If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Both the economic loss and welfare implications of lameness affect the dairy industry. Currently no analgesic drugs are approved to alleviate lameness-associated pain in lactating dairy cattle in the United States. In this randomized controlled trial, 48 lactating Holsteins were enrolled to evaluate the effect of oral meloxicam and i.v. flunixin meglumine on induced lameness. Cows were allocated to 1 of 4 treatment groups (n = 12 per group): lameness and flunixin meglumine (LAME + FLU); lameness and meloxicam (LAME + MEL); lameness and placebo (LAME + PLBO); or sham induction and placebo (SHAM + PLBO). Six hours before treatment, arthritis-synovitis was induced in the distal interphalangeal joint with 20 mg of amphotericin B, whereas SHAM cows were given an intra-articular injection of an equal volume (4 mL) of isotonic saline. Cows in LAME + FLU received 2.2 mg/kg flunixin meglumine i.v. and whey protein placebo orally; LAME + MEL were administered 1 mg/kg meloxicam orally and 2 mL/45 kg sterile saline placebo i.v.; LAME + PLBO were administered 2 mL/45 kg sterile saline placebo i.v. and whey protein placebo orally; and SHAM + PLBO received 2 mL/45 kg sterile saline placebo i.v. and whey protein placebo orally. The initial treatment of MEL, FLU, or PLBO was identified as time 0 h and followed by a second dose 24 h later with data collection for 120 h. The methods used to assess analgesic efficacy were electronic pressure mat, visual lameness assessment, visual analog score, plasma cortisol concentration, plasma substance P concentration, mechanical nociception threshold, and infrared thermography imaging. Linear mixed effect modeling was the primary method of statistical analysis. Visual lameness scoring indicated a lower proportion of the FLU + LAME group was lame at the T2 h and T8 h time points in comparison to the positive controls, whereas MEL therapy resulted in a lower proportion of lame cows at the T8 h time point. Cortisol area under the effect curve was lower following FLU therapy compared with LAME + PBLO for the 0–2 h (LSM difference = 35.1 ng·h/mL, 95% CI: 6.8, 63.3 ng·h/mL), 2–8 h (LSM difference = 120.6 ng·h/mL, 95% CI: 77.2, 164.0 ng·h/mL), and 0–24 h (LSM difference = 226.0 ng·h/mL, 95% CI: 103.3, 348.8 ng·h/mL) time intervals. Following MEL therapy, cortisol area under the effect curve was lower than LAME + PLBO for both the 2 to 8 h (LSM difference = 93.6 ng·h/mL, 95% CI: 50.2, 137.0 ng·h/mL) and 0 to 24 h time intervals (LSM difference = 187.6 ng·h/mL, 95% CI: 64.9, 310.4 ng·h/mL). Analysis of data from other assessment modalities failed to discern biologically relevant differences between treatment groups. We conclude that meaningful differences were evident for visual lameness assessment and cortisol from MEL and FLU treatment versus the positive control. Further clinical research is needed toward development of a model that will create reproducible events that are more pronounced in severity and duration of lameness which can be validated as a substitute for naturally occurring lameness cases.
The clinical manifestation of lameness is due primarily to the activation of local nociceptors, which in turn activates the brain's central recognition of the painful stimulus and is often accompanied by the recruitment of inflammatory mediators (
Benchmarking cow comfort on North American freestall dairies: Lameness, leg injuries, lying time, facility design, and management for high-producing Holstein dairy cows.
). In addition to the welfare component, there is a substantial negative economic effect on the US dairy industry due to lameness. Studies have indicated that a greater prevalence of lameness is associated with decreased production (
). To reduce the associated welfare issues associated with lameness, there is a need to determine effective modalities of analgesia. In the United States, nonsteroidal anti-inflammatory drugs (NSAID), such as meloxicam (MEL) and flunixin meglumine (FLU), are commonly used by veterinarians for pain control (
). In our clinical practice, MEL and FLU are commonly used to provide analgesia associated with lameness. From this, our clinical impression is that MEL is superior to FLU as it provides a longer duration of action.
The objective of this study was to compare the analgesic effects of meloxicam via oral administration versus flunixin meglumine via i.v. administration in lactating dairy cattle for improvement of experimentally induced lameness via intra-articular injection of amphotericin B. The null hypothesis was that lameness improvement would be equivalent between FLU and MEL lameness induced treatment groups. The alternative hypothesis was that lameness would improve with meloxicam or flunixin meglumine treatment compared with no treatment (positive control).
MATERIALS AND METHODS
This project was approved by Iowa State University's Institutional Animal Care and Use Committee (Protocol # 4–17–8500-B). Cows were group housed in a freestall barn meeting the requirements of the Guide for Care and Use of Agricultural Animals in Research and Teaching (
). All cows had ad libitum access to water. Milking occurred in a parlor at 0400, 1200, and 2000 h. Milk was discarded from all animals for the duration of the trial.
Study Animals
Study animals were determined to be healthy, free from previous lameness based on treatment records, free from current lameness based on visual lameness score (VLS) equal to 0 using a modified scoring system from
, and not treated with FLU or MEL within 30 d before enrollment. Forty-eight lactating Holstein cows were enrolled and randomly allocated to 1 of 4 treatment groups using the variables parity, DIM, and current daily milk production using the RAND function in a spreadsheet (Excel, Microsoft Corp., Redmond, WA) based on the cow's identification number. The treatment groups included: lameness induction + i.v. FLU (LAME + FLU); lameness induction + oral MEL (LAME + MEL); lameness induction + placebo (LAME + PLBO); and sham lameness induction + placebo (SHAM + PLBO).
All trial personnel responsible for data collection were blinded to group assignment, and cows were block enrolled across 4 wk with 12 cows enrolled each week. During each week of the trial, groups were housed in the same pen. For sample collection, all animals were required to stand for approximately 10 min in treatment chutes, which occurred at −24, 0, 2, 8, 24, 48, 72, 96, and 120 h. Study subjects were haltered briefly for jugular venipuncture restraint. The order of sample collection was blood draw for cortisol and substance P determination, mechanical nociception threshold (MNT) measures of each rear foot, infrared thermography (IRT) imaging of each rear foot, VLS and visual analog score (VAS) determination, and pressure mat analysis.
