Materials and Methods

Animals—Twenty-six healthy preconditioned Holstein steers were purchased from the Kansas State University Dairy Teaching and Research Center in Manhattan, Kan. Steers were (mean ± SD) 172.2 ± 27.5 days old at the time of the study. All steers were screened for antibodies against A marginale by use of a commercially available cELISA^4,29,30,^b and for the detection of 16S rRNA of A marginale by use of an A marginale—specific RT-PCR assay³¹ Negative disease status was assessed through interpretation of results of the cELISA and RT-PCR assay in series.³² The study was approved by the Kansas State University Institutional Animal Care and Use and Institutional Biosafety Committees.

Randomization, housing, and husbandry—Steers were assigned to 1 of 3 treatment groups: needle injection (group ND [n = 10 steers]), needle-free injection (group NF [10]), and a noninjected control group (5). Briefly, steers were ranked in descending order on the basis of body weight, assigned a random number by a random-number generator,^c and sorted by random number in descending order. Steers with the 5 largest random numbers were included in the noninjected control group, and the remaining 20 steers were assigned to the ND and NF groups on an alternating basis. Steers were individually housed in a biolevel 2 safety facility An insecticide^d was applied to each steer, in accordance with the manufacturer's recommendation, at the time of entry into the biolevel 2 facility. A total mixed ration was fed at 2.5% of body weight (on an as-fed basis); the ration for each day was divided into 2 equal portions and fed twice daily. Monensin (80 g of monensin/909 kg of the total mixed ration) was the only antimicrobial included in the ration. Water was available ad libitum. When it was necessary to restrain a steer for any procedure, a rope halter was used. Steers did not receive antimicrobials that would potentially interfere with transmission of anaplasmosis for the 30 days before the study or during the study period. Furthermore, procedures that were not included in the experimental design were not performed during the study period.

Splenectomy protocol and procedure—One steer was splenectomized and used to propagate an isolate of A marginale. A hand-assisted, laparoscopic procedure was used for the splenectomy. Briefly, the steer was sedated by IM injection of xylazine hydrochloride (0.1 mg/kg), ketamine hydrochloride (0.1 mg/kg), and butorphanol tartrate (0.05 mg/kg). The sedated steer was restrained in right lateral recumbency. A 2% solution of lidocaine hydrochloride was infused as a local anesthetic, and a 6-cm paracostal incision was made over the anesthetized area. The incision was centered over the costochondral arch and extended approximately 3 cm caudal to the 13th rib. The abdominal cavity was entered, and the surgeon inserted a hand through this incision and bluntly dissected the connective tissue between the spleen and rumen until the splenic hilus was isolated. A 1.5-cm incision was made in the left flank, and a laparoscopic stapler, which functioned as a ligating, dividing, and stapling device, was inserted into the abdominal cavity. The surgeon used the hand that was inside the abdominal cavity to guide the laparoscopic stapler around the vascular pedicle, and 2 staples were applied around the hilus. The spleen then was dissected free from the rumen and removed from the abdomen. All incisions were closed by use of a routine 3-layer closure. Postoperative analgesia was provided via IV injection of flunixin meglumine (1 mg/kg). Penicillin G procaine (10,000 U/kg, IM, q 24 h for 3 days) was administered beginning on the day of surgery. Skin sutures were removed 14 days after surgery.

Inoculation and monitoring of splenectomized steer—A blood sample that contained a tick-transmissible Virginia isolate of A marginale was collected from an infected cow by a researcher at another institution,e placed in heparin anticoagulant, and shipped on ice to our laboratory via overnight courier. This isolate was fully characterized in 1978 and was originally obtained by that researcher from the USDA Animal Parasitology Institute in Beltsville, Md.³³

Five milliliters of the heparinized blood sample was used to inoculate the splenectomized steer. The steer was inoculated IV on day 8 after splenectomy and was then monitored daily for clinical signs of anaplasmosis, including anorexia, lethargy and fever. Blood samples were collected daily and immediately used for determination of the PPE and PCV

