All work was completed at The University of Pennsylvania School of Veterinary Medicine, New Bolton Center, Kennett Square, PA.
Effect of Sample Storage on Blood Crossmatching in Horses
Article first published online: 28 MAR 2012
Copyright © 2012 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 26, Issue 3, pages 662–667, May-June 2012
How to Cite
Harris, M., Nolen-Walston, R., Ashton, W., May, M., Jackson, K. and Boston, R. (2012), Effect of Sample Storage on Blood Crossmatching in Horses. Journal of Veterinary Internal Medicine, 26: 662–667. doi: 10.1111/j.1939-1676.2012.00913.x
- Issue published online: 2 MAY 2012
- Article first published online: 28 MAR 2012
- Manuscript Accepted: 6 FEB 2012
- Manuscript Revised: 17 JAN 2012
- Manuscript Received: 31 OCT 2011
- Arabian Horse Association
- Blood storage;
- Pretransfusion testing;
Blood samples banked for up to 1 month are typically used to perform pretransfusion testing in humans and small animals, but this has not been validated using blood from horses.
Compatibility of equine blood samples is repeatable using fresh samples, and reproducible using donor blood samples stored for up to 4 weeks.
Six healthy adult horses.
Randomized, blinded experimental study. Immunologic compatibility of the blood of all horses was assessed using a major and minor saline agglutination and hemolysin crossmatch using blood samples refrigerated for 0–4 weeks and fresh blood from the same horses. Crossmatch results were scored and then compared to identify changes of compatibility in each of the 4 tests. In addition, repeatability of the crossmatch technique itself was assessed by performing 6 iterations of this procedure in immediate succession with fresh blood from 3 horses.
No significant difference in crossmatch results was found using fresh blood (P = .39–1.00). Reproducibility was poor using blood stored for 1–4 weeks, especially in tests using stored erythrocytes (major antigen crossmatches), with significant differences from baseline at all weeks (P < .05); 13 of these differences were positive, indicating poorer compatibility.
Conclusions and Clinical Importance
Equine blood crossmatching is repeatable using fresh blood, although decreased apparent compatibility after storage makes exclusion of compatible donors more likely than mistaken administration of incompatible blood. These data suggest that fresh samples should be collected from potential donors before crossmatching equine blood.
Whole blood transfusion is an essential tool in equine critical care and surgical practice. Indications for transfusion include severe anemia from surgical blood loss or acute hemorrhage, neonatal isoerythrolysis, hemolysis because of toxins, drugs, or immune-mediated conditions, coagulopathies, and nonregenerative disorders. Seven blood groups are recognized in horses by The International Society for Animal Genetics, namely A, C, D, K, P, Q, and U, along with over 30 factors, which are antigenic sites on red blood cell surface, distributed within these groups. Therefore, over 400,000 phenotypic variations are possible. Any allogenic transfusion has the potential to expose the recipient to foreign antigens and incite a hypersensitivity reaction, and therefore blood typing and crossmatching should ideally be performed before transfusion. However, for equids, no published data exist that definitively correlate in vitro compatibility of blood samples using any standard crossmatching technique with a certainty or increased risk of such an outcome.
Unlike dogs and cats where a point-of-care card1 and gel column agglutination test2 are available for typing, only a few diagnostic laboratories perform equine blood typing.[1, 2] In addition, typing will not prevent transfusion reactions attributable to poorly characterized red cell antigens, leukocytes, platelets, or proteins.[3, 4] Initial transfusions, however, are rarely associated with complications as strongly reactive alloantibodies are infrequently encountered in the equine population. Notable exceptions include exposure via pregnancy and parturition, as well as previous blood transfusion.