A priori sample size calculation was conducted using a power of 0.80, with the assumption of difference in effect size of 122 kg-force (kgf; 1 kgf = 9.8 N), a standard error of 36.8, and α of 0.05 based on previous contact force studies using pressure mat technology (
). Based on the aforementioned calculation, a sample size of 12 animals per treatment group was established.
Lameness Induction
Lameness was induced in LAME + FLU, LAME + MEL, and LAME + PLBO treatment groups by injecting 20 mg of amphotericin B (X-Gen Pharmaceuticals Inc., Big Flats, NY) into the lateral distal interphalangeal joint of the left hind leg, as previously described by
. The SHAM + PLBO treatment group underwent intra-articular injection of 4 mL of sterile water, which was the equivalent volume of amphotericin B. To facilitate intra-articular injections, cows were restrained in a tilt chute and the left lateral hoof was clipped on the proximal surface of the coronary band. At this time, the limbs were visually assessed for signs of hoof lesions. The surface was then aseptically prepared with alternating scrubs of chlorhexidine and 70% isopropyl alcohol. Needle placement was confirmed using anatomical landmarks, retraction of joint fluid, and ease of injection. All injections were performed by a board-certified surgeon to facilitate adequate joint placement and diminish operator bias.
Analgesic Treatments
Cows in LAME + FLU were dosed i.v. with 2.2 mg/kg FLU (Prevail, VETONE, Boise, ID) every 24 h for 2 treatments and concurrently administered oral whey protein powder (Walgreens, Deerfield, IL) in a gelatin capsule (size #07, Torpac Inc., Fairfield, NJ) to eliminate bias associated with dosing. Cows in LAME + MEL were dosed orally with 1 mg/kg meloxicam (Mobic, Carlsbad Technology, Inc., San Diego, CA) in a gelatin capsule every 24 h for 2 treatments and concurrently administered 2 mL/45 kg sterile saline i.v. to eliminate bias associated with dosing. Placebo groups received both 2 mL/45 kg i.v. saline and whey protein powder in a gelatin capsule every 24 h for 2 treatments to mimic treatment group handling procedures without drug administration. The first of the 2 treatments was considered time 0 for the study and was approximately 6 h after lameness induction. Trial personnel who induced lameness and administered the treatments were not responsible for data collection.
Blood Sample Collection
Blood samples were collected for drug and cortisol determination in heparin tubes (BD Vacutainer, Franklin Lakes, NJ), inverted 10 times, and placed on ice for transport. For substance P, 6 mL of whole blood was transferred into blood tubes containing EDTA (BD Vacutainer, Becton Dickinson, Franklin Lakes, NJ) and 200 µg benzamidine (#D150, Santa Cruz Biotech, Dallas, TX). The tubes were inverted 10 times to facilitate mixing and placed on ice. Within 2 h of collection, all blood samples were centrifuged at 2,500 × g for 20 min at 4°C, and at least 2 mL of supernatant was placed into appropriate premarked cryovials and stored at −80°C.
Determination of Analgesic Drug Concentrations
Analgesic drug concentration in plasma was determined to confirm proper treatment administration and compare determined values in treated animals with reported efficacious plasma concentrations. Flunixin and MEL concentrations were determined using validated plasma NSAID extraction methods as previously described (
). Briefly, samples were evaluated using liquid chromatography with tandem MS with inclusion of flunixin-D3 and piroxicam as internal standards. Standard curves for both drugs ranged from 1 to 5,000 ng/mL, with r2 values >0.997 for all runs. The lower limit of quantification was 2 ng/mL and the limit of detection was 0.4 ng/mL for each drug. Quality control samples for both FLU and MEL were included at 15, 150, and 1,500 ng/mL. The quality control accuracy for FLU was 92.1, 96.9, and 97.6%, respectively, and the precision for FLU was 2.2, 1.3, and 0.9%, respectively. Accuracy for MEL was 88.4, 96.3, and 97.5%, respectively, and the precision for MEL was 4, 2.6, and 2.1%, respectively.
Primary Analgesic Assessment: Pressure Mat Analysis
Gait analysis was performed using a commercially available pressure mat system (MatScan, Tekscan Inc., South Boston, MA) at time −24, 0, 2, 8, 24, 48, 72, 96, and 120 h after first treatment administration. Research software specific to the pressure mat (HUGEMAT Research 5.83, Tekscan Inc.) was used to collect the data from the pressure mat before use at each time point. The system was calibrated with a known mass before each time point to ensure accuracy. Video synchronization was used to optimize gait and positioning of each footstep. A multitude of measurements were determined including force, contact pressure, impulse of the right hind (RH) and left hind (LH) feet, and gait length, in accordance with previously described methods (
), which assigned a 0 to 4 score at each time point. See Table 1 for a description of the scoring system. To further characterize the degree of lameness, VAS was simultaneously used to analyze grade of lameness as previously described (
). Briefly, the VAS scoring system consists of a 10-cm line drawn on paper ranging from 0 cm for no lameness to 10 cm for non-weight-bearing lameness. To determine VLS and VAS, cows were assessed on a flat, concrete surface where they were allowed to walk for 20 m, then turned and walked back to the initial starting point. The single investigator that monitored cow movement was blinded to treatment group assignment. The use of a single investigator mitigated interobserver variability. Intra-observer variability was alleviated through reviewing the scoring chart seen in Table 1 before score administration each day. No intra-observer agreement scores were calculated.
Table 1Modified visual lameness scoring (VLS) system
Plasma samples were analyzed, in duplicate, for cortisol concentration with a commercial radioimmunoassay (MP Biomedicals, Santa Ana, CA) as previously described (
). The coefficient of variation for the intra-assay variability was 8.69%, and the interassay variability was 17.28%. The limit of detection was 0.313 ng/mL, and limit of quantification was 0.625 ng/mL.
