• Open Access

Plasma D-Dimer Concentrations during Experimental EHV-1 Infection of Horses




Central nervous system blood vessel thrombosis is a part of the pathogenesis of equid herpesvirus-associated myeloencephalopathy (EHM). D-dimers (DD) are stable breakdown products of cross-linked fibrin, and increased DD-plasma concentrations could reflect the degree of systemic coagulation during EHV-1 infection.


We hypothesized that blood DD concentrations will be increased during periods of EHV-1 fever and viremia, reflecting an activated coagulation cascade with fibrinolysis.


Twenty-eight equids were infected with EHV-1 in 3 experimental infection studies. Three (uninfected) horses were included in a separate study to evaluate methodology for DD concentration measurements.


Clinical data and quantitative viremia were evaluated, and DD concentrations were measured in blood samples on the day before the infection and during days 1–12 postchallenge. Uninfected horses were sampled every 3 hours for 48 hours. Logistic and linear regression was used to investigate the potential association between the fever and viremia with the presence or absence of DD concentrations in peripheral blood.


DD concentrations were increased for 1–8 days in the majority of infected animals. Both viremia (odds ratio [OR] 6.3; 95% confidence interval [CI] 3.4–11.8; P = .0013) and fever (OR 4.9; CI 2.3–10.1; P = .001) were strongly associated with the likelihood of detecting DD in peripheral blood.

Conclusions and Clinical Importance

EHV-1 viremia is associated with increases in DD concentration in horses and ponies. This indicates that EHV-1 viremia can lead to an activation of coagulation and fibrinolysis.


(quantitative) polymerase chain reaction


chemokine (C-C motif) ligand 2


confidence interval


central nervous system




endothelial cell


equid herpesvirus-associated myeloencephalopathy


equid herpesvirus


glyceraldehyde 3-phosphate dehydrogenase


glycoprotein B


generalized estimated equations




monocyte chemoattractant protein-1


odds ratio


post infection


peripheral blood mononuclear cell


plaque forming units


receptor antagonist


serum neutralization


tissue factor


tumor necrosis factor

Coagulation, anticoagulation, and fibrinolysis in mammals are the cornerstones of hemostasis.[1] The appropriate balance between these mechanisms ensures the formation of a functional thrombus, composed of fibrin, to stop bleeding from a damaged blood vessel. Fibrinolysis is the process of thrombus breakdown and aims to restore blood flow. Plasma D-dimers (DD) are breakdown products of factor XIII-activated, cross-linked fibrin, and DD can only be detected once coagulation is followed by fibrinolysis.[2] Hemostasis is tightly regulated and orchestrated by coagulation and anticoagulation factors, as well as by fibrinolysis and antifibrinolysis factors. All factors are circulating in plasma and/or are directly derived from platelets, endothelial cells (EC), and monocytes upon activation.[1] Similar to many mammalian species relevant systemic inflammation in the horse might alter the EC surface from a resting into a prothrombotic state, a process that is likely mediated by circulating proinflammatory cytokines.[3, 4] Although there is a continuous (low volume) production of DD during homeostasis, inflammatory conditions such as endotoxemia or neonatal septicemia in the horse have been found to cause DD product increases.[5-7]

Equid herpesvirus-associated myeloencephalopathy (EHM) is secondary disease that can follow respiratory tract infection with equid herpesvirus type-1 (EHV-1) in sporadic cases. Infectious dose, virus strain, pre-existing immunity, but also host genetic factors are thought to be responsible for differences in clinical outcome.[8, 9] In cases of EHM, lesions are localized to the central nervous system (CNS) and are commonly scattered throughout the spinal cord white and gray matter. Usually, lesions are consistent with vasculitis and thrombosis of the CNS vasculature, hemorrhage, and mononuclear cell extravasation into the neural parenchyma.[10] Vascular lesions are thought to develop in close proximity to EHV-1-infected EC, and the virus is assumed to reach the CNS-vasculature EC exclusively by intracellular viremia in peripheral blood mononuclear cells (PBMC).[11, 12] After experimental EHV-1 infection by nasopharyngeal route there are 3 clinical phases over a 20-day postinfection (p.i.) period. Within 12–24 hours of inoculation, a primary fever spike occurs, which is typically short-lived and returns to normal by day 3 p.i. (phase 1). A secondary fever spike may begin on days 3–6 p.i. in the majority of cases; it typically lasts for 1–3 days, and commonly coincides with PBMC-associated intracellular viremia (phase 2).[13] Phase 3 of the disease is the occurrence of EHM, and occurs from late in phase 2 or up to approximately 20 days p.i. Although the majority of EHV-1 infected horses will become viremic, only a variable but consistently lower percentage of viremic horses will develop neuropathology.[14]