Crossmatch methodology has undergone multiple changes since it was first used in human medicine in 1907. Although a gel agglutination test is available for dogs and cats, the saline agglutination crossmatch is by far the most common test used in horses. Like the latter test, there is a lack of evidence that other equine crossmatch methods, such as the addition of the Coombs antiglobulin reagent, reliably predict immunologic responses to allogenic transfusions in vivo. Eliminating crossmatch procedures altogether has been suggested in favor of transfusion of a small amount of donor blood and monitoring for signs of a transfusion reaction. In the saline agglutination and hemolysin crossmatch procedure, agglutination indicates incompatibility, as does hemolysis after the addition of exo-genous complement. Antibodies in equine blood that cause hemolysis require a source of complement to activate or enhance the classical pathway of complement activation. The saline agglutination test is best performed in a laboratory by an experienced technician as each component of the test is subjectively evaluated, especially the agglutination portion which is graded on a scale indicating severity.
Traditionally in equine practice, blood is freshly collected from potential donors for crossmatching with each recipient. This process is time-consuming especially if the potential donor is not on-site, increases the workload of the clinic staff, and can delay possibly lifesaving treatment in an emergency situation. Storage of human, canine, and feline blood in blood banks is well-established, but storage of equine whole blood is not commonly performed, although equine plasma is routinely frozen and stored. There has been minimal investigation of viability parameters of stored equine erythrocytes, and none evaluating changes in consistency of crossmatch results over time.
Our objective was to evaluate if aliquots of equine blood stored refrigerated for up to 1 month could be used for the purposes of crossmatching, thus eliminating the need to collect fresh blood from potential donors for each recipient. We chose to use anticoagulants and storage techniques that are already available at many equine hospitals, thus making the procedure easily adoptable in the common clinical setting. We hypothesized that crossmatching blood stored for up to 4 weeks with fresh blood from the same individuals would produce reproducible results. In addition, the repeatability of the saline agglutination crossmatch technique was evaluated with fresh samples. We hypothesized that this technique would produce repeatable results.
Materials and Methods
This study was performed using a randomized, blinded experimental design. Blood was obtained from 6 healthy adult horses (age range 8–23 years, mean 16 years) from a university teaching herd. Two were geldings, 4 were mares, with 1 Standardbred, 2 Thoroughbreds, 2 Thoroughbred crosses, and 1 Warmblood. Two of the mares were known to have been pregnant in the past, but none were currently pregnant or had been pregnant in the 3 months preceding blood collection. None of the horses were known to have received blood transfusions in the past and none received any medications during the study period. All procedures were performed in accordance with the University of Pennsylvania Institutional Animal Care and Use Committee approved guidelines.
Blood Collection and Storage
Blood was collected from each horse using a 20 gauge needle directly into evacuated glass tubes, both with no additive and with acid-citrate-dextrose (ACD) solution A.3 At week 0, 150 mL of blood was collected from each horse, and then another 30 mL was collected from each horse every 7 days for 4 weeks. The blood collected at week 0 was stored in a commercial refrigerator, the temperature of which was monitored daily and confirmed maintenance of temperatures between 2° and 6°C for 4 weeks of the study. The serum and plasma were separated from erythrocytes on day 0 and components were stored separately. For the repeatability component, 20 mL of blood was collected and blinded in the same fashion, and 6 identical crossmatch procedures were performed in succession.
All blood was collected by the same phlebotomist (MH) who randomly assigned a number or letter to each sample to blind the technician performing the crossmatch. All crossmatches were performed by the same technician with over 15 years of experience both in human and veterinary crossmatching (WA). At week 0, each of the 6 horses was crossmatched against the other 5 individuals. Once per week for 4 consecutive weeks, fresh blood was collected from the same 6 horses and crossmatched against stored blood samples stored since week 0 from all 6 horses, including the same individual matched against its own stored blood.
Repeatability of the crossmatch procedure was evaluated by collecting blood from 3 of the 6 horses 2 days in a row. Each sample of serum and plasma was divided into 3 aliquots and assigned random identifiers to ensure blinding during analysis. In this way, each horse was crossmatched against the others 3 times on 2 consecutive days for a total of 6 repeats of the major and minor saline agglutination and hemolysin crossmatches.
