Beef Quality Assurance has been an important part of beef production since the 1990s. Dairy Beef Quality Assurance programs are now also beginning to gain momentum, as dairy producers are paying attention to beef quality in addition to milk quality. Dairy producers understand that because approximately 15% of their income comes from the sale of market cows, they must be concerned about the quality of beef coming from those cows. Products from market dairy cows include whole meat cuts, not just ground beef, and poor quality cuts reduce consumer satisfaction.
Previous research about injection site lesions has focused on the effects on beef quality of antibiotics, vaccines, and anti-inflammatory drugs (George et al., 1995; Van Donkersgoed et al., 1999a). Effects of reproductive hormones commonly used in the dairy industry for estrus synchronization have not been well-described. The assumption for beef quality assurance is that all injectable products cause damage, and all injections should be performed in front of the shoulder to avoid damage to high-value meat cuts. However, injection behind the shoulder is common in the dairy industry, with reported rates of at least 50% of intramuscular (i.m.) injections of reproductive hormones in the hind leg (USDA, 2007; Knust et al., 2008).
The extent of muscle damage from injections has been evaluated by measuring blood concentrations of creatine kinase (CK) from damaged muscle cells (Steiness et al., 1978; Pyörälä et al., 1994). The objective of this study was to use serum CK concentrations to assess muscle damage associated with injection of GnRH and dinoprost compared with flunixin meglumine, a drug known to damage muscle. This estimation of muscle damage could provide information to improve decision-making about injection of reproductive hormones in dairy cows. Our hypothesis was that reproductive hormones are not as damaging to muscle tissue as flunixin meglumine and would not be different from saline injection.
MATERIALS AND METHODS
The protocol was approved by the North Dakota State University Institutional Animal Care and Use Committee.
Ten healthy nonpregnant first-, second-, and third-lactation dairy cows less than 60 d after calving were enrolled in the study. All cows were normal upon physical examination and had not been treated for illness in the current lactation. Cows were housed in free stalls in a closed, unheated barn with estimated ambient temperatures between 0 and 18.3°C. The cows were provided with ad libitum access to water and a total mixed ration formulated to meet NRC requirements for lactating dairy cows (NRC, 2001).
All cows received each of 5 treatments in a balanced 2-square, 5 × 5 Latin square design (Kuehl, 2000). Cows had not been treated with any reproductive hormones or other injectable drugs since calving. Cows were injected once weekly in the semimembranosus or semitendinosus muscles with GnRH (100 µg, 2 mL; Merial, Duluth, GA), dinoprost (25 mg, 5 mL; Pfizer Animal Health, Kalamazoo, MI), flunixin meglumine (volume equivalent to that of the usual dose of dinoprost, 5 mL, or 250 mg; Intervet/Schering-Plough, Summit, NJ), saline (5 mL), or needle alone (20-gauge dry needle inserted and removed), in a crossover design. Two groups of 5 cows were used, resulting in 2 treatment periods of 5 wk each, with period 1 starting in January and period 2 in April. Although the use of flunixin meglumine i.m. is extralabel, and the Food Animal Residue Avoidance Databank discourages i.m. injection due to potential for tissue damage and unpredictable pharmacokinetics (Smith et al., 2008), we used it as a positive control because the potential for damage to tissue with i.m. injection of flunixin meglumine is well known (Pyörälä et al., 1999). The sequence of drug administration in each cow was randomly assigned using a balanced 2-square, 5 × 5 Latin square design, which allowed inclusion of first-order carryover effects in the model. The semimembranosus and semitendinosus muscles were arbitrarily divided into 6 sections (high, medium, and low on each side), and the injection was begun at the same site on each cow for the initial injection, and then systematically rotated for each subsequent injection in each cow to avoid injecting more than once into the same site.