As increased DD concentrations in humans can be measured during transient ischemic attack to the brain, and because localized vasculitis and thrombosis are believed to be intrinsic components of EHM pathology, we speculated that DD concentration increases may be present as early as during viremia in the critical phase 2 of infection, consistent with a prothrombotic state.[15] We hypothesize that phase 2 (viremia) of EHV-1 infection induces a prothrombotic state in horses leading to increased DD concentrations.

To test this hypothesis, our objective was to measure DD concentrations in plasma during experimental EHV-1 infection studies collected from yearling horses, yearling ponies, and aged horses, and determine any temporal association with viremia.

Materials and Methods


This study included horses or ponies from 3 independent experimental EHV-1 infections (groups 1–3), as well as 3 uninfected horses to control for possible effects of frequent venipuncture on DD concentrations in plasma (group 4). Group 1 contained 8 healthy yearling (Western stock) mixed sex horses that were infected with 5 × 107 PFU/mL of EHV-1 strain Ab4. Group 2 contained 10 healthy yearling mixed sex ponies that were infected with 5 × 107 PFU/mL of EHV-1 strain Ab4. Group 3 contained of 10 aged (>18 years) horses, enrolled in an antiviral drug study, that were infected with 1 × 107 PFU/mL EHV-1 strain OH03. Both virus strains that were used for infection are so-called “neuropathogenic” EHV-1 strain variants with a single point mutation in the polymerase gene.[16] Animals in group 1 and 2 were tetanus vaccinated, and had a low or absent preinfection serum neutralization (SN) titer against EHV-1 of ≤1 : 32. In group 3 an EHV-1 specific (glycoprotein G) ELISA1 was used to select for horses with low or absent EHV-1 specific antibody concentrations based on optical densities <0.5 of absorbance. A vaccination history of group 3 animals could not be obtained as horses were purchased at different sales. Group 4 contained 3 healthy Standardbred intact males, all 3 years of age. Although ponies in group 2 were group-housed in a roofed enclosure, all other horses were housed in individual stalls. All animals were fed a diet of free choice grass hay or mixed alfalfa/grass hay diet with grain supplements and free choice water.

These studies were performed according to the guidelines of the Institutional Animal Care and Use Committees of Colorado State University (groups 1 and 2), Oklahoma State University (group 3), and Utrecht University, the Netherlands (group 4).

Infection, Monitoring, and Sample Collections

All animals in groups 1–3 were infected by nasopharyngeal instillation of the virus suspension on day 0. As the objective for group 4 was to evaluate the effect of frequent blood draws on DD concentrations, the 3 uninfected horses in group 4 were sampled every 3 hours for 48 hours and examined every 12 hours, which included recording of rectal temperature data and inspection of both jugular veins. In the 3 infection studies, clinical examination data (nasal discharge, conjunctivitis, lymphadenopathy, coughing), rectal temperatures, and all sample collections were performed once daily in the mornings between day 2 and day 12 with the exception of day 0 according to an established protocol.[13] Gait evaluations were carried out before the study and otherwise performed daily with the exception of day 0. Changes in neurological gait qualities were recorded and scored according to an established scoring scale ranging from 0 to 5 with “0” being normal.[17] Any blood collection was carried out by direct venipuncture of left or right jugular vein. Venous blood for viral genome (glycoprotein B, gB) PCR quantifications in studies 1 and 2 was collected on all days with the exception of day 0. Study 3 did not measure viremia on days 1 and 3 p.i. For DD measurements venous blood samples were collected on preinfection day −1, then on days 4–12 in group 1, and on days 1–12 in groups 2 and 3.