Substance P.
Samples were analyzed in duplicate using an in-house radioimmunoassay as previously described (
). The coefficient of variation for the intra-assay variability was 10.43% and the interassay variability was 25.85%. The limit of detection was 5 pg/mL, the limit of quantification was 10 pg/mL.
Mechanical Nociception Threshold.
A handheld algometer (FPX 100, Wagner Instruments, Greenwich, CT) was used to determine MNT at T−24 h, before treatment administration (T0 h), and 2, 8, 24, 48, 72, 96, and 120 h after first treatment administration. The MNT was determined using methods described by
. Mechanical nociception threshold measurements were sampled in triplicate on the lateral digit at the level of the coronary band between midline and the heel bulb. Measurements of both RH and LH limbs were recorded to individually normalize the treated limb to the unaffected RH limb for comparison. Therefore, the reported value for MNT is the difference of RH from LH. Triplicate measurements of individual limbs were averaged using arithmetic mean. To reduce bias, the same individual performed all measurements and was blinded to the treatment group assignment and to the threshold score. To eliminate bias during MNT measurement, the algometer screen was held downward during measurement and then the measured value shown to the individual responsible for recording data values, who viewed the score and assured that the algometer was cleared between measurements.
Infrared Thermography.
The lateral surface of RH and LH feet were photographed using an infrared camera (FLIR SC 660, FLIR Systems AB, Danderyd, Sweden). Infrared thermography was determined using methods described by
. Briefly, images were obtained at a 45° angle cranial to a line drawn through both rear legs, a 45° angle from the surface of the floor, and ~1 m from the coronary band with approximate distance marked. The focus circle was placed centrally on the coronary band of the lateral hoof. The focus circle for the IR camera had a diameter of 44.45 mm. Atmospheric temperature and humidity were determined at all time points from a local weather station and manually entered into the camera to calibrate it before completing animal measurements. Triplicate images of individual limbs at each time point were analyzed using research grade computer software (FLIR ExaminIR, Inc., North Billerica, MA) to determine maximum and minimum temperatures. The 3 images from each animal at each time point were compared with the unaffected RH limb for individual normalization and then averaged using arithmetic mean. Therefore, the evaluated value for IRT was the difference of mean maximum and mean minimum temperatures of the RH limb from LH limb.
Statistical Analysis
Using commercially available software (GraphPad Prism 8.2.1, La Jolla, CA), semilogarithmic transformations of individual animal plasma drug concentrations were plotted versus time. Measured peak plasma concentration (Cmax) for each animal was determined from the graph and geometric mean Cmax was determined. As FLU was injected intravenously, it likely reached peak concentrations immediately after injection. Because drug concentrations were not evaluated using a compartmental pharmacokinetic model, extrapolations to time points that were not measured cannot be determined. As a result, maximum measured drug concentration is reported as the Cmax of FLU. The time of maximum concentration (Tmax) was determined visually as the time at which Cmax occurred.
Response variables were analyzed using a mixed linear model using commercially available software (JMP version 14.3.0, SAS Institute Inc., Cary, NC), with cow as the experimental unit and using AR-1 as the covariance structure. Cows nested within treatment groups were considered random effects and treatment, time, and treatment by time interaction were designated as fixed effects. Response variables analyzed included pressure mat measurements (force, impulse, contact pressure, and gait length), VAS, cortisol, substance P, normalized MNT, and normalized IRT. Pairwise comparisons were performed using Tukey honestly significant difference (HSD) to further differentiate treatment effects of least squares means (LSM). Plasma cortisol and substance P concentrations underwent log-transformation for normality after evaluation of the residuals. Area under the effect curve (AUEC) cortisol and substance P were determined and compared using 1-way ANOVA, as previously described (
), using commercially available software (GraphPad Prism 8.2.1). Visual lameness scores were treated as categorical data, with scores compared using Fisher's exact test. To further compare VLS between groups at each time point, analysis of means for proportions tests were completed. Statistical significance was established a priori at P ≤ 0.05.
RESULTS
Study Animals
Average DIM, parity, and milk production for the respective treatment groups are reported in Table 2. One animal was removed from the study in the LAME + FLU treatment group at the T24 h mark due to misadministration of the second dose of flunixin meglumine and a second cow from the SHAM + PBLO group was removed after the T72 h time point due to an injury. Data from removed cows was included before removal from the study.
Table 2Cow characteristics (mean and SD) between treatment groups as indicated by parity, DIM, and average milk production
The geometric mean Cmax in the LAME + MEL treatment group was 1.78 μg/mL (95% CI: 1.53, 2.14 μg/mL) at a Tmax of 8 h for all cows in the group. The measured geometric mean Cmax in the LAME + FLU treatment group was 0.62 μg/mL (95% CI: 0.33, 1.19 μg/mL), with a corresponding Tmax at 2 h for all cows in the group. No cows had drug residues present in any sample that did not correspond to their treatment group assignments, with the exception of the cow removed from the study for incorrect drug administration.
Primary Analgesic Assessment Outcomes: Pressure Mat Analysis
The LSM of pressure mat measurements (force, impulse, contact pressure, and gait length) for the individual treatment groups are displayed in Table 3. There was no evidence of treatment, time, or treatment by time interaction effects for force applied by the LH foot. Least squares means for force on the LH foot were 255.35 kg (95% CI: 193.14, 371.56 kg) for LAME + MEL, 250.71 kg (95% CI: 193.51, 307.90 kg) for LAME + FLU, 253.18 kg (95% CI: 191.00, 315.36 kg) for LAME + PBLO, and 255.22 kg (95% CI: 197.91, 312.53 kg) for SHAM + PBLO.