PPE determination—Blood films were made from blood samples collected into evacuated tubes containing K₂-EDTA. An automated unit^f was used to stain blood films with modified Wright stain.^34,35 A Miller reticle^g (which has a large square with an additional square inset that is one-tenth the size of the large square) was used to determine the number of parasitized erythrocytes.³⁶ Only parasitized erythrocytes were counted in the large square; however, all erythrocytes (parasitized and nonparasitized) were counted in the smaller square. The numbers of parasitized erythrocytes and nonparasitized erythrocytes were recorded. A total of 1,000 erythrocytes were counted. The PPE was reported as a percentage and was calculated by use of the following equation, which was modified from an equation³⁶ used to measure the percentage of reticulocytes for similar conditions:

PPE = (number of parasitized erythrocytes in the large square/ [number of erythrocytes in the smaller square × 9]) × 100

Measurement of PCV—The PCV was determined by partially filling capillary tubes^h with blood samples that were collected into evacuated tubes containing K2-EDTA. Tubes were centrifuged at 12,600 × g for 10 minutes.³⁰ The PCV was determined by measuringⁱ the height of the RBC portion and comparing it with the total height for the sample.

Experimental procedures—Transmission experiments were initiated when the splenectomized steer achieved a PPE of 2.0%. The steer was medicated by IV injection of flunixin meglumine (1 mg/kg). Fifteen minutes later, a halter was used to restrain the splenectomized parasitized steer with its head extended and tied to 1 side to provide access to the left side of its neck. Ten naïve steers (group ND) were each individually and serially restrained in a similar manner adjacent to the splenectomized parasitized steer. A multiple-dose syringe^j was primed with a 50-mL aliquot acquired from a 1-L bag of saline (0.9% NaCl) solution.^k The syringe was fitted with a single-use, 1.7 × 25-mm hypodermic needlel and adjusted to deliver a 2-mL injection. A sham vaccination with 2 mL of sterile saline solution was administered IM into the cervical muscles on the left side of the neck of the splenectomized parasitized steer. The same needle and syringe were then used within 60 seconds to administer a sham vaccination into the muscles on the left side of the neck of a naïve steer from group ND. This 2-step procedure was repeated until all 10 naive steers in group ND were sham vaccinated via needle injection.

The head of the splenectomized parasitized steer then was moved to the other side to provide access to the right side of its neck. Ten naïve steers (group NF) were each individually and serially restrained in a similar manner adjacent to the splenectomized steer. A needle-free injection system^m was primed in accordance with the manufacturer's recommendations with solution from the same 1-L bag of sterile saline solution.^k The pneumatic pressure was adjusted to 55,160 Pa for IM delivery of 2 mL of saline solution to cattle weighing < 227.3 kg. The tip of the needle-free injection apparatus was placed against the skin on the right side of the neck of the splenectomized parasitized steer. The apparatus was agitated in a circular motion to ensure there was no hair between the apparatus tip and the skin. Manual, downward pressure was used to engage and ready the apparatus for injection, and the injection trigger was depressed to release the saline solution. This process was immediately repeated on the neck of a steer in group NE This multiple-stage procedure was repeated until all 10 naive steers in group NF were sham vaccinated by use of the needle-free injection system.

None of the injection sites in either group was swabbed with disinfectant, and none of the needles or the needle-free apparatus tip was changed or disinfected during either injection procedure. Different persons performed all injections for the steers in groups ND and NE Day of sham vaccination was designated as day 0.

Five control steers served as sentinel animals; furthermore, these steers were not sham vaccinated via either method (noninjected control group).

Steers were monitored daily for signs of illness for 61 days. Disease status was evaluated twice weekly (days 0, 2, 6, 9, 13, 16, 20, 23, 27, 30, 34, 37, 41, 44, 48, 51, 54, 57, and 61). On those days, 2 blood samples were collected from each steer (a 3-mL sample into evacuated tubes containing K₂-EDTA and a 5-mL sample into evacuated tubes that contained no additive [ie, serum tubes]). Samples in the serum tubes were centrifuged at 5,000 × g for 5 minutes at 4°C, and serum was harvested and stored at −80°C until subsequent analysis. Serum samples were evaluated by use of a cELISA.

cELISA—A commercial cELISA^b was used to evaluate serum samples; the cELISA was used in accordance with the method described by the OIE⁴ and recommended by the manufacturer.^b The optical density of each well was measured by use of an ELISA plate reader at a wavelength of 620 nm. The percentage inhibition of each sample was calculated by use of the following equation: percentage inhibition = 100 − (sample optical density × 100)/mean optical density of a negative control sample. Samples with a percentage inhibition < 30% were recorded as having negative results, whereas samples with a percentage inhibition ≤ 30% were recorded as having positive results.^4,37,b

