (SCC). Individual milk yield (kg/d) was available from the farm’s milking machine software (ALPRO, DeLaval, Tumba, Sweeden). After 28 DIM, the study cows left the fresh pen and were moved to different groups across the farm. Consequently, the assessment and reporting of clinical mastitis and respiratory disease were divided into <28 DIM, with cows in both treatments exposed to the same conditions, and ≥28 to 60 DIM, with cows subjected to multiple pen environments.

Reproductive events, including AI and pregnancy diagnosis, were monitored up to 150 DIM. Farmworkers and attending veterinarians were blinded to the treatment assignment of the study cows. Research personnel performing the metritis and clinical endometritis assessment received a list of cow identification numbers without reference to the treatment group.

Case definitions

Stillbirth was defined as a calf that was either born dead or died within the first 24 h after birth, and a case of retained fetal membranes was defined as a cow unable to expel the membranes within 24 h after the expulsion of the offspring (Kelton et al., 1998). Cows with clinical hypocalcemia were identified as a downer-cow (in head-down recumbency with paresis of the limbs) within 24 h after parturition, responding to intravenous calcium administration (Mahjoubi et al., 2018). Clinical ketosis was defined as the presence of acetoacetate in urine that resulted in any color change in the urine test strip (Ketostix, Bayer, Leverkusen, Germany), combined with staggering, excessive object chewing, and unusual behavior. Cows diagnosed with moderate or severe ketosis were administered dextrose 5% (500 mL i.v.) for 3 d, according to the farm health protocol.

One of the authors performed a metritis diagnosis on days 7 and 14 after calving, and clinical endometritis was assessed on day 28. Diagnosis of uterine disease was done by evaluation of vaginal discharge using the Metricheck device (Metricheck, SimcroTech, Hamilton, New Zealand), considering a 0 to 5 score system: 0 = no mucus; 1 = crystalline; 2 = flecks of pus; 3 = mucopurulent <50% pus; 4 = purulent, >50% pus; 5 = watery, reddish/brownish fetid discharge (McDougall et al., 2007). The size of the uterus was assessed through per-rectum palpation. The uterine horn at the greater curvature was categorized as enlarged (larger than 1 hand) or normal (modified from Zemjanis, 1970). Metritis was defined as a mucus score of 5 within 21 d after parturition and an abnormally enlarged uterus in the absence of systemic illness and fever (Sheldon et al., 2006). Puerperal metritis (PM) was defined as an abnormally enlarged uterus and a mucus score of 5, associated with signs of systemic illness (e.g., reduced milk yield and appetite, dullness) and fever (≥39.5 °C) within 21 d after calving. According to the farm health protocol, all cows diagnosed with PM received an intrauterine infusion of Optimum UterFlush (Van Beek Natural, Science, Orange City, IA), administered every other day for three times, hypertonic saline solution (500 mL 7.2% i.v.), dextrose (500 mL 50% i.v.), and oral aspirin (5 boluses/d). Clinical endometritis was defined as vaginal discharge with a mucus score of 3 to 4 at 28 DIM (McDougall et al., 2007). Cows diagnosed with endometritis received an intrauterine infusion of 500 mL of dextrose 50%.

Cows with pyometra were diagnosed by the farm veterinarian by a transrectal uterine ultrasonography performed at 2-wk intervals, and the disorder was defined as mixed echo density fluid (pus) in the presence of a persistent corpus luteum and a history of anestrous, usually after 40 DIM (Sheldon et al., 2006). Due to prohibitions in the use of hormones in organic-certified systems, no therapeutic actions were taken in pyometra cases.

Clinical mastitis was defined as a mammary quarter with signs of inflammation (heat, pain, redness, or swelling) and/or changes in the appearance in the milk (e.g., flakes, clots, and pus; International Dairy Federation, 1971). According to farm health protocol, the treatment approach for cows with mastitis was deep stripping (hand milking) until symptoms disappeared.

Attending veterinarians diagnosed displaced abomasum by auscultation of a metallic “ping” sound in the left flank side of the abdomen, paralleled with a history of reduced consumption, reduced milk yield, and lack of rumen motility (Caixeta et al., 2018). Trained farmworkers diagnosed ruminal acidosis mainly by the presence of profuse watery diarrhea and a parallel history of reduced consumption, reduced milk yield, and poor body condition (Oetzel, 2017). Trained workers and veterinarians diagnosed respiratory disease by the auscultation of lung sounds and the presence of cough, polypnea, nasal discharge, and elevated body temperature (>40 °C), as indicated in the farm’s standard operating procedures.