Table 3Least squares means (95% CI) for pressure mat measurements of force, impulse, contact pressure, and gait length for the left hind limb after lameness induction (n = 12/group)
For impulse of the LH foot, there was no evidence of treatment, time, or treatment by time interaction effects. Mean values for impulse were 149.12 kg × s (95% CI: 112.11, 186.13 kg·s) for LAME + MEL, 145.05 kg·s (95% CI: 111.05, 179.05 kg·s) for LAME + FLU, 150.16 kg·s (95% CI: 113.20, 187.12 kg·s) for LAME + PBLO, and 159.60 kg·s (95% CI: 125.45, 193.76 kg·s) for SHAM + PBLO.
There was no evidence of treatment or treatment by time interaction effects for contact pressure applied by the LH foot; however, a time effect was found (P = 0.0003). Mean comparisons for contact pressure on the LH foot were 6.84 kg/cm2 (95% CI: 5.23, 8.45 kg/cm2) for LAME + MEL, 7.05 kg/cm2 (95% CI: 5.57, 8.53 kg/cm2) for LAME + FLU, 6.95 kg/cm2 (95% CI: 5.34, 8.56 kg/cm2) for LAME + PBLO, and 7.10 kg/cm2 (95% CI: 5.62, 8.58 kg/cm2) for SHAM + PBLO. Based on Tukey LSM pairwise comparisons, contact pressure had a lower LSM at T2 h compared with T48 h (LSM difference: 1.53 kg/cm2, 95% CI: 0.09, 2.97 kg/cm2), lower at T8 h than T48 h (LSM difference: 1.73 kg/cm2, 95% CI: 0.29, 3.18 kg/cm2), lower at T24 h compared with T48 h (LSM difference: 2.03 kg/cm2, 95% CI: 0.57, 3.50 kg/cm2), and lower at T24 h than T72 h (LSM difference: 1.59 kg/cm2, 95% CI: 0.12, 3.05 kg/cm2).
For gait length of the LH foot, there was no evidence of treatment, time, or treatment by time interaction effects. Least squares means for gait length were 94.15 kg·s (95% CI: 88.05, 100.25 kg·s) for LAME + MEL, 98.04 kg·s (95% CI: 92.48, 103.60 kg·s) for LAME + FLU, 98.72 kg·s (95% CI: 92.61, 104.83 kg·s) for LAME + PBLO, and 97.33 kg·s (95% CI: 91.65, 103.01 kg·s) for SHAM + PBLO.
Secondary Analgesic Assessment Outcomes
Lameness Scoring.
Data from analysis of the VAS are displayed in Table 4 and comparisons of VAS over time between groups are displayed in Figure 1. Mean values for the groups were 0.58 cm (95% CI: 0.47, 0.68 cm) for LAME + MEL, 0.60 cm (95% CI: 0.50, 0.69 cm) for LAME + FLU, 0.84 cm (95% CI: 0.71, 0.97 cm) for LAME + PBLO, and 0.40 cm (95% CI: 0.36, 0.44 cm) for SHAM + PBLO. These analyses determined there to be significant treatment (P = 0.0045), time (P < 0.0001), and treatment by time interactions (P < 0.0001). Pairwise comparisons on the day before induction (T−24 h) indicated no evidence of a difference in VAS between the 4 groups. Additionally, the SHAM + PLBO group demonstrated no evidence of a difference in VAS within the group between any time points throughout the trial. At T0 h (before first treatment induction), the treatment groups that underwent lameness induction with amphotericin B had higher VAS than the SHAM + PLBO group. Compared with the SHAM + PLBO group, the LAME + MEL was 0.87 cm higher (95% CI: 0.16, 1.57 cm), the LAME + FLU group was 0.81 cm higher (95% CI: 0.10, 1.51 cm), and the LAME + PLBO was 1.03 cm higher (95% CI: 0.33, 1.74 cm). When evaluating VAS over time in Figure 1, it appears visually that LAME + FLU is lower than LAME + PLBO group at T2 h and that the LAME + MEL group is lower than the LAME + PLBO group at T8 h. However, based on pairwise comparisons, the difference in VAS between the LAME + FLU and LAME + PLBO groups at T2 h was 0.57 cm (95% CI: −0.14, 1.27 cm) and the difference between the LAME + MEL and LAME + PLBO groups at T8 h was 0.50 cm (95% CI: −0.20, 1.20 cm). No evidence for treatment differences in VAS were visible between groups after 24 h. Thus, we were unable to show evidence of a difference from treatment over placebo in lame animals using VAS.
Table 4Least squares means (95% CI) for visual analog score, cortisol concentration, substance P concentration, mechanical nociception threshold differences, and infrared thermography difference after lameness induction (n = 12)
VAS = visual analog score, range = 0–10 cm; MNT = mechanical nociception threshold difference between left hind (LH) and right hind (RH) limb at coronary band; IRT = infrared thermography threshold difference between the maximum LH and maximum RH limb values, measured at coronary band.
Different superscript letters within a row, least squares means differ (P ≤ 0.05).
<0.0001
<0.0001
0.0031
(0.79, 1.44)
(0.73, 1.37)
(0.91, 1.55)
(−0.36, 0.28)
a,b Different superscript letters within a row, least squares means differ (P ≤ 0.05).
1 VAS = visual analog score, range = 0–10 cm; MNT = mechanical nociception threshold difference between left hind (LH) and right hind (RH) limb at coronary band; IRT = infrared thermography threshold difference between the maximum LH and maximum RH limb values, measured at coronary band.
2 LAME + MEL = oral meloxicam and induced lameness; LAME + FLU = intravenous flunixin meglumine and induced lameness; LAME + PLBO = placebo and induced lameness; SHAM + PLBO = placebo and sham induction.
Figure 1Visual assessment score (VAS; with SE) over time for treatment groups. Assessment is indicated through visual assessment marked on a 10-cm line. LAME indicates amphotericin B induction; SHAM indicates sterile saline negative control; FLU indicates flunixin meglumine treatment group administration at t = 0 and 24 h; MEL indicates meloxicam treatment group administration at t = 0 and 24 h. PLBO indicates treatment groups that received no analgesic. Inset represents the −24 to 24-h time period.
