Effects of delivery via pressure-adjustable pneumatic gas-powered dart gun of three antimicrobial drugs (ceftiofur crystalline free acid, tildopirosin, and tulathromycin) on drug disposition and meat quality in cattle

Animals and drug administration

A total of Forty steers were available from a project in which sires were genetically verified. A total of Forty Bos taurus-Bos indicus crossbred steers were purchased from the Department of Animal Science at Texas A & M University. Study design was a randomized block, with sires serving as blocks. The steers belonging to each sire were randomly assigned to one of four treatment groups, ten per group. There were ten sires among the 40 animals, with seven sires represented across all four treatment groups and three sires that were balanced across treatment groups as much as possible. Treatment groups were (1) 6.6 mg/kg ceftiofur crystalline free acid (CCFA, Excede, Zoetis Animal Health, Kalamazoo, MI, USA), (2) 4 mg/kg tildipirosin (TILD, Zuprevo, Merck Animal Health, Madison, NJ, USA), (3) 2.5 mg/kg tulathromycin (TULA, Draxxin, Zoetis Animal Health, Kalamazoo, MI, USA), and (4) 0.9% saline (SALN). Drugs were delivered by pressure adjustable pneumatic gas-powered dart gun (X-Caliper compressed air rifle, Pneu-dart, Inc. 15223 Route 87 Highway Williamsport, PA, 17701, USA). Based on the manufacturer’s recommendations, a pressure of 6.9 bars (100 psi) was used for all treatments except CCFA. Based on trial and error, CCFA was delivered using a pressure of 7.2 bars (104 psi) due to the weight of the dart slowing the velocity enough to prevent full penetration of the skin at the pressure used for the other drugs. A 16 gauge 1/2 inch single-port needle was used for all treatments, and animals were darted at a distance of 25 feet aiming for a region approximately 4″ below the tuber ischii in order to target the biceps femoris muscle (also known as the bottom round). Although Beef Quality Assurance guidelines recommend all injections be in the neck whenever possible, the location of injection for this study was selected based on anecdotal reports of how darts were being used in stocker cattle. In addition, it should be noted that the administration of CCFA in this manner would be considered illegal, since cephalosporins may not be administered via an extralabel route in the US (United States Food & Drug Administration Center for Veterinary Medicine, 2012), and CCFA is labeled for injection subcutaneously at the base of or the middle posterior of the ear (United States Food & Drug Administration Center for Veterinary Medicine, 2020). The actual route of administration cannot be determined when animals are darted, so some drug may be subcutaneous and some may be intramuscular. All darts appeared to completely expel their contents, but darts were not weighed before or after administration. Darts were collected after administration, and no needles appeared to remain in any animals.

A total of Five ceftiofur-, six tulathromycin-, and six tildipirosin-treated calves were randomly selected from each group of ten animals for sampling for plasma concentrations of drug. Three additional steers from the herd (in addition from the 40 described above) were administered ceftiofur as previously described at a separate time for pharmacokinetic assessments only to provide a better estimate of pharmacokinetic parameters and verify the marked variability in drug concentrations; no carcass quality or tenderness data were collected from these animals. Blood samples for drug analysis were collected at 0, 4, 8 h, and 1, 2, 3, 6 and 10 days after drug administration. Animal availability and financial considerations prevented the inclusion of animals injected via the conventional route of administration via syringe.

After treatment and blood collection, steers were assigned to one of three feedlot pens for feeding so that sire and treatment groups were balanced across pens. Steers were fed a balanced ration that included dry rolled corn, dried distillers’ grains, chopped alfalfa, molasses, and a mineral premix including lasalocid (1,320 g/ton, resulting in 33 g/ton on a dry matter basis or 30 g/ton at 90% dry matter as specified on the product label) for approximately 195 days (6.5 months). Steers were then slaughtered with captive bolt and exsanguination.

All animal procedures were approved by the Texas A&M University Institutional Animal Care and Use Committee (IACUC #2014-0316).

Drug analysis

All chemicals and reagents were obtained from VWR Scientific, Randor PA USA and ACS grade unless otherwise noted. D9-clenbuterol (C570000) and tulathromycin (T897150) were obtained from Toronto Research Chemicals, Toronto, Canada. Tildipirosin (sc-397306) was obtained from Santa Cruz Biotechnology. Ceftiofur (32442) was obtained from Sigma Aldrich. St. Louis, IL, USA. Reagent grade water (RG-water) was provided by in-house PureLab ULTRA system, ELGA Water Technologies LLC, Lowell, MA, USA.