Sample Preparation and Analysis

Viremia was determined in jugular vein blood collected in heparinized syringes (groups 1 and 2) or in EDTA tubes2 (group 3). Viremia in groups 1 and 2 was determined by detection of viral (gB) copy numbers in purified PBMC using a qPCR as previously described, and results were reported as gB copy numbers/1 × 106 β-actin gene copies.[18] In group 3, gB copy numbers were determined in purified PBMC using a different qPCR assay, where results were reported as gB copies/1 × 106 GAPDH gene copies.[19] Both assays assume that 2 gene copies of β-actin or GAPDH represent one nucleated cell. To allow a comparison of viremia magnitude among studies, samples were assigned an ordinal score (0–3) based on qPCR results. For this the maximum gB-copy count of either qPCR assay was selected and divided by 3. For the assay used in groups 1 and 2, a score of 1 was assigned to PCR results between 50 and 1466 gB copies/1 × 106 β-actin copies; a score of 2 to PCR results between 1467 and 2932 gB copies/1 × 106 β-actin copies, and a score of 3 for 2934–4400 gB copies/1 × 106 β-actin copies. For the assay used in study 3, scores 1, 2, or 3 were assigned to gB copy results 50–4000, 4001–8000, and 8001–12000 gB copies/1 × 106 GAPDH gene copies, respectively.

Collection and processing of all citrated blood samples (groups 1–4) followed the same protocol. Jugular vein blood for determination of DD was collected by direct venipuncture in tubes containing 3.8% Na citrate.2 Tubes were immediately stored on ice for no longer than 30 minutes, followed by centrifugation (1000 g × 10 min). The plasma fraction was transferred in individual 2-mL containers, which were kept on ice until transported to the laboratory. Upon arrival, containers were immediately frozen at −80°C. For analysis, plasma samples from individual studies were allowed to thaw in batches of 4. A semiquantitative D-dimer assay3 was used to determine DD concentrations according to manufacturer's instructions. This assay determines DD concentrations using a mouse-derived antibody directed against the human epitope that is present in the FactorXIIIa cross-linked fragment D domain of fibrin. Step 1 is a qualitative determination of DD. Once positive agglutination was observed, a separate 100 μL plasma sample was serially diluted (1 : 2; 1 : 4; 1 : 8, 1 : 16, etc). Each dilution was combined with the antibody until no agglutination was observed. Results of each test were compared to a positive and negative control. An ordinal score (0–4) was assigned to each sample regarding relative DD concentration based upon agglutination of samples at different dilutions. A negative agglutination in the first sample corresponded to a DD concentration of <250 ng/mL (DD score = 0). A positive agglutination at a specific dilution corresponded to the manufacturer predetermined range of DD concentrations expressed in ng/mL: (i) 1 : 2 = 250–500 ng/mL (DD score = category 1); (ii) agglutination at 1 : 4 = 500–1000 ng/mL (DD score = category 2); (iii) 1 : 8 = 1000–2000 ng/mL (DD score = category 3); and (iv) 1 : 16 ≥ 2000 ng/mL (DD score = category 4). To control for a potential prozone effect plasma samples (day 1–8) from 3 horses out of group 3 were retested.[20] Two horses were selected because of previously measured significant DD increases during days 6–12, whereas a third horse was selected because of a modest DD increase on days 6 and 7 p.i. The assay was repeated in 1 : 5 and 1 : 10 saline-diluted plasma samples.