Blood samples (5 mL) were collected via jugular venipuncture immediately before injection, and at 2, 4, 8, 24, 48, and 72 h after injection. Samples were allowed to clot at room temperature and then centrifuged (approximately 100 × g for 15 min at room temperature), and serum was removed using a pipette and placed in a sterile polystyrene tube. After separation, serum was evaluated for CK concentrations using the VetTest 8008 Chemistry Analyzer (IDEXX Laboratories, Westbrook, ME) according to the manufacturer’s instructions.
Incorporating the work that correlated presence of increased concentrations of CK with muscle damage (Steiness et al., 1978; Pyörälä et al., 1994), Lefebvre et al. (1996) delineated a useful approach to using CK to quantify muscle damage, which the authors called Q. The investigators developed an equation using area under the curve (AUC) of the time course of CK to noninvasively compare muscle damage caused by injectable products (i.e., the area under the time-concentration curve from time 0 to the time of the last measured concentration for each animal). These investigators evaluated the disposition of intravenous CK to estimate clearance of CK in cattle and used published estimates of CK bioavailability (i.e., percentage of CK released from damaged muscle cells that reaches systemic circulation) to develop a population estimate of the degree of muscle damage associated with the time course of CK in blood. The equation they developed for cattle was Q = (4.4 × 10−6) × AUCdrug, where Q = muscle damage in grams per kilogram of BW, and AUCdrug = AUC of CK after drug administration; AUCdrug is calculated by subtracting AUC of basal CK production (using CK values before drug administration) from AUC determined after drug administration. This equation and approach has been used successfully by Ferré et al. (2001, 2006) in sheep and by other workers in cattle (Pyörälä et al., 1999). In the present study, CK values immediately before treatment were subtracted from each subsequent CK concentration, and then AUC were calculated based on CK concentrations greater than baseline using the trapezoidal method (Riviere, 1999); then, Q was calculated using the formula described above.
The Q values were analyzed as a balanced, 2-square, 5 × 5 Latin square design including square, period, treatment, and first-order carryover effect as fixed effects and sequence as a random effect using PROC MIXED (SAS Inst. Inc., Cary, NC). Parameterization of the model was yijkl = μ + αi + βj + γk + ρl(i) + eijkl, where yijkl is the observation at period j and treatment k on the lth cow within the ith square, μ is the overall mean, αi is the effect of the ith square (i = 1, 2), βj is the effect of the jth period (j = 1,…,5), γk is the effect of the kth treatment (k = 1,…,5), ρl(i) is the random effect of the lth cow (l = 1,…,10), and eijkl is the random error within the ith square with for all i, k, l and Default variance component variance-covariance structure of SAS was used. Additionally, interaction terms between treatment and period, square and treatment, and period and treatment were tested in the model. Multiple comparisons between treatments were estimated using the Tukey-Kramer method. Variance partitioning between cows and within cows were calculated. Statistical significance differences were declared at P < 0.05.
Cows were treated in groups (squares) of 5, with 1 group starting in January 2009 and the other group starting in April 2009. In the first group, 2 cows were second lactation and 3 were third lactation. In the second group, 1 cow was first lactation, 2 cows second lactation, and 2 cows third lactation. One cow in the first group developed mastitis and was removed from the study after only 1 treatment. Cow numbers in each treatment group included 10 sham-, 9 saline- and flunixin meglumine-, and 8 dinoprost- and GnRH-treated animals.
Six of the 50 measurements were excluded from the analysis because Q could not be calculated due to collection of insufficient number of samples over time for CK measurement, 4 from 1 cow (cow with mastitis) and 1 from 2 cows. In the initial model including square, period, treatment, and first-order carryover effect and the treatment × square 2-way interaction term as fixed effects and sequence as a random effect, the first-order carryover effect was not significant (P = 0.4462); hence, the first-order carryover effect was excluded from the final model. The final model included square (P = 0.0324), treatment (P = 0.0132), period (P = 0.2406), and a square × treatment interaction (P = 0.0718) as fixed effects, and cow nested within square as a random effect. Least squares means (±SEM) were 11.85 (±4.60), 23.68 (±4.91), 32.54 (±4.92), 33.15 (±5.20), and 15.45 (±5.19) for needle only, saline, flunixin meglumine, dinoprost, and GnRH, respectively. Significantly greater estimated grams of muscle damage were found for dinoprost and flunixin meglumine when compared with needle only, as shown in Table 1. Comparisons of GnRH to flunixin meglumine and dinoprost were marginally significant at P < 0.15. No statistically significant difference was found for comparisons of muscle damage between the other treatments. Carryover effects between treatments were not significant (P = 0.4462). Between- and within-cow variance was 2.4 and 97.6%, respectively.