For sample preparation, 1mL serum or plasma was aliquoted for analysis. Standard curves for the analytes of interest were prepared by fortifying the sample aliquots prior to Solid Phase Extraction (SPE) at concentrations of 0.5 to 1,000 ppb. D9-clenbuterol (40 µL) was added to all samples, calibrators and controls as an internal standard to improve the precision of the quantitative analysis. Tildipirosin and tulathromycin were analyzed by High Performance Liquid Chromatography Tandem Mass Spectrometry (LCMSMS) after isolation by SPE. SPE was processed using SPEWare CEREX48 (SPEWare Corp, Baldwin Park, CA, USA) processor and concentrator. A total of 3 mL/35 mg WWP SPE columns (SPEWare #676-0353) were used to extract the analytes of interest. SPE columns were preconditioned with 1 mL methanol, and 1 mL RG-water before application sample. Columns were washed with 1 mL RG-water prior to drying 10 min with nitrogen at 45 °C. SPE columns were eluted with 1 mL methanol before evaporation under nitrogen for 10 min at 45 °C. Residues were reconstituted with 100 µL 5% acetonitrile in RG-water prior to analysis by LCMSMS.

Ceftiofur and metabolites were analyzed after conversion to Desfuroylceftiofur Acetamide (DFCA). The residue from sample extract (from WWP) was added 450 µL 4% w/v Dithioerythritol in borate buffer pH 9 (4.75 g disodium tetraborate, and 925 mg potassium chloride in 250 mL RG-water). Samples were capped and heated at 50 °C for 15min. 200 µL Iodoacetamide solution 14% w/v in PO₄ buffer (0.025 M, pH7) was added to the vials and then heated at 50 °C for an additional 15 min. Samples were cooled and 500 µL 1 N acetic acid was added prior to SPE. A total of 3 mL/35 mg Trace-B (SPEWare # 711-335M) were used to extract the DFC using the equipment mentioned above. Columns were preconditioned with 1 mL methanol, 1 mL RG-water, and 500 µL 1 N acetic acid. After sample application, columns were washed with 1 mL RG-water, 500 µL 1 N acetic acid and 1 mL methanol. Columns were then dried under nitrogen at 45 °C prior to elution with 2% ammonium hydroxide in methanol. Eluents were dried under nitrogen at 45 °C. Residues were reconstituted with 100 µL 5% acetonitrile in RG-water prior to analysis by LCMSMS.

High Performance Liquid Chromatography tandem Mass Spectrometry was performed on a TSQ Endura, Thermo Instruments, San Jose, CA, USA and Agilent 6410 Triple Quadrupole LC/MS. Analytes were separated on an Acentis Express C18 2.1 $\times$ 100 mm HPLC column (#53823-U, Supelco Inc. Bellefonte, PA, USA) using a mobile phase of 0.1% formic acid in both RG-water (solvent A) and LC-MS grade acetonitrile (solvent B). A solvent gradient was applied to the column at a flow of 300 µL/min of 5% B to 95% B over 8 min. The LCMSMS was run in electrospray mode with a 0.7 mass resolution on both quadrupoles.

Pharmacokinetic analysis

Noncompartmental analysis was performed to estimate the pharmacokinetic parameters in plasma for each individual animal for each drug (see Tables 1–3 for parameters).

Post-mortem carcass quality and tenderness assays

Steers were humanely slaughtered at the Rosenthal Meat Science and Technology Center, Texas A&M University, College Station, TX, USA. Carcasses were chilled for 48 h postmortem and the Bicep femoris from the right and left sides was removed for further evaluation. The Longissimus dorsi lumborum was also removed from one side of each carcass. Products from animals treated with CCFA did not enter the food supply.

Live Weight of the animal just prior to slaughter was measured. Dressing percentage was defined as hot carcass weight divided by live weight times 100, where hot carcass weight was determined prior to chilling. Carcasses were ribbed at the 12^th and 13^th rib interface. Marbling score, ribeye area (cm²), kidney, pelvic and heart fat (%), and subcutaneous fat thickness (mm) were determined (USDA, 2017). Yield grade and quality were calculated (USDA, 2017).