RNA extraction and RT-PCR assay—Extraction of RNA from blood samples collected into evacuated tubes containing K₂-EDTA was performed by use of a commercially available product1ⁿ; extraction was performed in accordance with the manufacturer's recommendations. Briefly, plasma was separated by centrifuging blood samples at 2,750 × g for 5 minutes at 4°C. Plasma was removed with a single-use pipette. An aliquot (200 μL) of plasma-free blood was transferred to a microcentrifuge tube. One milliliter of a monophase solution1ⁿ of guanidine thiocyanate and phenol was added to lyse the RBCs. The microcentrifuge tubes were vigorously vortexed and then allowed to sit undisturbed for 10 minutes. Chloroform (200 μL) was added to each tube, and tubes were vigorously vortexed for 15 seconds. The solution was then allowed to sit undisturbed for 10 minutes, after which it was centrifuged at 12,000 × g for 15 minutes at 4°C. The colorless upper aqueous phase of the solution was transferred to a new tube; 500 μL of 2-propanol was added to each tube, and the contents were vortexed briefly. The solution then was allowed to sit undisturbed for 10 minutes, after which it was centrifuged at 12,000 × g for 10 minutes at 4°C. Supernatant was discarded, and each pellet was washed with 1 mL of 75% ethanol; the samples were briefly vortexed, after which they were centrifuged at 12,000 × g for 5 minutes at 4°C. Supernatant was discarded, and the remaining RNA pellet in each tube was allowed to air dry. Nuclease-free water (50 μL) was used to resuspend each RNA pellet. Samples were stored at −80°C until analysis by use of an RT-PCR assay.

A real-time quantitative RT-PCR assay³¹ was used for the identification of 16S rRNA of A marginale by use of previously designed³⁸ forward and reverse primers and a Taq polymerase probe designed as part of the present study. The forward and reverse primer sequences^o were 5′-CTCAGAACGAACGCTGG-3′ and 5′-CATTTCTAGTGGCTATCCC-3′, respectively. The A marginale probe sequence^o was 5′-/56-FAM/CGCAGCTTGCTGCGTGTATGGT/3BHQ_1/−3′. A commercially available, 23-μL RT-PCR assay mixture^p that included 10 pmol of each of the forward and reverse primers, 5 pmol of each deoxynucleotide triphosphate, 187.5 nmol of MgSO₄, 11.3 pmol of A marginale probe, and 8 U of ribonuclease inhibitor^q was placed in a tube, and 2 μL of template was added to achieve a final volume of 25 μL. Temperature cycles used for the RT-PCR assay were an initial complementary DNA generation cycle at 48°C for 30 minutes, then 3 minutes at 94°C, followed by 45 cycles of 94°C for 15 seconds, 50°C for 30 seconds, and 60°C for 60 seconds.

An in vitro transcript of A marginale plasmid DNA was prepared for use in development of the RT-PCR assay. Plasmid DNA (3 μg) was digested with the restriction enzyme SpeI. The SpeI-digested plasmid DNA was used to generate recombinant transcripts with a T7 RNA polymerase, as described in the instructions for the kit.^r Ten-fold serial dilutions (ranging from 1 billion to 1 molecule) of the in vitro transcript were analyzed in triplicate to optimize the assay in a commercially available real-time quantitative PCR system.^s Real-time formation of the RT-PCR product was monitored by measuring the emitted fluorescence in the extension phase of the PCR cycles within the real-time PCR system. A reaction was assessed as positive for the template when it detected 7 fluorescent units for the emission channel of the fluorescent probe. The PCR cycle at which fluorescence was evident, which is dependent on the concentration of the template in the reaction, was regarded as the Ct value.

Linear regression was used to quantify the number of 16S rRNA template molecules in each 25-μL reaction on the basis of the corresponding Ct value by use of the equation y = −3.4324x + 40.38, where y is the reported Ct value and x is the number of template molecules. The R² for the regression equation was 0.9973. Efficiency of the RT-PCR assay was calculated to be 95.6%.^30,39 Samples from a cow known to be a carrier of A marginale and a naïve cow were extracted and analyzed simultaneously to monitor assay performance and quality of the rRNA extraction technique.