Table 5 shows the proportion of cows with a VLS > 0 by time point. On the day before induction (T−24 h) no animal had a VLS > 0. Throughout the entire trial, LAME + MEL (11.1%), LAME + FLU (13.5%), and the LAME + PBLO (28.7%) had a higher proportion of VLS > 0 than did the SHAM + PBLO (0%) (P < 0.0001). At the time of the first treatment (T0 h), all 3 groups that underwent lameness induction had similar proportion of cows with an elevated VLS (LAME + MEL 50.0%, LAME + FLU 50.0%, and LAME + PBLO 66.7%, P = 0.0063). For the T2 h and T8 h time points, the LAME + PBLO group had 58.3% of the group with an elevated VLS, whereas the LAME + FLU group showed improvement (16.7%) and LAME + MEL (41.7%) had a similar proportion of elevated VLS at the T2 h (P = 0.0085). By T8 h, both the LAME + FLU (16.7%) and the LAME + MEL (8.3%) were substantially improved over the LAME + PBLO group (P = 0.0021). For the remainder of the trial, all groups maintained similar proportions of elevated VLS (P > 0.05).
Table 5Proportion of cows with visual lameness score >0 by time point following lameness induction (n = 12/group)
Log converted cortisol concentration analyses are presented in Table 4 and Figure 2 and the AUEC comparisons are presented in Table 6 and Figure 3. Over the entire 120 h period, evaluation of plasma cortisol concentrations indicated a time (P < 0.0001) and treatment by time interaction (P = 0.0008), but no evidence of a difference for an overall treatment effect. Log converted LSM were 1.06 ng/mL (95% CI: 0.98, 1.15 ng/mL) for LAME + MEL, 0.99 ng/mL (95% CI: 0.90, 1.07 ng/mL) for LAME + FLU, 1.08 ng/mL (95% CI: 1.00, 1.17 ng/mL) for LAME + PBLO, and 0.96 ng/mL (95% CI: 0.88, 1.05 ng/mL) for SHAM + PBLO. Pairwise comparisons using Tukey HSD of log cortisol concentration on the day before induction (T−24 h) indicated no evidence of a difference in cortisol between the 4 groups and the sham induction group had no discernable difference between any time point. At T0 h (before first treatment induction), only the LAME + MEL group had a higher log cortisol concentration (LSM difference = 0.49 ng/mL, 95% CI: 0.01, 0.97 ng/mL) compared with before lameness induction. At the T2 h time point, pairwise comparisons of log cortisol concentrations indicated a lower cortisol level in LAME + FLU cows compared with the LAME + PLBO group (LSM difference = 0.68 ng/mL 95% CI: 0.20, 1.16 ng/mL). Compared with the T0 h time point, log cortisol concentrations for the LAME + PBLO group continued to increase numerically through T8 h, whereas the LAME + MEL group decreased starting at T2 h and continued down until T8 h, where the log concentration was 0.95 ng/mL (95% CI: 0.74, 1.15 ng/mL). We were unable to discern a difference between the groups that received analgesics and the placebo groups after T2 h.
Figure 2Mean cortisol concentration versus time (± SE) for treatment groups. LAME indicates amphotericin B induction; SHAM indicates sterile saline negative control; FLU represents flunixin meglumine treatment group administration at t = 0 and 24 h; MEL indicates meloxicam treatment group administration at t = 0 and 24 h; PLBO indicates treatment groups that received no analgesic. Inset represents the −24 to 24-h time period.
Table 6Least squares means (95% confidence interval) for area under the effect curve (AUEC) for cortisol and substance P after lameness induction (n = 12)
AUEC0–2 = area under the effect curve from time 0 to 2 h; AUEC2–8 = area under the effect curve from time 2 to 8 h; AUEC8–24 = area under the effect curve for 8 to 24 h; AUEC0–24 = area under the effect curve from time 0 to 24 h; AUEC24–48 = area under the effect curve for 24 to 48 h; AUEC0–48 = area under the effect curve for 0 to 48 h; AUEC0–120 = area under the effect curve from time 0 to the last sample collection.
Different superscript letters within a row, least squares means differ (P ≤ 0.05).
0.63
(2,022.3, 3368.7)
(2,070.0, 3416.5)
(1,778.4, 3124.9)
(2,396.3, 3742.8)
a,b Different superscript letters within a row, least squares means differ (P ≤ 0.05).
1 LAME + MEL = oral meloxicam and induced lameness; LAME + FLU = intravenous flunixin meglumine and induced lameness; LAME + PLBO = placebo and induced lameness; SHAM + PLBO = placebo and sham induction.
2 AUEC0–2 = area under the effect curve from time 0 to 2 h; AUEC2–8 = area under the effect curve from time 2 to 8 h; AUEC8–24 = area under the effect curve for 8 to 24 h; AUEC0–24 = area under the effect curve from time 0 to 24 h; AUEC24–48 = area under the effect curve for 24 to 48 h; AUEC0–48 = area under the effect curve for 0 to 48 h; AUEC0–120 = area under the effect curve from time 0 to the last sample collection.
Figure 3Mean cortisol area under the effect curve (AUEC) boxplots characterized across 0 to 24 h. LAME indicates amphotericin B induction; SHAM indicates sterile saline negative control; FLU represents flunixin meglumine treatment group; MEL indicates meloxicam treatment group; PLBO indicates treatment groups that received no analgesic. a, b: Different lowercase letters indicate significant differences in means at P ≤ 0.05.