Statistical Analysis

Data were initially evaluated by examining frequencies for categorical variables and data distributions and descriptive statistics for continuous variables. Only data from days 4–12 p.i. were used. A fever was defined as a rectal temperature measured ≥38.5°C (101.3°F) (dichotomized as ≥38.5°C = 1, <38.5°C = 0); cell-associated viremia was defined as a viral load >50 EHV-1 gB copies/1 × 106 housekeeping gene copies and dichotomized as =1, and <50 EHV-1 gB copies/1 × 106 housekeeping gene copies =0. An agglutination test for DD was considered positive if agglutination occurred at the first stage (undiluted) or following dilutions (dichotomized as positive = 1, negative = 0). Logistic regression was used to investigate whether dichotomized data for fever or viremia were associated with the likelihood of being positive for DD presence. Then, Spearman rank correlation was used to analyze the relationships between DD and viremia scores on the same day, and also between DD with a 1 day lag for viremia score. DD and viremia scores were summed for all study days for each horse and divided by the number of sampling days to calculate the average DD score and average viremia score per day. Linear regression was used to investigate whether the average viremia score per day was a significant predictor of the average DD score per day. Generalized estimating equations (GEE)3 were used in logistic regression to control for the lack of independence among observations that were made in the same animals on multiple days. Confounding that might be associated with obtaining data from different studies was controlled in models by inclusion of study (nominal variable) as a fixed effect.4 The exposure variables of interest were also dichotomized as yes/no variables, as described above, to facilitate a logistic regression analysis, to evaluate an association between fever following day 4 p.i., detection of gB viral genomic copy numbers in PBMC (viremia), and DD concentrations. Formal comparisons between the different studies that generated these data were not performed for various reasons. Experiments were not conducted simultaneously and were performed at different locations; experiments differed with regard to virus strain and dose used for experimental infection, differed with regard to age and breed of the animals, and PCR methodology used to detect viremia differed between experiments.


Clinical Results

All 3 experimental infection studies elicited typical clinical signs of a primary EHV-1 infection: fevers (serous), nasal discharge, conjunctivitis, and mandibular lymphadenopathy. None of the group 4 horses were febrile or showed signs of disease during the sampling period. The majority of infected animals developed a biphasic fever with the first phase of fever during the immediate 24–72 hours postinfection. All horses in group 1 developed a secondary fever for at least 1 day. Two ponies in group 2 and 2 horses in study 3 failed to develop a secondary fever. The pre- and postinfection rectal temperature data for each study, shown as the mean and standard deviation (SD), are shown for each day in Figure 1 (panels b, d, f). Total number of febrile animals and the number of febrile animals per day during the experimental period are shown in Table 1.

Table 1. Data overview of 3 EHV-1 infection studies. Number of horses positive for each category “fever”, “viremia”, and “positive for D-dimer” (DD) are listed for preinfection day −1 and postinfection days 1–12 as well as the percentage of positive horses for various periods: day −1; days 1–4; days 5–9, and days 10–12.
Day Post-InfectionFevernPeriod (%)SummaryViremianPeriod (%)SummaryDD-PositivenPeriod (%)Summary
Figure 1.

Panels a&b; c&d and e&f show results for groups 1, 2, and 3, respectively. Panels a, c, and f show the D-dimer results in categories 0–4 (0 ≤ 250 ng/mL; 1 = 250–500 ng/mL; 2 = 500–1000 ng/mL; 3 = 1000–2000 ng/mL; 4 ≥ 2000 ng/mL, y-axis) of individual EHV-1 infection studies 1–3. D-dimer concentrations were determined on the day before infection (Day −1), and during postinfection days 1–12. D-dimer concentrations were not determined on days 1–3 in study 1. Panels b, d and f show the mean (± SD) of rectal temperature measurements in °C indicated on the left y-axis and the individual EHV-1 viremia data (□) for each animal (expressed as gB copies/1 × 106 beta-actin or GAPDH gene copies indicated on the alternative, right y-axis) on the preinfection day and during the postinfection observational period (days 1–12; x-axis).

One horse of group 3 was found with a mild (grade 1 out of 5) neurological gait abnormality before infection, which worsened by one grade p.i. All other animals appeared normal on preinfection gait evaluation. One horse in group 1 developed severe EHM on day 10 with acute signs of quadriplegia (score: 5/5), and was euthanized the same day. None of the ponies (group 2) showed clinical signs of EHM, whereas 6 horses in study 3 developed mild–moderate signs of EHM with a 1–3 grade worsening in neurological gait quality when compared to their individual preinfection scores.

Viremia Results

All horses in group 1 became viremic (Fig 1b, d, and f). Two ponies (group 2) were not found viremic but still developed a secondary fever for 1 or 2 consecutive days, respectively. In group 3 there was 1 horse without detectable viremia despite having a secondary fever. Furthermore, there were 2 horses that did not develop a secondary fever and the principal investigator of study 3 decided to not measure viremia.