Four crossmatch procedures were performed between each horse pair: major agglutination (MaA), minor agglutination (MiA), major hemolysin (MaH), and minor hemolysin (MiH). The major and minor agglutination procedures were performed by washing all erythrocytes from the ACD tubes 4 times with physiologic (0.85%) saline in a centrifuge4 at 1,000 rpm for 1 minute per wash. After washing, cells were resuspended in saline to a final concentration of 4%. All erythrocytes used in the crossmatch procedures were obtained from ACD tubes, and all serum was obtained from no-additive tubes. In 10 × 75 mm tubes, 2 drops of “donor” cell suspension was combined with 2 drops of “recipient” serum for the major match, and 2 drops of “recipient” cell suspension was combined with 2 drops “donor” serum for the minor match. All tubes were incubated at 37°C for 15 minutes, and then placed into the centrifuge at 1,000 rpm for 1 minute before being evaluated both macroscopically and microscopically for agglutination. An auto-control was performed combining “recipient” serum and washed erythrocytes. Results were reported on a scale of 0–4 depending on the degree of agglutination. The scale indicates an increasing amount of agglutination, with 0 assigned to a sample with no macroscopic or microscopic agglutination and 4 assigned to a sample with agglutination visible to the naked eye. The procedure for the hemolysin crossmatch was the same as for the agglutination match, except 2 drops of rabbit complement5 was added to both the major and minor tubes. These samples were incubated for 90 minutes at 37°C, and then centrifuged for 1 minute. The mixture was evaluated both microscopically and macroscopically, and results are reported as positive (score = 1) if hemolysis was observed and negative (score = 0) if no hemolysis was present.
At the conclusion of the 4-week storage period, the remaining samples were submitted to a commercial laboratory for bacterial culture using standard techniques.
Data were assessed for normality using the Shapiro-Wilks test and found to have a non-normal distribution. Simple regression with bootstrapping (200 iterations per test) was thus used to assess differences (Δscore) from mean score for each of the 6 repetitions (repeatability study), and from baseline score for the crossmatch results obtained after 1, 2, 3, and 4 weeks of storage (reproducibility over time). Significance was set at P < .05, and all analysis was performed using commercially available software.6
Thirty-six donor-recipient pair combinations were evaluated using 4 crossmatch assays at baseline and weekly for 4 weeks, resulting in a total of 600 data points. These included data pairs that were matched in all 4 tests (ie, compatible) as well as pairs which were incompatible to varying degrees in some or all tests. The repeatability data were performed using 3 donor-recipient pair combinations for a total of 72 data points. In 6 repetitions of each of the 4 crossmatch tests using fresh blood, no significant difference was found in the results of any test (MaA, P = .39; MiA P = 1.00; MaH, P = .42; MiH P = .86). However, reproducibility was poor using blood stored for 1–4 weeks with significant differences from baseline in some tests in all weeks (Fig 1). The greatest discrepancies were seen with the MaA assay, where scores were significantly different from baseline at week 1 (P = .001), week 2 (P = .041), week 3 (P = .044), and week 4 (P < .001). Differences from baseline values were also noted at all weeks in the MaH score, with significances of P = .003, P ≤ .001, P = .020, and P = .010, respectively. For MiA, however, the scores were not significantly different from baseline at any time (P = .70–1.0), and for MiH, scores were different from baseline at weeks 1 (P = .027) and 2 (P = .008) but not at weeks 3 and 4 (P = .14 and .33, respectively). Of the 4 tests evaluated over 4 weeks, there was a mean difference from baseline (Δ score ≠ 0) in 15 of 16; this difference was positive in 13/15 instances, indicating a worsening in compatibility in weeks 1, 2, 3, and 4 compared with baseline (Table 1). There was no evidence that the decreased compatibility became more severe with increased storage time beyond 1 week. No bacterial growth was obtained on any of the examples cultured after 4 weeks of refrigerated storage.