Disease events (excluding PM, metritis, and clinical endometritis) and dairy herd improvement SCC were exported from on-farm recording system software PCDART.

Assessment of WBC dynamics, metabolic status, and reproductive performance

A subsample of 34 cows (MCWF = 16, CON = 18) was randomly selected for characterization of the dynamics of circulating WBC count following treatment administration. A second group of study cows was added to this subpopulation to complete 42 cows (MCWF = 21, CON = 21) for peripartal metabolic status assessment (beta-hydroxybutyrate [BHB], fatty acids [FA], and Ca serum determination). Blood samples were collected from the coccygeal vessels of the tail in the morning (0600 to 0800 hours) at enrollment (7 ± 2 d before expected calving); 2 d after enrollment, and at days 1, 3, 7, and 14 after calving using 10 mL evacuated tubes containing ethylenediaminetetraacetic acid (EDTA)-K₂ (Becton Dickinson, Franklin Lakes, NJ). Blood for WBC was homogenized, stored in a transportation cooler containing ice packs, and processed within 12 h of collection in the Clinical Pathology Laboratory of the Veterinary Teaching Hospital at Colorado State University (Fort Collins, CO). Blood samples for the assessment of metabolic status were collected at days 1 and 7 postpartum into vacuum serum tubes free of anticoagulant (Becton Dickinson, Franklin Lakes, NJ) and stored in a transportation cooler. Serum was separated within 12 h of collection (centrifuged at 3,500 rpm for 15 min at room temperature) and stored frozen at −20 °C until submission to the Clinical Pathology Laboratory of the Veterinary Teaching Hospital at Colorado State University (Fort Collins, CO). Enzymatic colorimetric determination assays with commercial kits were performed to determine FA (Catachem Inc., Oxford, CT) and BHB (Catachem Inc., Bridgeport, CT) serum concentrations. Both FA and BHB absorbance were read at 550 nm to calculate concentrations. The limit of quantification of both assays was 0.1 to 2.5 mmol/L and the intra-assay coefficient of variation was below 5%. Calcium concentrations were measured using a Roche reagent cassette, which was analyzed by the Cobas c501 analyzer (Roche, Indianapolis, IN).

Cutoff values for serum FA concentrations as a surrogate of negative energy balance were set at 0.57 mmol/L (Ospina et al., 2010). Cows with serum BHB > 0.96 mmol/L at calving or at 7 DIM were considered as hyperketonemic cows (Ospina et al., 2010). Cutoff values for subclinical hypocalcemia were Ca < 2.12 mmol/L at calving (Goff, 2008; Martinez et al., 2012) and <2.20 mmol/L at 7 DIM (Chapinal et al., 2011).

Reproductive outcomes considered the percentage of cows inseminated at 150 DIM; median intervals from calving to first AI and from calving to pregnancy; pregnancy at first AI; and the percentage of pregnant cows at 100 and 150 DIM.

Data management and statistical analyses

Sample size calculations were performed using the data analysis application SAS Power and Sample Size (release 9.4; SAS, Inst. Inc., Cary, NC). The sample size was calculated anticipating the potential positive impact of MCWF on immunity and uterine health that would result in improved fertility. Considering the reported association between health disorders, such as metritis and PM, and fertility outcomes (Ribeiro et al., 2013, 2016), the proportion of pregnant cows by 100 DIM was used in the calculations. The goal was to detect a difference of 20 percentage points in the proportion of pregnant cows by 100 DIM in favor of the MCWF group. Based on the farm data, the anticipated proportion of pregnant cows in the CON group was set at 30%. Considering power = 80% and confidence = 95%, the number of cows required to show a significant difference between the two treatment groups was 93 cows per group, as determined by a two-sided hypothesis test.

All data were exported to Microsoft Excel (Microsoft Corp., Redmond, WA), where data were organized for further analysis using SAS. Descriptive statistics were completed using PROC MEANS and the appropriate randomization of the cows at the time of enrollment for lactation number, the average number of days from enrollment to calving, and average body condition at 1 and 7 d postpartum were analyzed by univariable ANOVA (PROC MIXED).

Assumptions of normality of residuals and homogeneity of variance of continuous data were evaluated using the UNIVARIATE procedure considering the Shapiro–Wilk test. Due to lack of normality, WBC counts (10³ cells/μL), serum concentrations of FA, BHB, and Ca, and SCC in milk were log₁₀-transformed. After the analyses, the log₁₀ least square means (LSM) and confidence interval (CI) were back-transformed for reporting in the original units (Van Schyndel et al., 2018).