Statistical analysis—Data were entered into a software package^c for subsequent calculations and manipulation. Geometric mean and CV were calculated for data acquired from recorded diagnostic test results. Diagnostic test results were also converted to a binary format (0 = negative and 1 = positive). Sensitivity and specificity with 95% CIs were calculated for light microscopy, cELISA, and RT-PCR assay for each sample (n = 25) at each time point.

Agreement between diagnostic results for each twice-weekly sample was assessed by calculating a κ statistic.⁴⁰ Results were compared by use of a software program' in a 2 × 2 contingency table to calculate κ via the following equations:

View Expanded

where EP is the expected proportion of equal outcomes according to chance; a is the number of true-positive results, b is the number of false-positive results, n is the total number of samples, c is the number of false-negative results, d is the number of true-negative results, ([a + d]/n) − EP is the observed proportion of equal outcomes beyond chance, and 1 − EP is the maximal proportion of agreement not attributable to chance. The κ statistic measures the agreement between tests on a scale from 0 to 1, with κ between 0 and < 0.4 indicating poor agreement, κ between 0.4 and ≤ 0.75 indicating good agreement, and κ > 0.75 indicating excellent agreement. When κ could not be determined because of a lack of concordant results in ≥ 2 of the 2 × 2 cells, an overall proportion of agreement was calculated by dividing the sum of concordant test results by the number of samples tested.⁴⁰

The association between disease outcome and results for the 3 diagnostic testing regimens was analyzed by use of generalized linear mixed models^u and generalized estimating equations.^v Methods were used to account for the lack of independence among repeated observations for the same animal over time. A semi-parametric survival analysis^w was performed to take into account the time point at which each respective diagnostic test first detected A marginale. A Kaplan-Meier survival analysis^w (nonparametric) was performed on the raw data to depict the amount of time elapsed after sham vaccination before a positive outcome for each of the 3 testing regimens. Values of P < 0.05 were used to determine significant differences.

Results

Prior to enrollment in the study, all steers were confirmed to have negative results for anaplasmosis when tested by use of the cELISA and an A marginale-specific RT-PCR assay. No adverse reaction attributable to surgery was detected in the splenectomized steer. The splenectomized steer was inoculated with 5 ml of a heparinized blood sample that had a PPE of 8.1% and PCV of 30%; however, an rRNA molecule count was not determined for the inoculum because of the heparin in that blood sample.

Time elapsed from inoculation of the splenectomized steer until the development of a PPE equal to 2% was 34 days. The PPE, PCV, and number of 16S rRNA molecules were 2%, 23%, and 8.9 × 10⁹ molecules, respectively, at that time. No adverse reactions were observed as a result of the injection methods used. In addition, the authors are not aware of any indications to suggest the use of flunixin meglumine would predispose cattle to clotting disorders, and no such complications were detected after administration of flunixin meglumine to the splenectomized parasitized steer. The splenectomized parasitized steer was euthanatized 45 days after inoculation because of anorexia, lethargy, and pyrexia; however, none of the other steers were removed from the study.

At the end of the study, 6 of 10 steers in the ND group had positive results for the anaplasmosis pathogen, A marginale, as determined on the basis of results of the cELISA, light microscopy, and RT-PCR assay. This represented the first, second, fourth, sixth, seventh, and tenth steers that were sham vaccinated via needle injection. All steers in the NF and control groups had negative results for anaplasmosis throughout the study. Thus, the number of steers infected with A marginale differed significantly between the ND group and the NF and control groups.

The mean ± SD predicted model-adjusted probabilities of becoming infected, comparing the ND group with the NF and control groups, were 0.60 ± 0.16 and 0.65 ± 0.18, respectively. Because all cattle in the NF and control groups had negative results for A marginale, the model-adjusted probability of a positive test result for A marginale was zero for each of those groups. A steer in the ND group was significantly more likely to become infected with A marginale (odds ratio, 44.6; 95% CI, 19.5 to 101.8) than was a steer in the NF or control groups.

In 25-μL reactions with template from the 10-fold serial dilutions, the standard curve of the RT-PCR assay ranged from 100 to 1 billion molecules (Figure 1). The RT-PCR assay identified 120 molecules of 16S rRNA (Table 1) in 1 steer as early as 9 days after sham vaccination (Figure 2). Sensitivity of the RT-PCR assay was consistent in identifying steers infected with A marginale in successive samples. The peak sensitivity (100%) for the RT-PCR assay was at day 20, and it was sustained through the end of the study. The peak number of 16S rRNA molecules recovered from 250 μL of plasma-free blood was 1.6 × 10⁸ molecules at day 41. There were no positive results for the NF and control groups by use of the RT-PCR assay.