After USDA grade assessment, color CIE (International Commission of Light, Vienna, Austria) L * (lightness; 100 = white and 0 = black), a * (overall hue; 100 = red and 0 = green), and b * (overall hue; 100 =y ellow and 0 = blue) color space values were assessed (Minolta Colorimeter, CR-300, 8 mm diameter head, 10° standard observer, D65 light source; Minolta Co., Ramsey, NJ, USA). White and black tiles were used daily to calibrate the Minolta Colorimeter. At the 12^th rib Longissimus dorsi interface, three values were obtained per carcass and averaged. A total of three random pH values were obtained from the same location using a puncture pH probe (Hach, H100, Loveland, CO, USA) and averaged. This value was defined as the final carcass pH. The pH meter was calibrated (pH 4.0 and pH 7.0 standard buffers) at the start of each day. Color and pH were determined on the same lean surface as used for grade assessment. Hump height measurement of the Rhomboideus muscle (hump) was taken at its highest point.

The presumed site of dart entry was identified via visual examination for increased connective tissue or marbling or by digital palpation for slightly thicker, denser areas of muscle by a trained veterinarian with no knowledge of treatment. From the darted side in the bottom round muscle, 5–2.54-cm bottom round steaks were cut, the first steak included the presumed site of the dart entry (Position 1), and then two proximal (Positions 2 and 3) and two distal (Positions 4 and 5) 2.54 cm steaks were removed from the presumed site of dart entry. One corresponding 2.54 cm steak from the control side (Position 6) was also collected in the similar anatomical location from the dart side.

From the Longissimus dorsi lumborum muscle, beginning at the anterior edge, 4–2.54 cm top loin steaks were cut, vacuum-packaged and randomly assigned to aging for 1, 7, 14, or 21 days at 4 °C (labeled as Positions 1–4 in Table 4). Bottom round steaks were vacuum-packaged and aged for 14 days. For Warner–Bratzler shear force determination, top loin and bottom round steaks were weighed, iron constantan thermocouples (TT-J-36-SLE, Omega Engineering, Inc., Stanford, CT, USA) were inserted in the geometric center of each steak, and steaks were cooked on flat top electric skillets (Hamilton Beach grill, Hamilton Beach/Proctor-Silex, Inc., Southern Pines, NC, USA) to a final internal temperature of 70 °C. Internal cook temperature was monitored with hand-held recorders (model HH-21, Omega Engineering, Inc., Stanford, CT, USA). Steaks were turned after reaching 35 °C internal temperature. After cooking, cooked weight was recorded and cook yield was calculated (Cook Yield = 100 − ((Cooked Chop Weight/Raw Chop Weight) * 100).

Cooked steaks were chilled for 24 h and then one 1.27 cm diameter core, parallel to the longitudinal orientation of the muscle fibers, was removed from six-seven positions designated on each steak. Steaks were assigned codes prior to cooking and laboratory personnel were blinded to treaments. Cores were sheared once with Warner-Bratzler shearing device (United Smart-1 Test System SSTM-500, United Calibration Corp., Huntington Beach, CA, USA; 200 kg load cell, 200 mm/min head speed) certified by United Testing Systems, Inc. using a Warner–Bratzler stainless steel blade (1.168 cm thick). Mis-shaped cores were not used. The maximum force in kg for cores from one steak were averaged.

Data analysis

Carcass grade data, color, pH and Warner-Bratzler shear force measures were analyzed using Analysis of Variance with the Proc GLM procedure of SAS (v9.4, SAS Institute, Cary, NC, USA) with an alpha of < 0.05. For grade data, color and pH, carcass was defined as the experimental unit. The dart treatment was defined as a main effect and the pen cattle were fed in was defined as a fixed effect. Warner–Bratzler shear force data were analyzed for top loin steaks. Main effects were defined as dart treatment, steak position, aging day and their subsequent two-way interactions and pen fed was used as a fixed effect. For bottom round steaks, data were analyzed with dart treatment and position of steak from the injection site, and their interaction as a main effects. The pen cattle were fed in was also defined as a fixed effect. A third analysis for bottom round steaks compared sides within a carcass for Warner–Bratzler shear force. The undarted side was defined as a control and the darted side was defined as the treatment. For all analyses, least squares means were calculated. Differences in least squares means were determined using the Tukey–Kramer function within SAS for multiple mean comparisons.