For all the analyses, only the first disease event was considered. However, in the case of mastitis and respiratory disease, the first disease event for each timeperiod (<28 DIM; 28 to 60 DIM) was considered. Disease frequency analyses considered incidence risk, as the time at risk of all the health-related outcomes was well delimited and short in comparison to the duration of the whole lactation (Dohoo et al., 2009). Binary variables, such as the presentation of peripartal diseases, were analyzed using logistic regression (PROC GLIMMIX). Models included treatment, parity category (primiparous; multiparous), and treatment by parity category interaction. Data from continuous variables with repeated measures, such as WBC counts and serum metabolites, were analyzed using repeated measures ANOVA (PROC MIXED) and ANOVA followed by the Tukey test (PROC MIXED) and least squared means calculation. A compound symmetry covariance structure was considered. This structure contemplates a correlation between two separate measurements, but it is assumed that the correlations of errors within subject are constant among measures. Models included treatment, parity category, DIM, and the interactions treatment by parity category and treatment by DIM.

Time-to-event analyses were included in the assessment of reproductive and survival outcomes. Hazard distributions for time to first AI and time to pregnancy were calculated using the actuarial method of the LIFETEST procedure of SAS to test the null hypothesis that the survivor functions (time from calving to a reproductive event) were identical for the two groups receiving a different treatment. Additionally, a Cox proportional regression model was developed to evaluate the effect of multiple variables (treatment, parity category, and the interaction treatment by parity category) on the risk of pregnancy (PROC PHREG).

Comparisons between treatment groups for milk yield (kg/d) were performed with PROC MIXED considering milk yield/day in the first 90 DIM. The model included the effects of treatment, parity category, time, and the interactions treatment by parity category and treatment by DIM with cow ID included in the repeated statement. A compound symmetry covariance structure was considered.

The effect of MCWF on survival up to 305 DIM (considering combined death and live culling) was assessed using logistic regression (PROC GLIMMIX) and survival analysis (PROC LIFETEST). The Wilcoxon test was used to assess the equality of survival times between treatment groups. Logistic regression models included treatment, parity category, and the interactions treatment by parity category.

For all the analyses, the interaction term treatment by parity was maintained in the models at P < 0.10. Significance and tendency levels were declared at P < 0.05 and P < 0.10, respectively.

Results

Descriptive statistics

Due to logistic constrains, the required sample size was not reached and a total of 155 Holstein cows were initially enrolled. From that population, 19 cows calved ≤2 or ≥14 d from the initial treatment administration and were excluded from the study. Consequently, 136 cows (MCWF = 65; CON = 71) completed enrollment, consisting of 103 multiparous (MCWF = 50; CON = 53) and 33 primiparous (MCWF = 15; CON = 18) cows. A total of seven cows left the herd (MCWF = 5) or died (MCWF = 1; CON = 1) within 45 DIM, with three (MCWF = 3) and four (MCWF = 3; CON = 1) cows not reaching the assessment at 14 DIM and at 28 DIM, respectively. Reasons for death were pneumonia (n = 1) and digestive impaction (n = 1), while reasons for culling were PM combined with pneumonia (n = 2), pneumonia and digestive problem (n = 2), and digestive disorders combined with injury (n = 1). Culling and disease information from these cows was considered in the subsequent analyses.

Descriptive statistics for lactation number, days from enrollment to calving, and body condition scores at 1 and 7 DIM by treatment group at enrollment are presented in Table 1. None of these descriptive variables were significantly different between the two groups.

Table 1.

Descriptive statistics by treatment group at enrollment

Variable	Treatment1	Median	Range	Mean	SD	P-value2
Lactation number	MCWF	3	1 to 6	2.66	1.34	0.42
	CON	2	1 to 7	2.48	1.32
Time from enrollment to calving, d	MCWF	7	3 to 13	7.02	4.37	0.36
	CON	6	3 to 13	7.8	5.56
Body condition score at day 1 postpartum	MCWF	2.75	2.50 to 3.25	2.85	0.21	0.95
	CON	2.75	2.50 to 3.25	2.85	0.22
Body condition score at day 7 postpartum	MCWF	2.75	2.50 to 3.25	2.7	0.19	0.61
	CON	2.75	2.25 to 3.00	2.71	0.18

¹Treatment groups, cows were assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as a CON (n = 71) or MCWF (Amplimune, NovaVive Inc., Napanee, ON, Canada; n = 65) 7 d before expected calving and within 24 h after calving.

²The reported P-values correspond to the comparison of means by treatment group.