Sensitivity and linearity with RNA concentrations for an RT-PCR assay (A) and fluorescent emission from serial dilution templates (B). Serial 10-fold dilutions of an in vitro transcript made from Anaplasma marginale plasmid DNA were used. The mean Ct values from 3 separate experiments were plotted against the log₁₀ number of rRNA molecules (log₁₀ 2 and log₁₀ 9 correspond to 100 and 1 billion molecules of 16S rRNA, respectively). The equation of the line is y = −3.4324x + 40.38 (R² = 0.9973).

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Sensitivity and linearity with RNA concentrations for an RT-PCR assay (A) and fluorescent emission from serial dilution templates (B). Serial 10-fold dilutions of an in vitro transcript made from Anaplasma marginale plasmid DNA were used. The mean Ct values from 3 separate experiments were plotted against the log₁₀ number of rRNA molecules (log₁₀ 2 and log₁₀ 9 correspond to 100 and 1 billion molecules of 16S rRNA, respectively). The equation of the line is y = −3.4324x + 40.38 (R² = 0.9973).

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

Sensitivity and linearity with RNA concentrations for an RT-PCR assay (A) and fluorescent emission from serial dilution templates (B). Serial 10-fold dilutions of an in vitro transcript made from Anaplasma marginale plasmid DNA were used. The mean Ct values from 3 separate experiments were plotted against the log₁₀ number of rRNA molecules (log₁₀ 2 and log₁₀ 9 correspond to 100 and 1 billion molecules of 16S rRNA, respectively). The equation of the line is y = −3.4324x + 40.38 (R² = 0.9973).

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

Sensitivity for the RT-PCR assay (white squares), cELISA (white diamonds), and light microscopy (white triangles) for detection of anaplasmosis in samples obtained from 6 steers that became infected with a Virginia isolate of A marginale after sham vaccination administered by IM injection with a needle. Day of sham vaccination was designated as day 0.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Sensitivity for the RT-PCR assay (white squares), cELISA (white diamonds), and light microscopy (white triangles) for detection of anaplasmosis in samples obtained from 6 steers that became infected with a Virginia isolate of A marginale after sham vaccination administered by IM injection with a needle. Day of sham vaccination was designated as day 0.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

Sensitivity for the RT-PCR assay (white squares), cELISA (white diamonds), and light microscopy (white triangles) for detection of anaplasmosis in samples obtained from 6 steers that became infected with a Virginia isolate of A marginale after sham vaccination administered by IM injection with a needle. Day of sham vaccination was designated as day 0.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

The cELISA yielded positive results for 1 steer at 13 days after sham vaccination (Figure 2); however, sensitivity of the cELISA was inconsistent for the 4 successive samples from that steer. No infected steer was identified by use of the cELISA at day 16. A steer identified as infected with A marginale at day 20 was not the same steer identified as infected at day 13. The cELISA did not have consistent sensitivity until day 34. Peak sensitivity (100%) for the cELISA was at day 41, and it was sustained through the remainder of the study. This coincided with the peak number of 16S rRNA molecules identified by use of the RT-PCR assay and a decreasing PCV (Table 1).

The PPE determination (use of light microscopy to examine stained blood films) detected 1 infected steer as early as 16 days after sham vaccination; however, this steer did not consistently have positive results until day 34. Interestingly, the day of peak sensitivity for light microscopy (day 41) coincided with the day of peak sensitivity for the cELISA, the peak in 16S rRNA molecules identified with the RT-PCR assay, and the decrease in PCV (Table 1). Furthermore, although they differed significantly, the odds of a positive test result for A marginale by use of light microscopy were only 0.5 times as likely as the odds for a positive test result by use of the cELISA.