A balanced 2-square, 5 × 5 Latin square design was chosen to be able to assess potential first-order carryover effects. However, the first-order carryover effects between treatments were not significant, which implies that the wash-out period between treatments was sufficient.
Injections into areas behind the shoulder appear to be common in the dairy industry. Knust et al. (2008) surveyed dairy veterinarians who reported more frequent use of the hindquarters for reproductive hormone injections (54%) compared with 24% who reported using the neck. On the other hand, for antibiotic drugs, 67% of veterinarians reported injecting in the neck compared with 14% who injected into the hindquarters. In a survey of dairy producers (Tozer et al., 2005), 65% reported not giving injections in the neck, with the tail-head or hip, rump, and flank areas most commonly used. The average number of injections per cow per year for estrus synchronization was 2.7 in that study. Similarly, the National Animal Health Monitoring System (NAHMS) Dairy 2007 study reported upper hip and hind leg injections of reproductive drugs were performed in 69.8% of herds (USDA, 2007). If the average cow is culled after 4 lactations, that equates to approximately 11 injections related to reproductive drugs in the lifetime of the cow.
There are several possible reasons why the neck is not often used for injections of reproductive hormones in dairy cows. First, facilities used during pregnancy diagnosis and other reproductive work may make injection into the neck somewhat challenging for dairy personnel. Many farms use headlocks only for restraining cows, which makes approaching and injecting the neck area more difficult than injecting the rear end of the cow. Second, some producers and veterinarians believe that efficacy of reproductive hormones varies with site of administration. Additionally, there is no specific site listed on the FDA-approved label for dinoprost or GnRH; the label only indicates that they are i.m. injections. In published studies, site and even route of administration, such as subcutaneously in the cervical area or in the ischio-rectal fossa (Chebel et al., 2007), did not affect efficacy of dinoprost in regression of the corpus luteum. Finally, the belief that reproductive hormones are less damaging to muscle tissue than vaccines, antimicrobial drugs, and anti-inflammatory drugs may be another reason for not using the neck for injection of reproductive hormones, although even injection of sterile water can cause lesions (Van Donkersgoed et al., 1999b; Sullivan et al., 2009). Because there are no published data on the formation of injection site lesions or impact on tenderness from use of reproductive hormones, determining the amount of muscle damage caused by injection of reproductive hormones in the hindquarters helps dairy producers and veterinarians to make more informed decisions about injection practices, balancing beef quality with human safety and convenience.
The assessment of muscle damage as it affects tenderness is ideally performed postmortem using Warner-Bratzler shear force testing. We were unable, however, to justify the financial and logistical factors necessary to perform postmortem testing in this commercial facility. This would have involved tracking cows through all subsequent lactations until culling, following the animals through the slaughter process, and documenting and accounting for the influence of subsequent injections in the animals. Because an effect on muscle tissue of injection of reproductive hormones has not been documented, a reasonable first step was to implement an antemortem, less invasive method of assessing muscle damage. Consequently, the use of CK as a surrogate marker of muscle damage was selected. Creatine kinase is a muscle-specific enzyme (Sattler and Furll, 2004), although cardiac muscle also releases CK when damaged. It is unlikely that healthy animals would demonstrate increased CK from cardiac muscle, however. Given the results of this study, further studies may now be designed to follow animals to slaughter and collect tissue samples for shear force testing.