Disease incidence risk and fertility

Table 2 reports the incidence risks of health-related events by treatment group. Although not significant, there was a greater presentation of stillbirth in MCWF vs. CON (4.6% vs. 0%; P = 0.11). Incidence risks of metritis and clinical mastitis <28 DIM were smaller in MCWF than in CON cows (36.9% vs. 50.7% and 6.3% vs. 19.7%, respectively). On the contrary, incidence risk of respiratory disease <28 DIM was smaller in CON (0%) than in MCWF (7.7% P = 0.02).

Table 2.

Incidence risk of health-related events by treatment group (% [n/total])

Health event	MCWF1	CON	P-value
Stillbirth	4.6 (3/65)	0 (0 /71)	0.11
Retention of fetal membranes	6.2 (4/65)	2.8 (2/71)	0.21
Clinical hypocalcemia	0 (0/65)	1.4 (1/71)	0.52
Clinical ketosis	10.8 (7/65)	9.9 (7/71)	0.22
PM	10.8 (7/65)	11.3 (8/71)	0.21
Metritis	36.9 (24/65)	50.7 (36/71)	0.04
Clinical endometritis	29.5 (18/61)	32.4 (23/71)	0.14
Pyometra	3.1 (2/64)	8.5 (6/71)	0.13
Clinical mastitis < 28 DIM2	6.3 (4/64)	19.7 (14/71)	0.01
Clinical mastitis ≥ 28 to 60 DIM	5.1 (3/59)	7.0 (5/71)	0.26
Clinical mastitis ≤ 60 DIM	11.9 (7/59)	22.5 (16/71)	0.05
Displacement of abomasum	3.1 (2/65)	1.4 (1/71)	0.36
Ruminal acidosis	3.1 (2/65)	1.4 (1/71)	0.36
Respiratory disease	13.8 (9/65)	5.6 (4/71)	0.06
Respiratory disease < 28 DIM	7.7 (5/65)	0 (0/71)	0.02
Respiratory disease ≥ 28 to 60 DIM	6.6 (4/61)	5.6 (4/71)	0.28

¹Treatment groups, cows were assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as a CON or MCWF (Amplimune, NovaVive Inc., Napanee, ON, Canada) 7 d before expected calving and within 24 h after calving.

²Clinical mastitis and respiratory disease include the first health event within the indicated time period. The initial models included treatment, parity category, and the interaction treatment by parity category. As no significant effect for the interaction treatment by parity was established, this term was removed from the model.

Somatic cell counts (LSM [SE]) measured at 14 ± 3 DIM showed no significant differences between treatment groups: MCWF = 221 (1.30) × 10³ cell/mL; CON = 195 (1.25) × 10³ cell/mL (P = 0.71). Similarly, the proportions of cows above 200 × 10³ cell/mL were not different between MCWF (33.8%) and CON (36.6%; P = 0.85).

Frequencies of cows inseminated and diagnosed pregnant at first AI, at 100 DIM, and 150 DIM are presented in Table 3 by treatment group and parity category. No treatment effect on the proportion of inseminated cows, on pregnancy at first AI, or on pregnancy at 100 and 150 DIM was established for primiparous cows. However, a difference of 9 percentage points in the proportion of multiparous cows inseminated (<150 DIM) was established in favor of MCWF (P = 0.04). Moreover, pregnancy at first AI and at 100 DIM in multiparous cows was about double in MCWF compared with CON cows (P = 0.035 and P = 0.045, respectively). Similarly, a difference of 24 percentage points in the proportion of multiparous cows pregnant at 150 DIM was established in favor of MCWF (P = 0.009; Table 3).

Table 3.

Frequency distributions for reproductive outcomes by treatment group and parity category

		Treatment1		P-value
Reproductive parameter	Parity	MCWF (59)	CON (70)2	Treatment3	T × P
Inseminated by 150 DIM % (n/total)	Overall	93.2 (55/59)	88.7 (63/70)	0.168	0.04
	Primiparous	78.6 (11/14)	94.4 (17/18)	0.182
	Multiparous	97.8 (44/45)	88.5 (46/52)	0.044
Pregnancy at first AI % (n/total)	Overall	32.2 (19/59)	22.5 (16/70)	0.074	0.12
	Primiparous	21.4 (3/14)	33.3 (6/18)	0.241
	Multiparous	35.6 (16/45)	19.2 (10/52)	0.035
Pregnancy by100 DIM % (n/total)	Overall	50.9 (30/59)	32.4 (23/70)	0.015	0.09
	Primiparous	50.0 (7/14)	55.6 (10/18)	0.266
	Multiparous	51.1 (23/45)	25.0 (13/52)	0.045
Pregnancy by 150 DIM % (n/total)	Overall	66.1 (39/59)	49.3 (35/70)	0.021	0.14
	Primiparous	71.4 (10/14)	77.8 (14/18)	0.291
	Multiparous	64.4 (29/45)	40.4 (21/52)	0.009

¹Treatment groups, cows were assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as a CON or MCWF (Amplimune, NovaVive Inc., Napanee, ON, Canada) 7 d before expected calving and within 24 h after calving.