Viremia was detected in all group 1 horses for 2–6 days with a median duration of 3.5 days; 0–7 days duration with a median of 2.5 days in group 2, and for 0–8 days with a median of 4 days in group 3. Total number of viremic animals and the number of viremic animals per day during the experimental period is shown in Table 1.

D-Dimer Results

All plasma samples obtained from frequent blood draws in group 4 horses tested negative for DD products (data not shown). DD concentrations in group 1–3 were all undetectable on preinfection days. All day 4 samples p.i. were negative in group 1—no DD samples were collected on days 1–3 in this group. DD concentrations in groups 2 and 3 were negative on days 1–4 p.i. with the exception of 1 horse in group 3 which showed a single category 2 DD increase on day 3. On that day the horse was not febrile or viremic, and DD concentrations in this horse were undetectable on previous days and on day 4 and 5. All animals in groups 1 and 3 had DD increases on at least 2 consecutive days. Two group 1 horses had 5 days of consecutive DD increases, whereas 2 horses of group 3 had 7 or 8 consecutive days of DD increases, respectively. Longest consecutive days of DD increases in group 2 were 3 days (2 ponies). Furthermore, 2 ponies in group 2 had DD increases on 1 day only, and 3 ponies did not have DD increases at all (Fig 1a, c, and e). The highest DD- concentrations measured in any study was category 3 (1000–2000 ng/mL). The highest increases in group 2 were category 2 increases (500–1000 ng/mL). DD increases were found between days 3 and 12 in groups 1 and 3 (Fig 1a and e). In group 2, DD increases were detected between days 5 and 8 (Fig 1c).

Three out of 4 horses or ponies without a secondary fever were still found positive for DD during 1–3 days; however, all measured DD concentrations were at category 1. Total number of animals positive for DD and the number of DD-positive animals per day during experiments are shown in Table 1.

Controlling for differences between studies and for lack of independence that was related to repeated sampling of horses, viremia and fever >day 4 were both strongly associated with the likelihood of horses being positive for DD in peripheral blood (OR, 6.3; 95% CI, 3.4–11.8; P = .001; and OR, 4.9; 95% CI, 2.3–10.1; P = .001, respectively).

Overall, DD, fever, and viremia were all detected with the same relative frequency between days 1 and 12 p.i. (25–30% of horse-days of observation; Table 1). DD increases typically were measured on the day of or the day following viremia. A fever before day 4 was not typically detected in combination with viremia or with DD where measured (groups 2 and 3) (Table 1). DD, viremia, and fever were all relatively common between days 5–9 p.i. (35–63% of horse-days), but DD and viremia continued to be detected on days 10–12 p.i., whereas fever generally was not. Increased DD concentrations commonly persisted for few more days once viremia became undetectable.

Average viremia score per day was a significant predictor of the average DD score per day (Fig 2). Regression analysis suggested that a 1-unit increase in the average viremia score resulted in a mean increase of 0.6 (95% CI, 0.06–1.1) in the average DD score per day. In general, viremia was detected just before detecting increases in DD in the peripheral blood. Viremia was detected 1 day before detection of DD for 50% of subjects (13/26) and on the same day in additional 2 horses. The general observation that viremia preceded DD increases is also supported by comparing the correlation between DD scores measured on the same day as viremia scores (Spearman r = 0.36) to the correlation measured with a 1-day lag for viremia scores (Spearman r = 0.45).

Figure 2.

Sums of scores per day were calculated for D-dimer (DD) (y-axis) or viremia results (x-axis) from 3 independent EHV-1 infection experiments. Scores were calculated as follows: DD scores 1–3 corresponded to agglutination results 1 = 250–500 ng/mL; 2 = 500–1000 ng/mL; 3 = 1000–2000 ng/mL of a semiquantitative plasma DD assay. The maximum measured viremia result in EHV-1 gB copies/1 × 106 β-actin copies of any study was identified and divided by 3, resulting in 3 classes. Individual viremia results were then assigned to 1 of 3 classes. D-dimer or viremia scores of individual animals on a particular day were summed and divided by the sum of dichotomized DD or viremia-positive animals on that day, resulting in an average score per day. Spearman rank correlation was used to analyze the relationships between DD and viremia scores on a same day.