Variability of the performance of each diagnostic method and agreement among diagnostic methods during the peracute, acute, and chronic stages of infection were compared (Table 2). Sensitivity was inadequate among all methods during the peracute phase after sham vaccination. However, use of the RT-PCR assay, cELISA, and light microscopy identified an infected steer on days 9, 13, and 16 after sham vaccination, respectively. Peak sensitivity for the RT-PCR assay, cELISA, and light microscopy was on days 20, 41, and 41, respectively. Diagnostic sensitivity was sustained throughout the remainder of the study for the cELISA and RT-PCR assay; however, sensitivity of light microscopy decreased to 0% by day 57. The RT-PCR assay was the only diagnostic method that maintained a specificity of 100% throughout the study. Poor specificity for the cELISA was caused by 8 false-positive results for steers in the control and NF groups on days 2 (n = 1 false-positive result), 13 (2), 16 (1), 23 (3), and 30 (1). The percentage inhibition recorded for each of these false-positive results was between 30% and 40%. The odds that an A marginale-infected steer would have a positive result for the RT-PCR assay was 1.34 times as likely as the odds for a positive result for the cELISA; these odds did not differ significantly (P = 0.07).

Agreement between results of the cELISA and RT-PCR assay as well as among results of all 3 diagnostic methods was determined. However, a proportion of concordance (0.76) was calculated on days 0, 2, and 6 after sham vaccination because of a lack of concordant results in ≥ 2 cells in the 2 × 2 contingency table. Sensitivity and specificity for each of these test days were 0% and 100%, respectively. The κ was calculated for the remainder of results for days 9 through 61. There was perfect agreement on day 41 when comparisons were made for 2 and 3 tests; however, excellent agreement was only sustained through the remainder of the study when results were compared for the RT-PCR assay and cELISA. At day 61, agreement between results for the cELISA and RT-PCR assay as well as among results for all 3 tests was 1.00 (95% CI, 0.72 to 1.28) and 0.75 (95% CI, 0.53 to 0.97), respectively.

Imprecision of diagnostic test results reported for the 6 steers infected with A marginale in the ND group was calculated as the CV of the geometric mean (Table 1). The geometric CV for the RT-PCR assay at peak sensitivity (20 days after sham vaccination) was 17%. However, this estimate did not exceed 22% throughout the study at any time point after infected steers were detected. The geometric CV for light microscopy, the cELISA, and the RT-PCR assay at day 41, which corresponded to the day with the highest agreement among the 3 tests, was 68%, 43.2%, and 1.8%, respectively The lowest estimates were at days 48, 51, and 44 for light microscopy, the cELISA, and the RT-PCR assay, respectively

Results for the semiparametric survival analysis were expressed as hazard ratios. Hazard ratios were interpreted similarly as odds ratios, were assumed proportional over time, and were representative of the effect of a unit change in the predictor on the frequency of the outcome.⁴⁰ The hazard ratio for cattle with a positive test result for A marginale in the ND group by use of RT-PCR assay, compared with that for a positive test result by use of the cELISA and light microscopy, was 1.15 and 1.58, respectively Similarly, the hazard ratio for cattle with a positive test result for A marginale in the ND group by use of the cELISA, compared with that for a positive test result by use of the RT-PCR assay and light microscopy, was 0.74 and 1.86, respectively Conversely, light microscopy was the least likely to yield positive results during the study period; the hazard ratio for cattle with a positive test result for A marginale by use of light microscopy, compared with that for a positive test result by use of the RT-PCR assay and the cELISA, was 0.74 and 0.81, respectively

Kaplan-Meier survival analysis was performed on raw data to compare the 3 diagnostic regimens (Figure 3). At day 0, the risk of infection for all steers exposed was l. However, this risk decreased over time as A marginale-infected steers were identified.

Kaplan-Meier survival estimate derived from the results for the RT-PCR assay (black line), cELISA (dashed line), and light microscopy (gray line) for detection of anaplasmosis in samples obtained from 6 steers that became infected with a Virginia isolate of A marginale after sham vaccination administered by IM injection with a needle. Day of sham vaccination was designated as day 0.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Kaplan-Meier survival estimate derived from the results for the RT-PCR assay (black line), cELISA (dashed line), and light microscopy (gray line) for detection of anaplasmosis in samples obtained from 6 steers that became infected with a Virginia isolate of A marginale after sham vaccination administered by IM injection with a needle. Day of sham vaccination was designated as day 0.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

Kaplan-Meier survival estimate derived from the results for the RT-PCR assay (black line), cELISA (dashed line), and light microscopy (gray line) for detection of anaplasmosis in samples obtained from 6 steers that became infected with a Virginia isolate of A marginale after sham vaccination administered by IM injection with a needle. Day of sham vaccination was designated as day 0.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1178

• Download Figure
• Download figure as PowerPoint slide

Additionally, the semiparametric survival analysis was used to test for significant differences in the injection sequence among steers within the ND group. Once the needle was contaminated by injection of the splenectomized parasitized steer, the steers that subsequently had a positive test result for A marginale were the first, second, fourth, sixth, seventh, and tenth animals in the ND group. The hazard ratio for the sequence of injection (0.96) indicated that the risk of becoming infected with A marginale (as determined on the basis of a positive test result) was the same for all steers in the ND group. Therefore, the sequence of injection was not significantly associated with a positive test result for A marginale.