²A total of seven cows left the herd (MCWF = 5) or died (MCWF = 1; CON = 1) before reaching the voluntary waiting period (45 DIM) and were excluded from the reproductive analyses.

³Models included treatment, parity category, and the interaction treatment by parity (T × P) as fixed effects.

Median ± SE time to first AI in MCWF and CON was 62.7 ± 2.9 vs. 66.5 ± 2.8 d (P = 0.37), while median intervals from calving to pregnancy were 90 vs. 121 d in MCWF and CON cows, respectively (P = 0.04). The curve originated from time to event analysis is consistent with the differential effects of treatment on the frequencies of pregnant cows at 100 and 150 DIM (Figure 1). When compared with cows in the control group, the hazard ratios (95% CI) for pregnancy up to 150 DIM in the MCWF cows were 1.63 (1.04 to 2.56; P = 0.03), 1.24 (0.54 to 2.80; P = 0.41), and 1.86 (1.07 to 3.22; P = 0.02) in overall, primiparous, and multiparous cows, respectively. No difference was observed in the mean ± SE number of AI per conception for MCWF (2.15 ± 0.21) and CON (1.78 ± 0.14) (P = 0.148).

Figure 1.

Kaplan–Meier survival curves for the proportion of nonpregnant cows by days after calving by treatment. Median intervals from calving to pregnancy for MCWF and CON were 90 vs. 121 d (P = 0.04). When compared with cows in the CON group, the hazard ratio (95% CI) for pregnancy in the MCWF cows was 1.63 (1.04 to 2.56; P = 0.03). Cows were assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as CON (n = 71) or MCWF (Amplimune, NovaVive Inc., Napanee, ON, Canada; n = 65) 7 d before expected calving and within 24 h after calving.

WBC count and metabolic status

Treatment did not have an overall effect on any of the WBC counts in the study. Accordingly, no differences were established between treatment groups at any of the sampling points (enrollment; enrollment +2 d; calving; and 2, 7, and 14 DIM).

No treatment effects on serum FA, BHB, and Ca concentrations at calving or at 7 DIM were established (Table 4). Similarly, no differences by treatment group were declared when cow metabolic status was assessed by categories, based on the proposed cutoff values. There was no difference in elevated FA at calving (MCWF: 42.9% vs. CON: 47.6%; P = 0.32) or at 7 DIM (MCWF: 26.2% vs. CON: 26.2%; P = 0.999). Proportions of cows above the BHB thresholds at 1 and 7 DIM were 0% and 4.8% for both treatment groups (P = 0.999). Similarly, no differences were determined for subclinical hypocalcemia at calving (MCWF: 16.7% vs. CON: 19.1%; P = 0.75) or at 7 DIM (MCWF: 9.52%; CON: 7.14%; P = 0.68).

Table 4.

Least square means (95% CIs) for serum FA, BHB, and calcium concentrations at 1 and 7 DIM by treatment group

		LSM (95% CI)		P-value
	Treatment1	1 DIM	7 DIM	Treatment (T)2	Parity (P)	Time (t)	T × P	T × t
FA, mmol/L	MCWF	0.63 (0.48 to 0.83)	0.58 (0.44 to 0.77)	0.775	0.831	0.143	0.075	0.477
	CON	0.71 (0.55 to 0.91)	0.56 (0.44 to 0.72)
BHB, mmol/L	MCWF	0.37 (0.30 to 0.47)	0.49 (0.39 to 0.61)	0.742	0.523	0.002	0.519	0.62
	CON	0.37 (0.31 to 0.45)	0.45 (0.37 to 0.55)
Calcium, mmol/L	MCWF	2.05 (1.96 to 2.14)	2.14 (2.05 to 2.23)	0.865	0.046	<0.001	0.967	0.369
	CON	2.03 (1.95 to 2.11)	2.17 (2.09 to 2.26)

¹Treatment groups, cows were assigned to receive two injections (5 mL s.c.) of either a placebo (saline solution) as a CON or MCWF (Amplimune, NovaVive Inc., Napanee, ON, Canada) 7 d before expected calving and within 24 h after calving.