Cortisol AUEC displayed treatment effects for the 0 to 2, 2 to 8 and 0 to 24 h intervals (P = 0.0010, < 0.0001, and 0.0018, respectively; Table 6). At the 0 to 2 h time interval, the AUEC for LAME + FLU was lower than LAME + PBLO (LSM difference = 35.1 ng·h/mL, 95% CI: 6.8, 63.3 ng·h/mL) and not discernable from the LAME + MEL (LSM difference = 17.3 ng·h/mL, 95% CI: −10.9, 45.6 ng·h/mL) or SHAM + PBLO (LSM difference = 7.7 ng x h/mL, 95% CI: −20.5, 36.0 ng·h/mL). Cortisol AUEC were lower for LAME + FLU than LAME + PLBO for both the 2 to 8 h (LSM difference = 120.6 ng·h/mL, 95% CI: 77.2, 164.0 ng·h/mL) and 0 to 24 h time intervals (LSM difference = 226.0 ng·h/mL, 95% CI: 103.3, 348.8 ng·h/mL). Likewise, cortisol AUEC for LAME + MEL were also lower than LAME + PLBO for both the 2 to 8 h (LSM difference = 93.6 ng·h/mL, 95% CI: 50.2, 137.0 ng·h/mL) and 0 to 24 h time intervals (LSM difference = 187.6 ng·h/mL, 95% CI: 64.9, 310.4 ng·h/mL). There were no discernable differences between any groups for the 8 to 24 h time interval and for any time intervals beyond T24 h.
Substance P.
Table 4 displays LSM for individual treatment groups and Table 6 displays substance P AUEC. The analyses were unable to discern any differences between treatments or over time for any time points. The LAME + MEL had a log LSM substance P concentration of 2.03 pg/mL (95% CI: 1.93, 2.14 pg/mL), whereas LAME + FLU, LAME + PBLO, and SHAM + PBLO had concentrations of 2.00 pg/mL (95% CI: 1.90, 2.11 pg/mL), 1.98 pg/mL (95% CI: 1.88, 2.09 pg/mL), and 2.07 pg/mL (95% CI: 1.97, 2.17 pg/mL), respectively (Figure 4).
Figure 4Substance P time course with SE for treatment groups. LAME indicates amphotericin B induction; SHAM indicates sterile saline negative control; FLU represents flunixin meglumine treatment group; MEL indicates meloxicam treatment group; PLBO indicates treatment groups that received no analgesic. Inset represents the −24 to 24-h time period.
The MNT difference LSM detected differences for treatment (P < 0.0001) and time (P < 0.0001), but not for the treatment by time interaction (Table 4). Mean MNT difference (LH − RH) values were −1.60 kgf (95% CI: −2.14, 1.06 kgf) for LAME + MEL, −2.46 kgf (95% CI: −3.00, −1.92 kgf) for LAME + FLU, −2.36 kgf (95% CI: −2.90, 1.83 kgf) for LAME + PBLO, and −0.01 kgf (95% CI: −0.37, 0.36 kgf) for SHAM + PBLO. Treatment effects were discernable between SHAM + PLBO and the other 3 groups. Pairwise comparisons on the day before induction (T−24 h) indicated no evidence of a difference in MNT difference between the 4 groups. At each time point, the SHAM + PLBO LSM oscillated around 0, indicating no difference between limbs and thus the sham induction did not create a discernable lameness with MNT (Figure 5). Pairwise comparisons were unable to demonstrate a difference between T = −24 h and T0 h for each of the groups that received amphotericin B induction, however by T2 h and continuing through T24 h, the LAME + PBLO group demonstrated a larger difference in MNT compared with T−24 h. Evaluating Figure 5, it appears that the LAME + MEL group approaches pre-induction values and that of the SHAM + PBLO group at the T8 h and T48 h time points, however we were unable to discern a difference between LAME + MEL and the other 2 lameness induction groups via Tukey HSD pairwise comparisons.
Figure 5Mechanical nociception threshold (MNT) change from baseline time course with standard error for treatment groups. Baseline is indicated by the difference between left and right hind limbs. LAME indicates amphotericin B induction, and SHAM indicates sterile saline negative control. FLU represents flunixin meglumine treatment group administration at t = 0 and 24 h. MEL indicates meloxicam treatment group administration at t = 0 and 24 h. PLBO treatment groups received no analgesic. Inset represents the −24 to 24-h time period.
Analysis of the maximum IRT difference showed a time (P < 0.0001), treatment (P < 0.0001), and treatment by time interaction (P = 0.0031; Table 4). Mean maximum IRT difference was 1.12°C (95% CI: 0.79, 1.44°C) for LAME + MEL, 1.05°C (95% CI: 0.73, 1.37°C) for LAME + FLU, 1.23°C (95% CI: 0.91, 1.55°C) for LAME + PBLO, and −0.04°C (95% CI: −0.36, 0.28°C) for SHAM + PBLO. Pairwise comparisons on the day before induction (T-24 h) indicated no evidence of a difference in maximum IRT difference between the 4 groups. SHAM + PLBO displayed an LSM difference of −0.007°C (95% CI: −0.373, 0.360°C), which oscillated around 0, indicating no difference between LH and RH thermography readings (Figure 6). For all groups that underwent lameness induction with amphotericin, larger maximum temperature differences were noted for these groups compared with the sham group at T0 h, which indicated that the inflammation initiated with the induction was discernable from the sham group. Infrared thermography was not able to discern differences between the treatment groups for the remainder of the time points. Minimum IRT differences were unable to discern any differences between groups (data not shown).
Figure 6Infrared thermography change from baseline time course with standard error for treatment groups. Baseline is indicated by the difference between left and right hind limbs. LAME indicates amphotericin B induction; SHAM indicates sterile saline negative control; FLU represents flunixin meglumine treatment group; MEL indicates meloxicam treatment group; PLBO indicates treatment groups that received no analgesic. Inset represents the −24 to 24-h time period.