A prozone effect, because of overwhelmingly high DD concentrations in samples blocking the agglutination reaction of the antibody in the assay, was not detected.


DD were detected during all 3 experiments in the majority of infected horses/ponies. DD increases in individual animals were associated in time with the duration of EHV-1 viremia as well as with its associated fever. In addition, high viremia scores were positively associated with high DD scores. However, DD increases during the primary fevers were absent (group 2 and 3). It is noteworthy that DD increases during viremia and viremic fevers were consistently noticed in all 3 groups that were studied, despite the use of different age groups, different breeds, and two different viral strains during the 3 different experiments.

The DD assay utilizes a mouse-derived antibody to detect specifically human DD. Although the assay has been used for a number of studies in the horse, its sensitivity and specificity have not been fully determined. However, presumed DD increases found during sepsis and endotoxemia in equids were measured in conjunction with other abnormal parameters of hemostasis, such as a prolonged PT and APTT, thrombocytopenia, decreased antithrombin activity, and others with good correlation.[5-7] Based on this validation we elected not to include more parameters of coagulation into this study.

Although a number of publications on experimental EHV-1 infections have shown the association between a secondary fever and cell-associated viremia, DD concentration increases have, so far, only been measured in horses with other significant systemic disease.[4, 5] We showed that DD production, as an indicator of an activated coagulation, appears to be linked to EHV-1 viremia; however, at this stage, we can only speculate about possible pathways that lead to DD increases. Potential pathways can be 2-fold. One possibility could be an indirect pathway where viremia induces proinflammatory cytokine production from infected and uninfected PBMC, which induces a prothrombotic state with, subsequently, an increased fibrin generation and break-down. Circulating cytokines TNF-α, IL-1, IL-6, IL-8, and MCP-1 (CCL2) in humans are associated with a prothrombotic state,[8-10] and whereas an activated coagulation and/or increased DD concentrations were found during severe gastrointestinal, ischemic disorders in horses, Lopes et al successfully detected mRNA up-regulation of TNF-α, IL-1RA, IL-6, and IL-8 in PBMC under similar conditions of gastrointestinal disease.[6, 21, 22] Furthermore, TNF-α concentration increases in plasma and abdominal fluids from horses with these conditions have also been detected.[23, 24] Although these results suggest that there is evidence of concurrent cytokine and DD production, it is possible that DD production in horses/ponies during EHV-1 viremia is also cytokine-driven. A mainly proinflammatory cytokinemia could also provide a good explanation of the viremia-associated fevers during EHV-1 infections. Few studies, in vitro or in vivo, have focused on cytokine and chemokine production during EHV-1 infection mainly using PBMC.[25-27] However, none of the cytokines/chemokines identified in these studies could be linked to DD production as it occurs in other species; neither was there significant overlap with viremia. This was unexpected when we searched for possible explanations; however, as cytokine production and breakdown is rapid and usually arranged in cascade-like pathways, a low sampling frequency, as is typical during EHV-1 infection experiments, may have caused a lack of cytokine detection. Furthermore, all studies used PBMC, whereas other cell populations, eg, EC, pericytes, myocytes, osteoblasts, or fibroblast with a potential of inflammatory cytokine production could not be included in these studies.[28]

An alternative possible explanation of DD-production during EHV-1 viremia has been recently suggested by Yeo et al. They describe significantly increased release of tissue factor from EHV-1-infected equine monocytes in vitro when compared to uninfected monocytes.[29] As the percentage of infected PBMC is very low during EHV-1 viremia and the monocyte is not the predominant virus carrier during viremia, these interesting findings warrant further investigation.[30, 31]