²Models for repeated measures analyses included treatment, parity category, time, and the interactions between treatment and time and treatment and parity as fixed effects. The analyses included a subsample of 42 randomly selected cows (MCWF = 21; CON = 21).

Milk yield and survival

No significant effects on daily milk yield up to 90 DIM were established for the interaction treatment by DIM (P = 0.49) or for the main effect of treatment (MCWF = 31.8 ± 1.02 vs. CON = 31.3 ± 0.95 kg/d; P = 0.75). Parity category had a significant effect on milk yield, with multiparous cows having higher milk yield compared with primiparous cows (35.4 ± 0.72 vs. 27.8 ± 1.12 kg/d; P < 0.0001). No significant effect was determined for the interaction treatment by parity.

No treatment effect on the proportion of cows that survived up to 305 DIM was observed (P = 0.8). Overall, 14 cows from MCWF (dead = 5 and culled = 9) and 14 cows of the CON group (dead = 7 and culled = 7) left the herd. Although no difference was established in the survival functions across the 305 DIM (P = 0.59), the median interval from calving to leaving the herd was smaller in MCWF (96 d) than in CON (189 d; P = 0.03).

Discussion

This study assessed the effect of peripartal administration of a commercial nonspecific immune stimulant based on fractions of M. phlei cell wall on disease incidence and fertility of dairy cows. Additionally, the resulting dynamics of circulating WBC were analyzed.

Previous research has reported the effects of MCWF on the health of neonatal and feedlot calves (Nosky et al., 2017; Romanowski et al., 2017; Omontese et al., 2020) and early and middle lactation adult cows (Masic et al., 2017; Chesta et al., 2018). In the current study, MCWF administration had a significant effect on the incidence of some of the health disorders in analysis. A relevant finding was the smaller incidence risk of metritis demonstrated in MCWF compared with CON cows (36.9% vs. 50.7%). Another significant result was a reduction in incidence risk of clinical mastitis before 28 DIM in cows receiving MCWF (6.3%) compared with CON cows (19.7%).

Cows in MCWF had a greater presentation of stillbirth (4.6% vs. 0% for MCWF and CON, respectively; P = 0.11), although this difference was not significant. The effect of inflammatory or immune responses external to the reproductive tract on embryonic mortality has been reported (Hansen et al., 2004). However, this association is not evident in advanced gestation. Nonetheless, in approximately one-third of perinatal mortality cases, no cause of death can be detected (Berglund et al., 2003; Khodakaram-Tafti and Ikede, 2005; Jawor et al., 2017), which opens the potential for speculation about unregulated immuno-inflammatory responses, including an upregulation in production and secretion of acute-phase proteins that could be related with early calf survival.

Some of the results, however, were unexpected. Contrary to what could be anticipated, the reduction in metritis incidence risk was not reflected in subsequent levels of clinical endometritis, with both groups showing similar incidences. Also, the incidence risk of respiratory disease <28 DIM was smaller in CON (0%) than in MCWF (7.7%). Nonetheless, some of the cases of respiratory disease in MCWF cows occurred in animals that previously were diagnosed with PM (n = 2), clinical ketosis (n = 1), acidosis (n = 1), and/or displacement of abomasum (n = 1); hence, it is possible that these comorbidities made the cows more susceptible to developing respiratory signs.

Research from Huber et al. (2016) suggested that MCWF might improve the leukocytes’ cell adhesion to pathogens, antigen presentation, phagocytic capacity, and antimicrobial activity, stimulating the humoral and cellular branches of the immune system before the appearance of an infectious disease. Moreover, the effects of immunomodulatory compounds, such as trehalose 6,6′-dimycolate (TDM) and muramyl dipeptide (MDP) found in the cell wall of Mycobacterium species, have been described in recent studies. Liu et al. (2018) and Walton et al. (2018) indicated that the lung vasculature remodeling, the inflammation occurring in tuberculosis cases, and the activation of macrophages were affected by TDM. On the other hand, mice immunized with MDP showed increased IgG, IgA, and cytokine (IL-2, IL-4, IL-6, and IFN-γ) concentrations in serum, combined with elevated serum subpopulations of lymphocytes T (CD3⁺, CD4⁺, and CD8⁺) and dendritic cells (Christiana et al., 2016; Zhou et al., 2018).

