Corresponding author: L.N. Nielsen, Department of Small Animal Clinical Sciences, Faculty of Life Sciences, University of Copenhagen, Dyrlaegevej 16, Dk-1870 Frederiksberg C, Denmark; e-mail: email@example.com.
The role of antiphospholipid antibodies in the prolonged activated partial thromboplastin time (aPTT) previously identified in healthy Bernese Mountain Dogs remains unknown. In people, an isolated prolonged aPTT without evidence of bleeding might be because of a thrombophilic condition caused by antiphospholipid antibodies.
To examine if prolonged aPTT in healthy Bernese Mountain Dogs is because of antiphospholipid antibodies.
Twenty-two healthy Bernese Mountain Dogs and 10 healthy adult dogs of various breeds.
Prospective case control study. Healthy Bernese Moutain Dogs were examined twice over 6 months. Dogs were investigated for the presence of lupus anticoagulants and anticardiolipin (aCL) antibodies by the use of multiple aPTT tests with low and high lupus anticoagulant sensitivities, a mixing study, and an ELISA test for aCL antibody optical density to detect solid phase antiphospholipid antibodies.
In all, 15 of 22 healthy Bernese Mountain Dogs were positive for lupus anticoagulants. The Bernese Mountain Dogs had markedly higher levels of aCL antibodies compared with the control dogs (P = .006). In all, 7 of 21 of the Bernese Mountain Dogs were positive for both lupus anticoagulants and aCL antibodies, whereas 4 of 21 Bernese Mountain Dogs were negative for both.
Conclusions and Clinical Importance
Lupus anticoagulants and aCL antibodies could be the cause of prolonged aPTT in healthy Bernese Mountain Dogs. The importance of the antiphospholipid antibodies in the dogs remains unknown.
The Scientific Subcommittee of the International Standardization of Thrombosis and Hemostasis
phosphate buffered saline
Antiphospholipid syndrome (APS) is a well-characterized syndrome in people, consisting of clinical signs of thrombosis, pregnancy-related complications, and thrombocytopenia. The international classification criteria for definite APS involve clinical findings along with the laboratory evidence of presence of antiphospholipid antibodies, which among others include the detection of lupus anticoagulants and anticardiolipin (aCL) antibodies. Lupus anticoagulants are circulating blood coagulation inhibitors or antibodies directed mainly but not exclusively against phospholipid-binding plasma proteins. The anticoagulant effect of lupus anticoagulants is strictly an in vitro phenomenon identified in a functional coagulation assay. The detection of aCL antibodies is a sensitive but not specific test for APS, and is quantified by ELISA. Increased concentrations of aCL antibodies can also be detected in people with infectious disease, immune-mediated conditions, neoplastic disorders, and in healthy individuals.[4, 5] People with APS are not necessarily positive for both lupus anticoagulant and aCL antibodies, and studies have shown that patients positive for lupus anticoagulants are more likely to experience thrombotic incidents than people only positive for aCL antibodies. However, removal of aCL antibodies from the diagnostic criteria has resulted in failure to diagnose APS in a subset of patients, which lead to the inclusion of both parameters in the diagnostic criteria for APS. Precise laboratory diagnostic guidelines from the Scientific subcommittee of the International Society of Thrombosis and Hemostasis (ISTH/SCC) have been proposed to diagnose the presence of lupus anticoagulants.[8, 9]
Although the presence of aCL antibodies has been reported in both healthy and diseased dogs in a single study, it remains unknown if these dogs had normal or abnormal hemostasis in the form of clinically detectable bleeding, hemorrhagic diathesis, or laboratory evidence of prolonged coagulation tests. Only a single case report has suggested the presence of the APS in the dog. Consequently, the importance of the APS, lupus anticoagulants, and aCL antibodies in dogs remains unknown.