Typically, there is a biphasic fever in most animals during experimental EHV-1 infection. The first fever is noticed within 24–36 hours of infection and is believed to be caused by viral replication and tissue destruction within the upper respiratory tract. Although the second fever is associated with EHV-1 viremia it is also the phase where DD production was consistently increased. The start of EHV-1 viremia also corresponded to the initiation of DD production. Although the locality, respiratory epithelium versus blood stream, and potential causes for the 2 distinctly separate periods of fevers are different, there may be another reason for the lack of DD production during the primary fever period. A prozone effect could have occurred which can be detected when an overwhelming amount of antigen is present in the test sample. A high DD concentration, produced in the first phase of the infection, could have blocked the agglutination reaction of the assay, resulting in false negative results. However, the prozone effect can be overcome by careful sample dilution to decrease the amount of antigen, which was performed. Still, after dilution of the samples none of the previously negative samples turned positive for DD, and previously measured low DD-concentration results did not increase after dilutions. This strengthens the idea that DD-production is linked to viremia and its associated fever, but not to the primary fever.

In individual animals, on occasions, we noticed discrepancies between simultaneous presence of fever, viremia, and DD concentrations in same-day samples. An explanation may be the once-a-day sampling technique that is typical for experimental EHV-1 infection studies, which may not be sensitive enough to detect fluctuations of parameters of interest within a 24 hour period. This has to be addressed in future studies.

DD research in humans showed that increases, in combination with other biomarkers, help the emergency room clinician to identify transient ischemic attack or cerebral stroke.[32] This is an intriguing finding as EHM pathology has similarities with transient ischemic attack or cerebral stroke in humans.[11] However, our studies were not designed to evaluate correlations between DD production and the development of (stages of) myelopathy or EHM, as we used yearling horses (group 1) and ponies (group 2), which are known to be least likely to develop EHM under experimental conditions.[13] This was different in group 3 where aged horses were selected because of an increased likelihood to develop EHM; however, some of these horses were also simultaneously treated with virustatics during phase II-fevers to evaluate the drugs' effects on EHM outcome and severity.[9, 12, 33] EHM most often follows the last day of viremia, and thrombosis of spinal cord vessels is a feature of EHM, but it is not the predominant feature. High DD concentrations are therefore more likely to be part of a systemic response and not associated with the spinal cord vasculature. Furthermore, DD concentrations started to rise immediately with the beginning of viremia, and not toward the end of viremia.

In a previous article we showed that in vitro infection of EC with EHV-1 only occurs in a contact model between the virus-infected PBMC and the EC.[34] We further presented in vitro data that demonstrate significantly decreased EC infection in our established contact model when PBMC and EC were pretreated with anti-inflammatory drugs.[35] We postulate that the same or simultaneously produced inflammatory stimuli that are capable of systemic DD production may also play a role during EC infection with EHV-1 by inducing contact molecules on EC and PBMC. Contact between a virus-infected PBMC and an EC would then be facilitated and EC infection may be increased. DD increases are markers of inflammation and associated with increased proinflammatory cytokine production in many species. With this in mind, DD may be a surrogate marker for inflammation and the prudent use of anti-inflammatory drugs early on during the course of an EHV-1 infection may be warranted to decrease the rate of EC infection.

This study suggests that mechanisms to cause a fever during viremia are different from those responsible for the primary fever during an EHV-1 infection. Only during viremia there is substantial inflammation and activated coagulation leading to a detectable DD increase. DD increases alert us to the extent of inflammation and the possible effects this inflammation might have on vasculature surfaces and the infection of EC with EHV-1. Whether there is a role for coagulation during the early stages of EHM, however, needs to be determined.


This study was in parts made possible through a grant from the Grayson Jockey Club Research Foundation. Results of this study were presented at the 2011 ACVIM Forum, Denver, CO (June 15–18, 2011), and during the European Veterinary Congress “Voorjaarsdagen” in Amsterdam, NL (April 2–5, 2012).

Conflict of Interest: Authors disclose no conflict of interest.


  1. 1

    Svanovir® EHV1/4-Ab, Boehringer Ingelheim Svanova, Uppsala, Sweden

  2. 2

    BD Vacutainer, Becton Dickinson, Franklin Lakes, NJ

  3. 3

    Minutex® D-dimer assay, Biopool Int, Ventura, CA

  4. 4

    Proc GENMOD SAS Inc, Cary, NC