Considering these reports, it is plausible to speculate a positive role for MCWF in the regulation of postpartum immunity, including uterine inflammatory responses that are crucial for clearing the uterine environment from pathogens (Cheong et al., 2017; Bogado Pascottini et al., 2020). In support of this idea, a recent study by Milovanovic et al. (2018) indicated that dairy cows administered an intrauterine infusion of MCWF had an increased influx of functional polymorphonuclear leukocytes into the uterus, and these cells had a stronger oxidative burst activity. Likewise, previous studies on MCWF in mares with post-breeding endometritis showed a reduction of cytotoxic factors that could threaten the uterus, favoring a better cleansing of uterine fluids and providing a uterine environment similar to that of normal mares resistant to endometritis (Fumuso et al., 2003, 2007; Rogan et al., 2007; Christoffersen et al., 2012; Woodward et al., 2013).

According to the studies performed on calves and lactating cows (Masic et al., 2017; Nosky et al., 2017; Romanowski et al., 2017), the main effect of MCWF is to stimulate the immune response; however, anti-inflammatory ILs (e.g., IL-10) have been shown to increase after MCWF administration in mares (Fumuso et al., 2007). In the case of the cows in our experiment, it is inferable that MWCF was able to improve the uterine environment, refining the endometrium for future pregnancy development and better reproductive performance. However, it is difficult and problematic to infer conclusions from studies performed on horses; hence, further uterine cytological analyses are encouraged for future studies to assess this hypothesis of MCWF in dairy cows in more detail.

In our study, MCWF had a significant impact on the general fertility of multiparous cows by improving the proportion of pregnant cows by 100 and 150 DIM and by reducing the median interval from calving to conception. A similar effect was reported by Chesta et al. (2018), where intrauterine application of MCWF (1.25 mg) in lactating cows without purulent discharge at 21 to 35 DIM resulted in pregnancy rate at 100 DIM of 44.4% compared with 29.2% in untreated controls.

Cow survival was an outcome of interest in the current study. The difference in the number of cows per treatment group that did not reach the end of the voluntary waiting period was an unexpected finding. Although smaller incidences of some diseases in the MCWF group suggest potential improvements in immunity, more cows did not survive in this group. Moreover, group differences in the number of cows that left the herd (MCWF = 5) or died (MCWF = 1; CON = 1) before reaching the voluntary waiting period (45 DIM) could have an effect on the fertility outcomes, as these seven cows in MCWF were excluded from the reproductive analyses. Furthermore, it is plausible to consider that if these cows had stayed in the herd, their performance had likely been compromised.

The significant effect of MCWF treatment reducing the incidence risk of cows could have contributed to an improved reproductive performance in multiparous.

Notably, the incidence of metritis in the study population (44%) was high compared with previously reported values for conventional dairies (Santos et al., 2010; Chapinal et al., 2011). As a reference, in a multistate study (Pinedo et al., 2020), incidences ranged from 21.0% to 33.9% depending on the region. Part of this discrepancy with reported values could be explained by inconsistency between observers classifying animals as healthy or metritic (Sannmann et al., 2013) and a potential for false positives in the metritis diagnosis at days 7 and 14 should be considered. Nonetheless, this high incidence could have affected the reproductive outcomes and the external validity of this study.

Another possible factor improving fertility in the MCWF group is the energy saving that might result from reduced incidence of disease, as glucose utilization markedly increases during infection (Lang and Dobrescu, 1991). As reported by Lochmiller and Deerenberg (2000) and Kvidera et al. (2017), immunoactivation following infection is characterized by decreased milk yield, inefficient feed utilization, and poor subsequent reproduction, presumably due to immune system nutrient consumption.

The association between events taking place in the peripartum and the subsequent fertility and performance of dairy cows has been reported widely. In particular, uterine diseases have great impacts on reproductive efficiency in dairy herds (Ribeiro et al., 2013; Carvalho et al., 2019). Moreover, suboptimal health predisposes cows to other fertility stressors, such as loss of body condition and anovulation (Santos et al., 2010; Ribeiro et al., 2016). Notably, these detrimental effects of disease conditions occurring during early lactation extend beyond this period, affecting subsequent biological processes, such as conception and embryo survival (Ribeiro et al., 2016; Carvalho et al., 2019).

The reduction in clinical mastitis incidence risk in MCWF could also influence fertility in this group. As reviewed by Dahl et al. (2017), several observational studies have reported that clinical or subclinical mastitis is a predisposing factor for pregnancy loss in dairy cows (Risco et al., 1999; Pinedo et al., 2009; Hernandez et al., 2012). Moreover, subclinical mastitis, measured by an SC score of ≥4.5, significantly increased the time from calving to first AI and the median interval from calving to pregnancy, while high SCC in the proximity of first breedings had a detrimental effect on conception (Pinedo et al., 2009). Nonetheless, contrary to what could be expected, the reported positive effects of MCWF administration on fertility and incidence of some diseases, including mastitis, were not accompanied by similar effects on daily milk yield up to 90 DIM.