To further validate results for the diagnostic methods, the 15 naïve steers from the NF and control groups were each inoculated IV with 5 ml of blood obtained from 1 of the 6 ND steers iatrogenically infected with the Virginia isolate of A marginale. All 15 steers became infected with A marginale as a result of IV inoculation, as determined by interpretation of results of the cELISA, light microscopy, and RT-PCR assay in series. The 4 remaining steers in group ND were not challenge exposed because they were enrolled in a subsequent study.

Discussion

Anaplasmosis poses many problems to the cattle industry because of complications with disease control, eradication, and treatment. When cattle of unknown disease status are being vaccinated, it is highly recommended to use hygienic techniques. Although repetitive use of needles to inject cattle is not recommended, this practice still is in use. It may be argued that a cattle population with a stable prevalence of endemic disease may be advantageous as it results in minimizing clinical disease in adult cattle. However, this approach does not permit commingling of cattle of unknown disease status. Another disadvantage would be the transmission of other blood-borne pathogens of cattle. If the prevalence of anaplasmosis were allowed to increase, culling practices may be amplified and trade restrictions intensified between endemic and nonendemic countries because of the current use of unreliable diagnostic methods.

The study reported here was designed to evaluate the use of a needle-free injection system for the control of anaplasmosis transmission among cattle during vaccination. Needle-free injection has been validated as a tool for use in controlling horizontal transmission of A marginale. Additionally, a novel RT-PCR assay was developed and evaluated for use in detecting A marginale in bovine blood samples. This data set is clinically relevant because of the potential spread of A marginale to naïve cattle during the performance of routine animal husbandry practices as well as for identifying deficiencies in the sensitivity and specificity of currently available diagnostic methods. To our knowledge, this is the first report in which the use of needle-free injection techniques for the control of iatrogenic transmission of anaplasmosis as well as the performance of first-, second-, and third-generation diagnostic methods at sequential time points after a single exposure of cattle to A marginale have been evaluated.

Transmission of anaplasmosis from a known carrier to susceptible cattle via needles has been reported.¹⁵ In that report, only 1 of 5 cattle became infected. Furthermore, the authors of that report described visual detection of blood contamination on the needle between injections. In the study reported here, 6 of 10 steers became infected. Blood contamination on the needle was visible only prior to injection of the last steer in the ND group. Otherwise, the needle appeared to be safe (ie, no blood contamination) for injecting multiple cattle. Because of the random pattern of transmission and lack of significance associated with the injection sequence, the repetitive use of a needle among cattle of unknown disease status should be regarded as unacceptable. Furthermore, the fact that 6 of 10 steers exposed to an A marginale-contaminated needle became infected after an IM injection leads to the hypothesis that this route of infection may be extremely common in current production systems.

Transmission of blood components during needle-free injection techniques has been reported.²⁸ This apparently was negligible for the transmission of anaplasmosis for the conditions of this study reported here. Sham-vaccination of the splenectomized parasitzed steer prior to sham vaccination of each steer in the NF group robustly challenged the potential for iatrogenic transmission of a Virginia isolate of A marginale via needle-free injection. However, transmission via needle-free injection techniques is not known for situations in which anaplasmosis carrier cattle may have a PPE > 2%. Similar or worse conditions are reasonably unlikely to be encountered in field settings. However, it may be necessary to account for the temporal association of previous vaccinations, disease prevalence, and timing of vaccination in regard to the seasonal distribution of clinically affected cattle when considering the use of needle-free injection techniques. Furthermore, it should be mentioned that disease resistance among breeds has not been verified.^41–43

The use of needle-free injection techniques in production settings will aid in the reduction of biohazard waste, alleviate operator injury resulting from accidental needle punctures, and eliminate the possibility of needle contamination attributable to vaccination in consumable meat products while enabling producers to maintain rates for processing of cattle that are necessary for minimizing handling and stress in the animals. Even though use of needle-free injection techniques is superior to use of needles for preventing iatrogenic transmission of A marginale, it is recommended that care be used to avoid unwarranted inoculation of cattle that might result from improper removal of previously used vaccine products from the injection system because of poor cleaning techniques.²⁴ Further studies are necessary to fully evaluate the use of needle-free injection techniques for the control of other blood-borne diseases of cattle.