|Repeatability mean score (range)||0.77 (0–1)||0 (0–0)||2.61 (0–4)||0.61 (0–1)|
|Week 0 (baseline)|
|Mean (range)||0.53 (0–2)||0 (0–0)||1.06 (0–4)||0 (0–0)|
|Mean (range)||1.13 (0–4)||0.17 (0–1)||1.13 (0–4)||0.13 (0–1)|
|Mean Δscore (±SD)||0.6 (±1.10)||0.17 (±0.38)||0.07 (±1.36)||0.13 (±0.35)|
|Mean (range)||0.96 (0–4)||0.33 (0–1)||0.96 (0–3)||0.20 (0–1)|
|Mean Δscore (±SD)||0.43 (±1.16)||0.33 (±0.48)||−0.1 (±1.37)||0.20 (±0.41)|
|Mean (range)||0.93 (0–4)||0.17 (0–1)||1.06 (0–4)||0.07 (0–1)|
|Mean Δscore (±SD)||0.4 (±1.07)||0.17 (±0.38)||0 (±1.41)||0.07 (±0.24)|
|Mean (range)||1.06 (0–2)||0.17 (0–1)||0.96 (0–4)||0.03 (0–1)|
|Mean Δscore (±SD)||0.53 (±0.68)||0.17 (±0.38)||−0.07 (±1.36)||0.03 (±0.18)|
This study evaluated the temporal stability of equine blood compatibility testing using stored “donor” blood samples against fresh “recipient” blood, as well as the intrinsic repeatability of the standard, 4 part equine crossmatch assay. Consistency in test results is usually evaluated using the metrics of repeatability and reproducibility. Repeatability is defined as the agreement of measurements taken under identical conditions where any variation can be assumed to be because of errors in the testing process itself, whereas reproducibility examines discrepancies in measurements under changing conditions, such as time or technique, where variations could be caused by biological processes. We found that the saline agglutination and hemolysin crossmatch procedure itself shows acceptable repeatability, giving consistent values across 6 iterations of the crossmatch procedure using a technician blinded to the identity of the samples. However, reproducibility of these results using stored, refrigerated “donor” blood samples matched against fresh “recipient” blood was poor even after as little as 1 week of storage. Because crossmatch assay results are repeatable, intrinsic failure of the test technique can be eliminated as a cause of the poor reproducibility after storage. Other potential explanations for the lack of consistency in crossmatch results using stored blood samples include human error, changes in expression of red cell surface antigens as a type of storage lesion, as well as alterations in structure or activity of serologic antibodies or complement during storage. Because the plasma was decanted from the red blood cell fraction before storage, the small quantity of ACD preservative remaining during refrigeration also may have affected erythrocyte stability. Although the crossmatch procedure is dependent on interpretation by the laboratory technician, the fact that our study was both blinded and had good repeatability using the same technician decreases the possibility that human error or bias is responsible for the results after storage. Bacterial contamination of the samples was also eliminated as an explanation for lack of reproducibility after storage.
There has been surprisingly little formal evaluation of the reproducibility of crossmatch results in stored specimens for both humans and veterinary species. In canine transfusion medicine, guidelines recommend maximum refrigerated storage of 28–35 days for products containing erythrocytes, whereas human samples can be stored for up to 42 days before transfusion. These storage recommendations are based on the suitability for transfusion in regard to development of red cell storage lesions. Remarkably, we were unable to find evidence that evaluation of pretransfusion compatibility in human or small animal samples have actually been shown to maintain fidelity after this duration of storage, although 1 study showed that human red cells in saline produced consistent crossmatch results after 10 days of refrigerated storage. In human medicine, stored donor samples are typed and crossmatched, either electronically or via laboratory methods, against recipient samples to establish pretransfusion compatibility. Recipient samples used for pretransfusion testing from a patient who has been pregnant or transfused within the last 3 months must be used within 3 days because of concerns about alloantibody production, whereas there is no set limit on storage time for patients outside these categories. Samples are typically kept under refrigerated conditions, although some evidence suggests that pretransfusion samples could be stored frozen at −30°C for up to 6 months. As discussed, equine blood is unusual in that a crossmatch rather than blood typing is required to most accurately predict transfusion reaction because of its vast array of groups, factors, and other antigenic stimulators, such as leukocytes and hemolysins.