By quantifying muscle damage using AUC of CK after injection as described previously (Lefebvre et al., 1996), we found that in this study, of the 2 most commonly injected reproductive hormones in dairy cows, only dinoprost caused more muscle damage than needle alone. However, the volume of the other reproductive drug, GnRH, is less than dinoprost (2 mL compared with 5 mL). Therefore, the finding of no significant difference between GnRH and saline may be confounded with this smaller volume. Volume of injection has been associated with size of the injection site lesion (Mawhinney et al., 1996), although this was for the long-acting antibiotic oxytetracycline. We elected to use the dosages of drugs used clinically in this first investigation into the assessment of muscle damage because we had no hypothesis as to the impact of volume of injection. If the impact of volume is to be investigated as a confounding factor, another option, aside from diluting the GnRH, which would be a less clinically relevant approach, is to use a PGF2α analog such as cloprostenol. Cloprostenol requires 2 mL of product to achieve luteolysis, so this drug could also be investigated alongside the same volume of dinoprost to assess its potential contribution to injection site lesions and effect of PG volume. Given the constraints related to number of available animals, in this study, we elected to use only dinoprost because it is the more commonly used drug in ovulation synchronization protocols. Another contributing factor to the difference between dinoprost and GnRH is the vasoconstrictive potential of PG, which could result in muscle damage due to compromised blood supply. Injections of PG have been anecdotally associated with clostridial myositis, suggesting that the vasoconstrictive effects of the drug might last long enough to cause tissue damage and effects on tenderness from resulting anaerobic environment. As for the significant estimated muscle damage from injection of flunixin meglumine, this result is not surprising. Certainly, anecdotal reports of tissue damage related to i.m. or perivascular flunixin meglumine abound. In addition, a study of muscle damage using Q estimates as in this study reported significant muscle damage from i.m. flunixin meglumine at the labeled dose when compared with saline, and the volume of flunixin meglumine injected in that study was one-half the volume of saline (Pyörälä et al., 1999). The labeled dose of flunixin meglumine would be approximately 4 times the volume we administered in this study, but we were attempting to control for a volume factor with the flunixin meglumine/dinoprost comparison.
Creatine kinase values are known to have considerable variance within and between animals, and in this study variance attributed to treatment within animals (97.6%) was, as expected, larger than variance between cows (2.4%) due to relatively large differences in effect on Q by different treatments. Additionally, there was at least 1 animal whose baseline CK was greater than some of the CK values after injection. Therefore, in future studies, baseline CK should be measured more than once before injection with treatment drugs and averaged for a more accurate assessment of muscle damage in each cow. Although it is true that CK can be increased by poor sampling technique resulting in muscle damage associated with the needle stick, our observation of sampling during the study suggests that this was not a major concern because multiple needle sticks and difficulty with blood sampling were not noted.
Another potential confounding factor in measuring CK concentrations is effect of the drugs on the pharmacokinetics of CK. This is unlikely given the small dosages (flunixin meglumine) and short elimination half-lives (GnRH and dinoprost) of the drugs in this study. In addition, changes in metabolism or elimination of CK have not been reported in other studies of this nature.
The significance of actual grams of muscle damage should not be overinterpreted. This value is useful primarily as a way to compare damage associated with various injections. In light of this, verification of effect on muscle tissue such as effects on tenderness as assessed with Warner-Bratzler shear force would be the next logical step in evaluating effects of reproductive hormones on meat quality. Creatine kinase concentrations after injection in the neck as compared with semimembranosus and semitendinosus muscles could also be assessed to further inform decision making. Volume of injection and brand of PG should also be investigated because the volume of GnRH used in this study was much less than the volume of dinoprost. In addition, other PG drug products, such as cloprostenol, should be evaluated to determine if it is the drug formulation, injection volume, or the drug itself contributing to the muscle damage. Results of this study suggest that the assumption that reproductive hormones cause little to no muscle tissue damage is probably incorrect, and further studies of the effect on meat quality should be performed.