Previously, we have found breed specific differences in hematologic serum and biochemical and hemostatic analytes in adult healthy Bernese Mountain Dogs. The thromboelastographic (TEG) parameter MA had a moderately higher de novo generated reference interval, which in most other dogs would indicate platelet or fibrinogen associated hypercoagulability, but most strikingly, we found that activated partial thromboplastin time (aPTT) was markedly prolonged with a new interval of 6–100 seconds, whereas the remaining parameters of hemostasis (fibrinogen, d-dimers, prothrombin time, and the TEG evaluation [R, K, angle]) were within standard reference limits. In this study, none of the dogs had clinical evidence of coagulopathy. Single coagulation factor analyses performed for FVIII, FIX, FXI, FXII were within reference range for all samples, and therefore no apparent cause of the prolonged aPTT in these dogs could be detected. In people, an isolated prolonged aPTT without evidence of bleeding should be fully investigated, as it could be because of a thrombophilic condition caused by antiphospholipid antibodies rather than an indication of bleeding. The objective of this study was to investigate if these healthy Bernese Mountain Dogs with prolonged aPTT met the laboratory classification criteria for APS by (1) examining the dogs for lupus anticoagulants by following the human ISTH/SCC guidelines, and (2) identification of aCL antibodies in the dogs.
Material and Methods
The ethical committee of the Department of Small Animal Clinical Sciences, University of Copenhagen, approved this study.
Twenty-two Bernese Mountain Dogs were examined twice with 6 months in between. According to the ISTH/SCC criteria, screening and confirmatory aPTT tests were performed at both the 1st and 2nd visit. In case of prolonged screening, aPTT and normal confirmatory aPTT at both visits, mixing studies to assess for lupus anticoagulants and solid phase immunoassays (ELISA) to detect aCL antibodies, were performed on samples from the second visit. The Bernese Mountain Dogs were clinically healthy with no abnormalities detected during the examination procedures except for a prolonged aPTT-SP. The mixing study was performed for all 22 dogs; data on the aCL IgG antibody test were only missing for 1 dog.
Bernese Mountain Dogs
The Bernese Mountain Dogs for this study have previously been described in detail.
In short, healthy adult Bernese Mountain Dogs (14 female entire, 7 male entire and 1 male neutered) with a median age of 66 months (range 60–87 months) with known pedigree status were recruited for this study through an initial questionnaire and were part of a larger ongoing study. These dogs were selected based on information regarding their health status, the fact being that they did not receive any medication and there was a lack of any signs of disease.
All the dogs were examined and blood sampled at the Veterinary Teaching Hospital, Department of Small Animal Clinical Sciences, University of Copenhagen.
Determination of Health
The Bernese Mountain Dogs were determined to be clinically healthy by history; physical examination; hematology and serum biochemistry analysis, urinalysis by cystocentesis including culture; abdominal ultrasonography and thoracic radiography in 3 planes evaluated by a board-certified radiologist. None of the dogs had documented episodes of excessive bleeding.
Dogs were excluded from the study if they were pregnant, were clinically abnormal or if receiving medication at the time of examination, or if they developed disease between examination periods, or disease was present at the second examination.
Control dogs were used for the mixing study (1 dog), the aCL antibody study (10 dogs), and the total IgG ELISA test (6 dogs). The control dogs for the mixing study (Labrador Retriever), for the aCL study (Labrador Retriever, Dachshund , Crossbreeds , Border Terrier , Border Collie ), and the total IgG ELISA (Labrador Retriever, Dachshund , Border Terrier , Crossbreed ) were determined to be healthy by physical examination, normal hematology and serum biochemistry, urinalysis, and a normal coagulation profile, including aPTT, prothrombin time, d-dimer, fibrinogen and thromboelastography and no prior history of disease, medication, or bleeding tendencies. The Bernese Mountain Dogs and the control dogs were not statistically different with regard to sex and age.
Blood samples were collected from the dogs at each visit. The dogs were fasted at home for 12 hours before blood sampling. Using minimum stasis, cephalic venous blood was obtained using a 21-gauge disposable butterfly needle and a vacutainer system into first serum (4 mL) tubes and then into 3.2% citrated (4 mL) tubes.1
The citrated whole blood and serum samples were kept at room temperature and processed within 30 minutes of sampling.
Citrate stabilized samples for aPTT were centrifuged (4,440 × g for 3 minutes), and the plasma supernatant was processed (aPTT measurements) or immediately frozen at −80°C. For the mixing study, the citrated plasma was double centrifuged (once before freezing and once after thawing) and the platelet count was measured to ensure adequate platelet depletion (<10 × 109/L). Serum was acquired by centrifugation (4,440 × g for 3 minutes) and stored at −80°C for later analysis of aCL immunoglobulin G antibodies. Lipemic or hemolytic samples were not processed.