In addition to storing small aliquots of donor blood to use for crossmatching purposes, in humans and small animal species, larger volumes of blood or components are often banked for a period of time for subsequent transfusion. Erythrocyte storage lesions have been investigated in veterinary species, such as dogs, cats, and horses, and include changes in erythrocyte metabolism and cell membrane.[10, 18, 19] In humans, decreased in vitro and in vivo survival of stored erythrocytes has been demonstrated. Equine erythrocytes stored for up to 4 weeks met in vitro criteria suggesting acceptability for transfusion, although, similar to other species, storage lesions have been identified. In our study, it is plausible that an erythrocyte storage lesion developed which affected the antigenic sites associated with the red cell membrane, resulting in erythrocytes which were perhaps acceptable for transfusion, but not for use in crossmatching evaluation. The major agglutination and the major hemolysin portions of the assay utilize the stored donor erythrocytes, which in this case were stored erythrocytes. Both of the major assays had the poorest reproducibility after storage (Table 1), supporting the theory that some form of red cell storage lesion is responsible for these results. Alternatively, the antibodies contained in the serum component could also have undergone alterations in structure or function that changed their immunologic interactions with erythrocyte surface antigens. The crossmatch procedure depends upon the interaction between the antigen and antibody, as well as the complement cascade in the case of the hemolysis component, and therefore any alteration in these systems would produce changes in the crossmatch itself.
We chose to use an anticoagulant and refrigeration method that we judged to be clinically applicable in a standard equine hospital or clinical laboratory setting. The anticoagulant ACD was used because this is commonly used to separate red blood cells for the crossmatch procedure and is readily available in glass tubes, although a recent study showed that citrate-phosphate-dextrose-adenine (CPDA-1) solution is likely the optimal solution for storing equine blood. The blood samples were refrigerated instead of frozen to facilitate the crossmatch procedure, as well as remove the effects of freezing on red cell integrity. Therefore, it remains unclear if our results would have been the same if a different anticoagulant or storage temperature, including freezing as employed in certain long-term storage situations in humans, was used. The crossmatch itself is a time-consuming procedure that must be performed by a skilled technician. To simplify the transfusion procedures, it would be ideal to use stored instead of fresh blood samples for crossmatching. In addition, care should be taken in interpreting these results, as this study indicated only the reproducibility of in vitro test results and not the potential in vivo effects of transfusing compatible or incompatible donor blood. Test results indicating “compatibility” may sometimes result in adverse reactions; and conversely test results indicating “incompatibility” might not always lead to adverse effects post transfusion.The decrease in in vitro compatibility after storage suggests that exclusion of potentially compatible donors may be more likely than mistaken administration of incompatible blood. These data suggest fresh samples should be collected from potential donors before crossmatching equine blood.
The authors acknowledge the staff of the clinical laboratory and the animal caretakers at New Bolton Center.
This study was supported by a grant from the Arabian Horse Association.
Rapid Vet-H, DMS Laboratories Inc, Flemington, NJ
DiaMed, Cressier sur Morat, Switzerland
ACD vacutainer blood tubes, BD, Franklin Lakes, NJ
Rabbit complement, Life Technologies, Invitrogen, Brown Deer, WI
Stata 11.0, StataCorp, College Station, TX
- 5How to approach whole blood transfusions in horses. Proc Annu Conv AAEP 2001;47:266–269., .