The second objective of the current study was to characterize the dynamics of circulating WBC following treatment administration. Cellular immunity is key to protecting transition cows against opportunistic pathogens. In specific, neutrophils play a major role in separating fetal membranes after calving and maintain the host immune tolerance against uterine and mammary infections (Alhussien and Dang, 2019). In addition, the possible role of neutrophils in ovarian function (folliculogenesis and corpus luteum formation and regression), in fertilization, and in embryo implantation have been described in transition cows (Alhussien and Dang, 2019). In consequence, neutropenia around calving appears to be a strong indicator of future diseases in dairy cows and neutrophil recruitment by the uterus and subsequent activation around calving is desirable (Moretti et al., 2016; Shimizu et al., 2018).

Contrary to data reported by Griebel (1999) indicating an increment in circulating WBC at 1 and 2 d post-MCWF administration in newborn calves, no treatment effect was established on the dynamics of circulating WBC in the current study. The innate immunity of cows goes through many adaptations especially around calving, leading to impairments in general immune function. Nightingale et al. (2015) indicated that cows with an excessive acute-phase response in the early postpartum may present cellular immunosuppression, observed by the reduction of neutrophil count, reduced functions of lymphocytes, and cytokine secretion. Therefore, cellular immune suppression around calving may be one of the reasons why MCWF had no effect on WBC counts in the current study. Another possible explanation might be that the effect in circulating WBC happened previously or after the time of measurement (Christoffersen et al., 2012).

The effect of MCWF on immunity seems to be supported by potential cytokine interactions, as suggested in studies with mares (Fumuso et al., 2003, 2007). It is encouraged for future studies to evaluate the humoral response and the variation in levels of cytokines, such as IL-1β, IL-2, Il-4, IL-6, IL-8, Il-10, IL-12, TNFα, and IFNγ, as well as the changes in serum amyloid A and haptoglobin.

There are no implications to suggest a direct effect of MCWF on metabolic status. However, as elevated concentrations of FA and BHB and reduced calcium serum levels can be directly detrimental to immune function (Hammon et al., 2006; Kimura et al., 2006), these metabolites were considered in the study. Nonetheless, no treatment effects on serum FA, BHB, and Ca concentrations at calving or at 7 DIM were established.

Due to resource limitations, the required sample size for this study was not achieved. Moreover, from the 155 Holstein cows initially enrolled, 19 cows calved ≤2 or ≥14 d from the initial treatment administration and were excluded from the study. Consequently, 136 cows (MCWF = 65; CON = 71) completed enrollment. This is a weakness that resulted in power limitations for the statistical testing. Moreover, binary outcomes, such as the presentation of PM and other health disorders, require considerable differences among treatments to determine statistical significance. Another point for consideration is the USDA organic-certified status of the study dairy, where specific health, reproductive, and management practices, including grazing during a portion of the year, may affect the external validity of the current findings. In consequence, the moderated sample size and the specific health and reproductive management in agreement with the organic certification in the study farm suggest caution regarding the external validity of these findings. Further research with a larger population under multiple management settings should be conducted to confirm the results reported in this study.

Conclusions

The subcutaneous administration of MCWF in the peripartum reduced the incidence risk of metritis and clinical mastitis during early lactation, whereas the incidence risk of respiratory disease was smaller in control cows. Fertility was improved in cows receiving MCWF, which evidenced greater proportions of cows pregnant at first AI, at 100 DIM, and at 150 DIM, as compared with CON cows. No treatment effect was established on the dynamics of circulating WBC or on postpartum metabolic status. As this study was performed in an organic-certified dairy herd, specific health and reproductive management practices may affect the external validity of the current findings.

Glossary

Abbreviations

AI

artificial insemination

BHB

beta-hydroxybutyrate

DIM

days in milk

FA

fatty acid

IFN

interferon

Ig

immunoglobulin

IL

interleukin

MCWF

mycobacterium cell wall fraction

MDP

muramyl dipeptide

PM

puerperal metritis

SCC

somatic cell count

TDM

trehalose 6,6′-dimycolate

TNF

tumor necrosis factor

WBC

white blood cells

Conflict of interest statement

The authors declare that they have no competing interests. One of the authors (A.M.) worked at NovaVive Inc. at the time of the study but he did not participate in the data analyses.