Detection of Lupus Anticoagulants
Detection of Lupus Anticoagulants
The ISTH/SCC guidelines for the detection of lupus anticoagulants were modified for veterinary use and applied accordingly (Table 1).
Table 1. International Society of Thrombosis and Hemostasis guidelines from the Scientific Subcommittee of lupus anticoagulants (ISTH/SCC) for the detection of lupus anticoagulants in humans and corresponding criteria applied in this study.
Prolongation of a phospholipid-dependent clotting assay
A screening aPTT (aPTT-SP) with high sensitivity for antiphospholipid antibodies using low concentration of phospholipids (according to manufacturer) and silica as activator
Evidence of the presence of an inhibitor demonstrated in a mixing study
Mixing performed using 1 : 1 proportion of patient and donor normal control plasma measuring aPTT-SP immediately
Confirmation of the phospholipid-dependent nature of the inhibitor
A confirmatory aPTT test (aPTT SynthAFax) with a moderate sensitivity for antiphospholipid antibodies, using high concentration of phospholipids (according to manufacturer) and ellagic acid as activator
Lack of specific inhibition of any one coagulation factor
Single factor activity to detect factor deficiency for factor VIII, IX, XI and XII
The difference between the reagents for the screening assay aPTT-SP and the confirm assay aPTT SynthAFax is in the concentration of calcium, phospholipid, and the type of activator. While the aPTT-SP reagent has a low concentration of phospholipid and calcium and is activated by silica, aPTT synthAFax reagent has a high concentration of both calcium and phospholipid and is activated by ellagic acid. The aPTT-SP test is sensitive and the aPTT synthAFax test is unaffected by the presence of antiphospholipid antibodies (ILS Laboratories, Scandinavia, Denmark, personal communication, 2010), making them applicable as screening and confirmatory assays in accordance with the ISTH SCC guidelines. The standard laboratory range for aPTT-SP (95% confidence interval) was 8.0–24.9 seconds (n = 40 healthy dogs) and for aPTT synthAFax was 9.1–12 seconds (n = 54 healthy dogs).
The use of a 3rd aPTT reagent (synthASil) with high concentration of calcium and phospholipid and activated by colloidal silica for optimal activation of the contact phase of coagulation was run in duplicates on each Bernese Mountain Dog along with the pool to screen for contact factor and prekallikrein deficiency.
The mixing study was performed using a 1 : 1 proportion of patient and healthy control dog plasma. After the mixing of the plasma, aPTT-SP analysis was run immediately. For interpretation of the mixing study, a relative correction formula was calculated and adopted from human medicine, where a less than 70% relative correction of the aPTT-SP reflected the presence of inhibitors.
where PP is patient plasma, and NCP is normal control plasma.
Detection of Anticardiolipin Antibodies
The detection of aCL IgG antibodies was performed using a direct ELISA test with the use of bovine cardiolipin, developed at the Central Laboratory at The Department of Small Animal Clinical Sciences, LIFE, KU. The test was a modification of the methods described by Papini and others () and Pierangeli and Harris (). Microtitre plates4 were coated with 50 μL bovine heart cardiolipin5 in 99% ethanol and incubated at 25°C for 45 minute. The plates were blocked with 200 μL 3% bovine serum albumin (BSA) in phosphate buffered saline (PBS) (0.15 M, pH 7.2) and incubated overnight at 4°C. After washing with PBS, 50 μL sample diluted in 1% BSA in PBS were added in duplicate to the plates. The plates were incubated for 2 hours at 25°C and washed with 1% BSA in PBS. Horseradish peroxidase conjugated rabbit anti-dog IgG5 was diluted to 1 : 20.000 in 1% BSA in PBS, and 50 μL was added to each well. After an incubation period of 60 minute at 25°C at room temperature, the plates were washed with 1% BSA in PBS, and 50 μL TMB substrate6 was added to each well. The plates were incubated in the dark at 25°C. Color development was monitored and stopped after 30 minutes with 0.2 M sulfuric acid. The optical density (OD) was read at 450 nm and corrected at 620 nm using a spectrophotometer.7 OD qualitatively measured the level of aCL antibodies. Each sample was run in duplicate and with an internal control (pooled canine samples being highly positive, medium positive and low positive for aCL IgG antibodies) to assess for inter- and intra assay variation. The highest OD of the control dogs was the cut-off for differentiation, if the test was positive for aCL IgG antibodies.