). To date, the US Food and Drug Administration has approved a transdermal formulation of FLU for analgesia associated with bovine foot rot in beef and nonlactating dairy cattle (
). Its use in lameness-associated pain is extra-label drug use in lactating dairy cattle. Intravenous FLU and oral MEL can be justified for use under the Animal Medicinal Drug Use Clarification Act regulations (
). Pharmacokinetic and drug depletion profiles for meloxicam have been performed to determine withdrawal times for judicious use in lactating dairy cattle (
Before this study, it was our clinical impression that MEL worked more effectively than FLU for pain control based on clinical outcomes and pharmacokinetic data. Previously, no studies have been done to compare injectable FLU to oral MEL in lactating dairy cattle for the relief of pain caused by naturally occurring lameness. The reasons for this lack of previous studies are likely multiple, including the logistics of enrolling sufficient numbers of animals with a consistent disease definition from naturally occurring lameness cases. For this reason, lameness induction models have been developed in cattle and other species for evaluation of analgesic efficacy and lameness detection modalities. The most commonly employed model uses amphotericin B, an antifungal pharmaceutical, to produce transient arthritis and synovitis following intra-articular injection in cattle (
). Using this model, previous research has shown lameness improvement via intravenous injection of flunixin meglumine at a dose of 2.2 mg/kg in both swine (
described a difference in lameness severity at onset between the 2 groups that underwent lameness induction (mean VAS = 6.7 vs. 2.4 cm). In the current study, the model proved consistent between treatment groups with a predictable onset, as indicated by proportion of VLS > 0 and time to lameness onset. At T0 h, both VAS and VLS indicated differences in visible lameness between the groups that received amphotericin B and the SHAM group. It was evident, however, that based on the VAS and VLS assessment, only a mild degree of lameness was achieved and the lameness was short-lived, even in animals that received no analgesics (LAME + PLBO).
Assessment of visual, mechanical, and biological differences were used to assess analgesia effectiveness of FLU and MEL on lactating dairy cattle. Considering these methodologies, it was possible to determine improved analgesia from both FLU and MEL over placebo following lameness induction with VLS and cortisol, whereas other modalities failed to uncover any differences between treatment groups.
For pressure mat analysis, significant time effects were present for contact pressure, but not between treatment groups. Additionally, pressure mat analysis did not discern any effects for force applied, impulse, or gait length of the LH foot. The changes found with contact pressure are logical considering that contact pressure increased as time since induction progressed. The use of pressure mat analysis has been used in studies to gain US Food and Drug Administration approval of transdermal flunixin (
), therefore it has been validated for discerning the effect of analgesics associated with lameness pain. In that study, a different model for inducing lameness was used. Comparing this outcome to the work reported here and other work by our group (
), it is possible to conclude that pressure mat may not be accurate at detecting lameness in dairy cattle associated with this model or that this model as used doesn't create a severe enough lameness for the pressure mat to detect associated changes.
A greater degree of initial lameness may have strengthened the pressure mat's ability to differentiate between treatment groups as our induced lameness groups had T0 h VAS ranging from 1.20 to 1.45 cm. However,
, pressure mat analysis failed to discern differences between induced lameness groups that had initial VAS of 2.4 and 6.7 cm compared with 0.2 cm for the negative control group. In another study by our group, pressure mat analysis proved to be useful for some categories of assessment to determine the efficacy of meloxicam in postpartum dairy cattle that had initial VLS ≤ 1 (Kleinhenz et al., 2018). In that study, it was presumed that the pain study animals was experiencing would have been associated with uncomplicated parturition, as cows that experienced no calving difficulty were enrolled within 24 h of parturition. Nonetheless, the trial made no attempt to assess the degree and duration of assumed parturition pain so we have no metric to compare the time period the cows experienced pain to the length of time we monitored the cows.
A complicating factor in the usefulness of the pressure mat to discern treatment differences in the current study may have been associated with the large number of animals that had a VLS = 0 at T0 h following induced lameness and the duration of lameness compared with the length of the monitoring period. From T24 h to the end of the study, the largest number of cows that experienced induced lameness and a resultant VLS > 0 was 4 of 36 animals at any time point. Additionally, the categories of contact pressure and impulse have a function of force embedded in them, so effects of nonlame limbs over time are interrelated (
), which include the inability to control velocity and acceleration when moving across the mat, reliance on a single foot placement to be representative of true lameness, and walkway design that may prevent natural, uninhibited flow across the mat. In the current and previous studies by our group (Kleinhenz et al., 2018,
), the same pressure mat system was used, along with similar walkway setups and cow handling protocols. Therefore, variation between trials with regards to pressure mat limitations should have been similarly present in all 3 trials.
Visual lameness scoring is a subjective measurement that was used in this study and is commonly used by veterinarians and dairy farmers to determine the need for analgesia. Before drug administration, there was a mild degree of lameness induced in both drug treatment groups and the positive control, corresponding to VLS = 1. At T0 h, 50% of the LAME + FLU and LAME + MEL groups had a VLS > 0, whereas 66.7% of the LAME + PBLO group scored similarly. Two hours after FLU administration, the number of animals with a VLS > 0 had been reduced to 16.7%. The rapid onset of analgesia is consistent with a drug administered via the i.v. route. Similarly, MEL treatment reduced the number of animals with elevated VLS to 1 of 12 animals at T8 h and 0 animals at T24 h. No cow in the LAME + MEL group had an elevated VLS throughout the remainder of the trial. In the LAME + FLU group, 2 and 1 cow demonstrated an elevated VLS at 72 and 120 h, respectively, indicating that FLU therapy may provide a shorter duration of effectiveness. The number of cows in the LAME + PBLO that had elevated VLS remained similar at T8 h and only reduced to 25% of the group with elevated scores at T24 h. Throughout the remainder of the trial, 1 to 2 cows in the LAME + PBLO group had an elevated VLS.
In this work, we reported the measured geometric mean FLU Cmax to have a geometric mean of 0.62 μg/mL at a Tmax of 2 h. The reported Cmax is substantially lower than previous work by our group, where we reported the Cmax of 22.40 µg/mL (
Comparative plasma and interstitial fluid pharmacokinetics of flunixin meglumine and ceftiofur hydrochloride following individual and co-administration in dairy cows.
J. Vet. Pharmacol. Ther.2018; 41 (28731206): 76-82
). In that manuscript, plasma concentrations underwent pharmacokinetic modeling, which allowed us to predict the initial FLU concentration immediately after i.v. administration. In the same manuscript, plasma concentrations over time were reported, demonstrating a rapid decrease in FLU. The T2 h geometric mean plasma concentration of 0.49 µg/mL, which is similar to the values measured in the current work. The slower onset of analgesia from MEL is consistent with the oral administration of the drug. In this work, we reported Cmax of 1.78 µg/mL at a Tmax of 8 h, which is close to previous work by our group where we determined a Tmax of 11.6 h in mid-lactation dairy cows that received the same dose as the current trial (
). In that work, the Cmax was 1.80 µg/mL and the T8 h plasma concentration was 1.63 µg/mL. Unfortunately, due to the mild, transient degree of lameness, the effects of the second drug treatment at 24 h could not be adequately evaluated.
