A radioimmunoassay test for tetracyclines (Charm II) was compared with high-pressure liquid chromatography (HPLC) for detection of oxytetracycline (OTC) residues in milk samples from individual lactating cows. Oxytetracycline was administered by 1 of 3 routes (IV, IM, or intrauterine) to 21 lactating dairy cows. A total of 292 duplicate milk samples were collected from milkings before and through 156 hours after OTC administration. Concentration of OTC in these samples was determined by use of the Charm II test and an HPLC method with a lower limit of quantitation, approximately 2 ng of OTC/ml. Samples were also classified with respect to presence of OTC residues relative to the FDA safe concentration (less than or equal to 30 ng/ml), using the Charm II (by control point determination) and HPLC methods. There was a significant (P less than or equal to 0.05) difference between test methods in classification of milk samples with respect to presence or absence of OTC at the FDA safe concentration. A total of 48 of the 292 test results (16.4%) did not agree. Using the HPLC test results as the standard with which Charm II test results were compared, 47 false presumptive-violative test results and 1 false presumptive-nonviolative Charm II test result (a sample containing 31 ng of OTC/ml, as evaluated by HPLC) were obtained. The samples with false presumptive-violative Charm II results contained (less than or equal to 30 ng of OTC/ml, as evaluated by HPLC. In some respects, the Charm II test performed appropriately as a screening test to detect OTC residues in milk samples from individual cows. However, the tendency for the test to yield presumptive-violative test results at OTC concentrations lower than the FDA safe concentration (as evaluated by HPLC), suggests that caution should be exercised in using the test as the sole basis on which a decision is made to reject milk.

Eight Holstein cows, 4 inoculated intracisternally in 1 quarter of the mammary gland with Escherichia coli and 4 noninfected controls, were administered ceftiofur sodium (3 mg/kg of body weight, IV, q 12 hours) for 24 hours, beginning at 14 hours after inoculation of infected cows. All challenge-exposed cows became infected, with mean +/-SEM peak log10 bacterial concentration in milk of 5.03 +/-0.69 colony-forming units/ml. The infection resulted in systemic signs (mean peak rectal temperature, 41.5 +/- 0.3 C; anorexia; signs of depression) and local inflammation (mean peak albumin concentration in milk, 7.89 +/- 1.71 mg/ml). Ceftiofur was detectable in milk from all challenge-exposed cows, compared with only 1 of 4 noninfected cows, and the mean period after inoculation that ceftiofur was detectable in milk was longer (P < 0.05) in infected (147.7 +/- 27.5 hours) than noninfected cows (1.3 +/- 1.3 hours). However, maximal ceftiofur concentration attained in milk for all cows was 0.28 micrograms/ml, and was 0.20 micrograms/ml or less for all but 2 milk samples collected for 10 days after challenge exposure. Mean serum concentration of ceftiofur peaked at 1.0 +/- 0.3 micrograms/ml and 0.7 +/- 0.1 micrograms/ml for infected and noninfected COWS, respectively. After each ceftiofur dose, mean peak and trough concentrations of ceftiofur in serum did not differ between groups; however, concentration of ceftiofur in serum was higher at 7 hours after each dose in noninfected cows, suggesting more rapid clearance of the drug in infected cows. Ceftiofur was not detected in serum (< 0.05 micrograms/ml) of any cow at or after 120 hours following inoculation of infected cows Storage of serum samples at -20 C for 3 weeks resulted in a 98.8% decrease in ceftiofur activity, compared with that in fresh serum samples. Eighty-seven percent of this loss occurred 30 minutes after mixing serum and ceftiofur; thus, about 13% of the original activity was lost in storage.