At the end of the study reported here, the prevalence of anaplasmosis was 24% (6/25 steers). It may be argued that the evaluation of sensitivity and specificity among the diagnostic methods included in this study should be interpreted with caution because of the small study population and low number of infected steers; however, the accurate and precise diagnosis of anaplasmosis has historically been problematic because of diagnostic methods that lack adequate sensitivity and specificity^{30,37,44–46} Therefore, it is important to describe the inequality of these diagnostic methods.

First-generation diagnostic methods rely on the growth or visual identification of the organism of interest. These methods have limited sensitivity and lack adequate specificity to differentiate between morphologically similar pathogens, normal intracellular structures, and stain artifacts. During the present study, light microscopy was proven to be unreliable because of false-negative results. This was attributable to the low number of circulating rickettsial organisms encountered. Even though all steers infected with A marginale were accurately classified at some time during the study, light microscopy had a diagnostic sensitivity of 100% only at day 41.

Second-generation methods, which rely on the identification of cell components, metabolic products, and detection of antigenic components, are currently the most commonly used techniques for disease classification in clinical medicine and research. One disadvantage of these methods is the potential for cross-reactivity among coexisting diseases. Because of similarity among MSP5 surface proteins, cross-reactivity among A marginale, A centrale, and Anaplasma phagocytophilum for the cELISA has been reported.^37,46 The cELISA has had better and more sustainable sensitivity, compared with that of light microscopy. However, the cELISA did not have a sensitivity of 100% until day 41 for the ND group. A disadvantage for this technique is the cutoff value used to classify disease status. A cutoff value of ≥ 30% inhibition was used in the present study to classify a steer as infected with anaplasmosis. This value led to multiple false-positive results for the NF and control groups. However, the sensitivity of this assay at the earlier time points would have been compromised if a cutoff value of 40% inhibition had been used.

Third-generation methods use nucleic acid-based techniques for the classification of disease status. These methods offer superior sensitivity and specificity over first- and second-generation methods. This study is not the first in which investigators used a nucleic acid-based technique for the diagnosis of A marginale in bovine blood samples.^14,29,47,54 However, it is the first study in which a real-time quantitative RT-PCR method was used to identify 16S rRNA of A marginale in infected cattle. The advantages of this assay are the enhanced sensitivity for identifying rRNA targets that are present in higher quantities than a single copy of DNA per organism, the ability to quantify the genetic template, and elimination of the need for PAGE. In addition, this assay could serve as a substitute for the inoculation of splenectomized cattle with blood from cattle of unknown disease status. The major disadvantage is the cost of the reagents and equipment as well as the need for equipment that may not be readily accessible.

The RT-PCR assay had a sensitivity of 100% by day 20. This was a noticeable improvement over the use of light microscopy and cELISA for diagnosis in the prepatent period; however, false-negative assay results were attributable to the inability of the RT-PCR assay to detect disease in 250 μL of plasma-free blood obtained prior to day 20. Because of the performance of the diagnostic methods used during this study, the authors recommend repeated collection of blood samples from cattle of unknown disease status at 3- and 6-week intervals when the RT-PCR assay and cELISA, respectively, are used. However, light microscopy would not be recommended for determining disease in cattle with unknown disease status.

Anaplasmosis is a complex and challenging disease for stakeholders in the cattle industry, foreign policy, and research arenas. Because of the lack of substantial success with treatment strategies and problems with vaccine availability and vector control, anaplasmosis control strategies should primarily concentrate on established methods for disease prevention. This data set is clinically relevant because of the potential spread of A marginale to naïve cattle during performance of routine animal husbandry practices as well as the fact that it indicates deficiencies in the sensitivity and specificity of currently available diagnostic methods. Our results identified needle-free injection as a superior method for controlling the iatrogenic transmission of anaplasmosis. Furthermore, use of a novel A marginale-specific RT-PCR assay has the potential to impact the future of disease classification prior to local, interstate, or international movement of cattle between endemic and nonendemic countries.