Treatment with FLU resulted in lower cortisol levels at the T2 h time point and the 0–2 h AUEC compared with the LAME + PBLO, whereas MEL treatment showed improvement by AUEC cortisol levels at the 2 to 8 h time period. As with changes in lameness scores, the plasma concentrations of cortisol correlated with increasing analgesic drug concentrations. Cortisol concentration is highly correlated with internal or external stressors (
). As cows were blocked so each treatment was equally represented in each replicate, and the cows were housed and managed similarly, the contribution of external stressors should have been represented across animals within a group simultaneously. As such, this indicates that analgesic drug treatment caused a reduction in stress levels in the face of mild lameness.
Substance P is a neuropeptide that acts as a neurotransmitter (
). In evaluation of substance P as an indicator of pain relief, no distinguishing characteristics were seen, which is consistent with previous work by our group (
). We do not believe that this discredits the ability of substance P to evaluate the effects of analgesics to decrease pain, but rather that a myriad of factors may influence substance P values including pain type, sample collection, storage, or laboratory variation that may be influencing the stability or detection ability of the molecule. Further research needs to focus on whether substance P is a reliable biomarker in analgesic studies in the bovine with naturally occurring lameness.
The use of MNT to noninvasively test for hyperalgesia has been validated previously (
). Multiple pairwise comparisons indicated that cows that underwent lameness induction with amphotericin B displayed, as a whole, a lower threshold of pressure than the negative control. These results indicated that the induction procedure was effective in causing increased sensitivity in the left limb. However, MNT did not discern differences between the groups that underwent lameness induction.
Maximum temperature difference, as measured by IRT, has been highly associated with inflammation of the foot (
). The negative control did not show apparent differences between hind limbs, indicated by a level line at zero, as all induced lameness groups trended above the negative control. These results indicate that amphotericin B induced lameness was detectable with thermography due to the collective LSM similarity between LAME + FLU, LAME + MEL and LAME + PLBO and difference from the negative control group. Though lameness was mild in nature, IRT was able to differentiate between LAME and SHAM inductions. Previous work has displayed IRT's sensitivity to inflammation associated with lameness, but decreased specificity when identifying true negative controls (
). Our work displayed adequate specificity, but poor sensitivity in its ability to detect reduction of heat following drug administration.
The induction model used for this study induced mild, short-term lameness. At the outset of the study, we intended to create more severe lameness, corresponding to a VLS > 2. In our opinion, most veterinarians and dairy producers question the biological significance of mild lameness and likely would not have administered analgesia due to the mild clinical presentation. Finding improvement in lameness pain management in clinical trials is more difficult when the degree of lameness is mild as indicated by the small changes in observations and variables. However, we were able to demonstrate the advantage of analgesics utilizing VLS and cortisol assessments, indicating that veterinarians should consider the application of analgesics for animals with mild lameness. To determine treatment effects of the analgesics in this trial, many different statistical tests were done on the data, which sometimes results in finding differences where no difference exists.
Future investigations should concentrate on creating a lameness induction protocol that will create reproducible events that are more pronounced in severity and duration of lameness. Additionally, comparisons of outcomes from induced lameness trials need to be completed with cows suffering from naturally occurring lameness, such as sole ulcers or digital dermatitis, to validate this model's ability to be a suitable substitute for naturally occurring lameness cases. The ability to achieve this goal will strengthen the potential for distinguishing analgesic efficacy in future comparison studies.
CONCLUSIONS
In comparison of the analgesics FLU and MEL, FLU had a similar effect on our outcomes as MEL in treating mild, transient lameness in lactating dairy cattle with both analgesics providing effective analgesia when compared with no treatment. These results corroborate previous studies by our group that suggest that NSAID are effective for providing analgesia in mild lameness and should be considered for treatment protocols for lameness. It is important to note that the NSAID used in this study are not labeled for use in lactating dairy cattle, and therefore, the herd veterinarian should be consulted to ensure compliance with Animal Medicinal Drug Use Clarification Act regulations, including that no violative drug residues are present in milk or meat of animals following treatment.
ACKNOWLEDGMENTS
This research was funded via a grant from the American Association of Bovine Practitioners Foundation (Ashland, OH). We thank the staff from the Iowa State University Dairy Farm and the Analytical Chemistry Section at the Iowa State University Veterinary Diagnostic Laboratory (Ames) for their assistance in conducting the trial. Last, thank you to the following individuals for their assistance with animal handling and data collection: Austin Ashbacher, Samuel Gorden, Caitlyn Dierks, Cassandra Rice, Kristin Raymond, Michael Burchard, and Wesley Oltman (Iowa State University, Ames). The authors declare that they have no conflicts of interest.
REFERENCES
Adams A.E.
Lombard J.E.
Fossler C.P.
Román-Muñiz I.N.
Kopral C.A.
Associations between housing and management practices and the prevalence of lameness, hock lesions, and thin cows on US dairy operations.
Comparative plasma and interstitial fluid pharmacokinetics of flunixin meglumine and ceftiofur hydrochloride following individual and co-administration in dairy cows.
J. Vet. Pharmacol. Ther.2018; 41 (28731206): 76-82
Benchmarking cow comfort on North American freestall dairies: Lameness, leg injuries, lying time, facility design, and management for high-producing Holstein dairy cows.
31005-5&id=gr1.jpg){kind=link}
31005-5&id=gr2.jpg){kind=link}
31005-5&id=gr3.jpg){kind=link}
31005-5&id=gr4.jpg){kind=link}
31005-5&id=gr5.jpg){kind=link}
31005-5&id=gr6.jpg){kind=link}