Drug concentrations and pharmacokinetics

Mean drug concentrations at sampling time points for the 3 drugs are shown in Figs. 1–3, and mean and median pharmacokinetic estimates are in Tables 1–3.

Carcass quality and tenderness

Carcass quality and yield factors by dart treatment indicated that dressing percentage and L* values were significantly different (P = 0.05 and P < 0.001, respectively) across treatment groups (Table 5). Dressing percentage was lower (P = 0.05) in cattle treated with SALN than cattle treated with CCFA and TILD. CIE color space values indicated L* values were significantly higher (P < 0.001) in cattle treated with CCFA, TILD and TULA resulting in lighter lean color than the cattle treated with the SALN control. Dart treatments had no effect on other carcass quality and yield grade factors.

As expected, top loin steak Warner-Bratzler shear force values across all treatments and positions were lower (P < 0.001) at 14 and 21 d age postmortem than at 1 and 7 d aging times (Table 4). Shear force values were also lower (P < 0.001) in steaks aged 7 d when compared to steaks 1 d postmortem. Neither the dart treatments (P = 0.21) nor position within the strip loin (P = 0.73) affected Warner–Bratzler shear force values.

There was no difference (P > 0.05) among core sample tenderness within steaks. There was also no difference in tenderness between steaks collected at the various positions within the bottom rounds (Table 4). All steaks (from darted and non-darted sides) from cattle treated with ceftiofur and saline were more tender than from cattle treated with tulathromycin or tildipirosin (P = 0.003).

There was a trend (P = 0.08) toward more tenderness in steaks from the non-darted compared to the darted side (Table 4). Steaks from the darted side for one treatment, tildipirosin, were less tender than the non-darted side (P < 0.05).

Drug concentrations and pharmacokinetics

It is important to acknowledge that a limitation of this study is the lack of direct comparisons to conventionally injected antimicrobial drugs. Cattle availability and financial constraints prevented us from including additional animals in our study, and future studies could provide controlled comparisons of drug disposition after different routes or methods of administration. The following discussion reviews comparisons to previously published estimates of the pharmacokinetics of the three drugs after conventional subcutaneous injections. These comparisons are acknowledged to be subjective in nature, and pharmacokinetic parameters can be variable across studies and across individual animals.

Carcass quality and tenderness

We are not aware of previous reports on the effects of remote delivery devices on tenderness at harvest, although a recent study of pneumatic darts similar to those used in the present study evaluated muscle damage by quantification of serum creatine kinase up to 6 days after injection of tulathromycin via dart (Coetzee et al., 2018). In that study, no difference was detected in creatine kinase concentrations between dart-injection and hand-injection of tulathromycin, but sham-injected animals were not included as a comparison in the study. In other studies, gross lesions have been reported at injection sites after intramuscular injection, including two reports of remote delivery devices that differ from the one used in the present study. Investigators in Canada used a crossbow to deliver oxytetracycline and tilmicosin to the neck or round (exact location not specified) region in cattle 30 days from slaughter, and at slaughter they noted lesions in darted and hand-injected cattle, including granulation, neovascularization, and abscesses (Van Donkersgoed et al., 1999). In the other darting study, the drug was delivered in patented devices called biobullets, which were designed to dissolve once delivered into the muscle (Edwards & Stokka, 1993). In that study, some of the biobullets did not enter the skin properly, and the ones that did enter the skin caused significant tissue damage when the animals were necropsied 2 h after injection. In the present study, we did not evaluate tissue damage in the early days after injection except during gross examination of the pilot study animals, so it is unknown if muscle was damaged at dart delivery during the study. However, the only statistically significant differences at slaughter in tenderness as evaluated via the Warner–Bratzler method (see Table 4) were in bottom round steaks from tildipirosin-and tulathromycin-darted animals as compared to saline-and ceftiofur-darted animals. Consumers can detect differences of 0.5 kg in Warner–Bratzler tenderness measures (Miller et al., 2001), so the statistically significant differences noted between tildipirosin-darted and non-darted steaks (Table 4) are potentially detectable by the consumer. The differences between ceftiofur-and saline-treated as compared to tildipirosin-or tulathromycin-darted animals (see Table 4) may not be detectable by the consumer even though statistically significantly different.