To ensure that the origin of the aCL antibody measurements were not because of nonspecific IgG binding, total IgG was measured using a commercial canine ELISA kit8 in all dogs.
Gaussian distribution was tested for the different types of aPTT and aCL antibody OD using Kolomogorov–Smirnoff tests. Descriptive statistics (mean and standard deviation) were performed for aPTT for the first and second visit. aPTT (SP and synthAFax) for the two visits were individually compared using a Student's t-test for paired observations. aCL antibody OD in Bernese Mountain Dogs and the 10 control dogs were compared using a Student's t-tests. The relationship between aCL antibody OD and total IgG for the Bernese Mountain Dogs were assessed using a Spearmann's rank correlation analysis. The proportion of sex groups was compared between the control dogs and the Bernese Mountain Dogs using chi-squared test, whereas the age of the dog groups was compared by Student's t-test. An association between lupus anticoagulants and aCL antibody test was sought using a Fischer's exact test. For all these analyses, α was set at 0.05. All statistics were performed using a commercial statistical package.9
The aPTT synthASil values were within reference range at both visits for all Bernese Mountain Dogs and control dogs (mean and standard deviation: 1st visit 11.55 ± 0.69 seconds; 2nd visit 11.56 ± 0.77 seconds). The aPTT-SP values and the aPTT synthAFax values were normally distributed for both visits, respectively. The aPTT synthAFax values remained within the standard laboratory reference interval for both visits. There was no significant difference in comparison by paired t-test of the aPTT values (aPTT-SP tests) of the 1st and 2nd visit (P = .3) and no significant difference in comparison by paired t-test of the aPTT values (aPTT synthAFax tests) between the 1st and the 2nd visit (P = .7) (Fig 1). In the mixing study, 68% of the dogs (15/22) had less than 70% relative correction of aPTT-SP, indicating the presence of an inhibitor. Because all dogs had a prolonged aPTT-SP screening test at 2 occasions 6 months apart, a normal aPTT synthAFax confirmatory test, and 15 of the dogs had less than seventy percent relative correction of aPTT-SP in the mixing study following the ISTH/SCC guidelines, it was suggestive that at least 15 of these dogs had lupus anticoagulants.
Both aCL antibody OD for Bernese Mountain Dogs and control dogs were normally distributed. One of the control dogs had unacceptable analytical variation in the duplicate determination and was excluded from the analysis. The mean aCL antibody OD was significantly higher in the Bernese Mountain Dogs compared with the control dogs (P = .006) (Fig 2). A correlation between aCL antibody OD and total IgG was not found (r = 0.30; P = .092). Thirty-three percent of the Bernese Mountain Dog group (7/21 dogs) were positive for both lupus anticoagulants and aCL antibodies, whereas 19% (4/21 dogs) were negative for both and an association between the presence of lupus anticoagulants, and aCL antibodies was not statistically significant (P = 1.00).
In this study, the aPTT synthASil tests were within reference range at both visits for all Bernese Mountain Dogs, indicating that contact factor deficiency was not the cause of the prolonged aPTT-SP. Little attention has been made to antiphospholipid antibodies in dogs, and the diversity of these antibodies and their individual significance therefore remain unknown.
Healthy people can have aCL antibodies or lupus anticoagulant and remain asymptomatic. Lupus anticoagulant is more prevalent in healthy young females. The majority of dogs in this study were females. Although thrombotic events are rare in dogs, they have been documented and include stroke, pulmonary thromboembolisms, and deep vein thrombosis.[20-22] In future studies, it may therefore be of use to test dogs with these types of diseases for the presence of antiphospholipid antibodies. Although all dogs in the study had prolonged aPTT-SP, only 68% of the dogs were identified as having lupus anticoagulants in the Bernese Mountain Dog and healthy donor mixing study. According to the human studies, a more pronounced relative correction than 70% suggests a factor deficiency, and a less pronounced relative correction suggests the presence of an inhibitor. Single factor analysis and TEG were previously performed in the dogs in this study not identifying factor dysfunction, and the aPTT synthAFax was within the reference interval for all dogs. In humans, it has previously been suggested that low concentration of lupus anticoagulants can remain undetected with a 1 : 1 portion patient and healthy donor mixing study, or slow binding of the proteins is not detected in an immediately run sample. Some studies therefore suggest a 4 : 1 patient healthy donor mixing study or a time-delayed aPTT measurement. Currently, the true lupus anticoagulant concentration of the remaining 7 dogs is therefore not clear, in particular, because these dogs had abnormal aCL IgG antibody concentration. In people, apart from using the aPTT test, the presence of lupus anticoagulants can be evaluated by performing the screening and confirmatory diluted Russell's Viper Venom tests with various concentrations of phospholipid. Presently, however, these tests have not been validated in the dog.
High levels of aCL antibodies are involved with thrombosis and recurrent pregnancy loss, and can also be present in human cancer patients.[24-28] The dogs neither had clinical signs of thrombosis, nor were they pregnant at the time of blood sampling or there were indication of increased pregnancy morbidity in this small population of dogs. Although the extensive examination protocol of the dogs did not detect tumors, occult neoplasia could not fully be excluded. Bernese Mountain Dogs are prone to disseminated histiocytic sarcoma, which is a neoplasia associated with the immune system,[29-31] and it is possible that levels of aCL antibodies exceeding the range of the control dogs without reason represent a dysfunctional immune system, which previously has been suggested in this breed. Presently, however, aCL antibodies should be evaluated in a greater number of Bernese Moutain Dogs to determine if this is a generalized breed trait or localized to a particular geographic area. The aCL antibody test was semi-quantitative. Although a quantitative test for dogs using human aCL antibodies has been described, we were not successful at validating this procedure in laboratory. As a means of standardization in this study, 3 internal controls (high levels, medium levels, and low levels of aCL antibody) were used for inter- and intraplate control procedures run using each ELISA plate, a method that has been previously described. Furthermore, total IgG was assessed to judge whether aCL test signal could be because of unspecific binding, and showed the lack of correlation between total IgG and the aCL test. There was no association between the dogs having the presence of lupus anticoagulants and high concentration of aCL antibodies. In humans, the presence of both lupus anticoagulants and aCL antibodies in a patient is not necessarily more strongly associated with the risk of thrombosis; however, there are suggestions that co-factor (β2 glycoprotein I) dependent antiphospholipid antibodies are more significant for morbidity in APS than co-factor-independent antiphospholipid antibodies. Specifically because of these conditions, when testing aCL antibodies in humans, the sample diluent in the ELISA test is considered greatly as, for example, adult bovine serum contain β2 glycoprotein I as opposed to BSA. Furthermore, within the last couple of years, it has appeared that high concentrations of antiβ2 glycoprotein I antibodies can be more seriously associated with clinical signs than aCL antibodies in patients with thrombosis.[3, 34, 35] Co-factors have only been examined once in dogs previously, and the importance of these factors was doubtful. The significance of co-factors in dogs therefore remain unknown, and the presence of antiβ2 glycoprotein I antibodies in dogs needs to be evaluated.
In conclusion, we suggest that lupus anticoagulants and aCL antibodies contribute to the prolongation of aPTT in a subset of apparently healthy adult Bernese Mountain Dogs. The significance of the findings in the dogs remains unknown at present, and needs further confirmation in a larger set of dogs. Further studies including canine specific quantitation of aCL antibodies, further antiphospholipid antibody identification, and association with disease processes are in progress.
We acknowledge Veterinary Technician Michelle Dupont and Bioanalyst Annette Urbrand Martinsen for technical assistance in this study. The study was funded by a PhD project.
The paper was presented in part as a poster at the ECVIM/ECVCP congress in Toulouse, France, September 2010.
VACUETTE Greiner Bio-one, Hannover, Germany
HemosIL reagents, Instrumentation Laboratory, Bedford, MA
ACL 500, Instrumentation Laboratory
Immuno 96 MicroWell, polysorb, F bottoms, Nunc, Herlev, Denmark
Sigma-Aldrich, Broenby, Denmark
Kem-En-Tec Diagnostics A/S, Taastrup, Denmark
Labsystems Multiskan RC, Labsystems Inc, Vienna, VA
Dog IgG ELISA kit, ICL Immunology Consultants Laboratory, Inc, Lake Oswego, OR
GraphPad Prism, 5.0b, GraphPad Software Inc, La Jolla, CA