This study was partially presented at the 9th International Equine Colic Research Symposium in Liverpool (UK), June 2008, and at the 3rd Congress of the European College of Equine Internal Medicine (ECEIM) in Barcelona (Spain), January 2009. The abstract has been published at the J Vet Inter Med 2009;23:434. The manuscript represents a portion of a thesis submitted by the first author at the Universitat Autònoma de Barcelona as partial fulfillment of the requirements for a PhD program, supervised by the corresponding author.
Corresponding author: Luis Monreal, Sevei de Medicina Interna Equina, Departament de Medicina i Cirurgia Animals, Facultat de Veterinària, Universitat Autònoma de Barcelona, 08193-Bellaterra, Barcelona, Spain; e-mail: email@example.com.
Background: Plasma d-dimer concentration is a useful marker to assess systemic coagulation and fibrinolytic activities in humans, dogs, and horses. Peritoneal fibrinolytic activity increases in horses with colic, especially in horses with endotoxin in the peritoneal fluid.
Hypothesis/Objectives: Peritoneal d-dimer concentration can be used to assess peritoneal fibrinolytic activity in horses with severe gastrointestinal (GI) disorders and altered peritoneal fluid.
Animals: Two hundred and twenty-one colic horses and 15 control horses.
Methods: Prospective observational clinical study. Blood and peritoneal fluid were collected on admission. Horses were grouped according to diagnosis, peritoneal fluid analysis, and outcome. Peritoneal d-dimer concentration was determined, together with peritoneal tissue-plasminogen activator (t-PA) and plasminogen activator inhibitor (PAI-1) activities. Plasma d-dimer concentration also was measured.
Results: Peritoneal d-dimer concentration was significantly higher in all colic groups compared with controls, and in horses with enteritis, peritonitis, and ischemic disorders compared with horses with large intestinal obstructions. Peritoneal d-dimer concentration was significantly higher in horses with altered peritoneal fluid (modified transudate and exudate) compared with horses with normal peritoneal fluid analysis. Plasma d-dimer concentration also was significantly higher in the peritonitis group, and in horses with altered peritoneal fluid analysis. Peritoneal and plasma d-dimer concentrations also were significantly higher in nonsurvivors. Peritoneal d-dimer concentration was significantly correlated with decreased peritoneal t-PA activity and increased peritoneal PAI-1 activity.
Conclusions and Clinical Importance: Peritoneal d-dimer concentration is markedly higher in severe GI disorders, and it can be used to assess peritoneal fibrinolytic activity in horses with colic.
Mesothelial cells of the peritoneum participate in initiating and resolving inflammation and repair by secreting various pro-, anti-, and immunomodulatory mediators.1,2 These include products of the coagulation and fibrinolysis systems, chemokines, cytokines, and growth factors, among others. Thus, mesothelial cells not only have procoagulant activity, but also marked intrinsic fibrinolytic activity. Procoagulant activity (fibrin deposition) is aided mainly by the release of mesothelial plasminogen activator inhibitor type 1 (PAI-1), whereas fibrinolytic activity is mediated by secretion of tissue type-plasminogen activator (t-PA) that is essential to destroy the fibrin formed after different gastrointestinal (GI) injuries and to decrease subsequent risk of adhesion formation.1–3 When the balance between peritoneal fibrinogenesis and fibrinolysis is markedly altered and fibrin formation exceeds fibrinolysis capacity, permanent adhesions appear, which are considered as the main cause of pain, intestinal obstruction, and repeat laparotomy in horses and in human beings.1,4–6
In humans, intra-abdominal adhesion formation after surgical trauma or peritonitis has been attributed to a decreased peritoneal fibrinolytic capacity. Studies performed on peritoneal biopsies during and after surgery have shown changes consistent with a decrease in t-PA and an increase in the main fibrinolytic inhibitor (PAI-1).6–12 These changes were related to surgical trauma, duration of the surgery, presence of previous adhesions, or peritonitis. Similar changes in peritoneal fibrinolytic activity have been observed in patients with peritonitis in continuous ambulatory peritoneal dialysis, in which adhesion formation was more likely to occur. These patients had a decrease in peritoneal t-PA activity and fibrin(ogen) degradation products (FDPs), and an increase in peritoneal PAI-1 activity, which were consistent with a decrease in peritoneal fibrinolysis activity and a higher risk for adhesion formation.13–16
In horses, few studies have been performed assessing peritoneal fibrinolytic activity, but it is thought to have some differences compared with humans. Baxter et al17 demonstrated an increase in anticoagulation proteins (antithrombin and protein-C) and fibrinolytic components (plasminogen and FDPs) in peritoneal fluid after abdominal surgery. These changes could be consistent with a peritoneal hyperfibrinogenesis and hyperfibrinolysis response during the first days after abdominal surgery. Later, Collatos et al18,19 detected an increase in peritoneal t-PA and FDPs in horses with colic, consistent with an increase in fibrinolytic activity, especially in horses with endotoxin in their peritoneal fluid and in horses that did not survive.
d-dimer is a degradation fragment released exclusively by plasmin-mediated lysis of cross-linked fibrin. Thus, d-dimer is a specific indicator of fibrinolysis, in contrast to FDPs, which are fragments released by plasmin degradation of either fibrinogen or fibrin.20 Plasma d-dimer has been very useful in assessing blood hypercoagulation and hyperfibrinolysis in horses after endurance competition, in horses with acute GI disorders, in those with laminitis, and in septic foals.20–28,a,b It is also one of the main diagnostic tests in humans suspected to have thromboembolic diseases.29–31 Thus, plasma d-dimer concentration currently is considered a sensitive marker in assessing fibrinolytic activity, and consequently coagulation activity. Any increase in d-dimer concentration is strongly related to an increase in fibrin destruction (hyperfibrinolysis) secondary to an increase in fibrin formation (hypercoagulation or hyperfibrinogenesis).
Although plasma d-dimer determination is a very useful test to assess fibrinolysis activity, few studies have been performed on peritoneal fluid. Most of the studies have been carried out in humans, and few have been performed in experimental animal models.15,16,32,33 However, to our knowledge, no studies have been conducted to evaluate fibrin formation and fibrinolytic activity in peritoneal fluid of horses with GI diseases using d-dimer measurements. Therefore, the aim of the present study was to assess the fibrinolytic activity in the peritoneal fluid of horses with colic by determination of peritoneal d-dimer concentration, together with peritoneal t-PA and PAI-1 activities, according to the GI disorder, the type of peritoneal fluid (transudate, modified transudate, and exudate), and the outcome. Plasma d-dimer concentration also was measured to detect any relationship with peritoneal fluid analysis results.
Materials and Methods
In this prospective study, all horses with colic admitted to the Equine Teaching Hospital of Barcelona between September 2004 and September 2007, from which peritoneal fluid together with blood samples were collected for diagnostic purposes with the owner's permission, were included. Additionally, blood and peritoneal fluid from horses with different processes not affecting the GI tract also were collected by identical techniques and with the owner's consent. These animals were considered as controls because physical examination, diagnostic tests, peritoneal fluid analysis, and postmortem examination (in those in which it was performed) confirmed that they did not have GI alterations.
Horses with colic were allotted to 5 groups according to the diagnosis, based on clinical history, complete physical examination, and results of complementary diagnostic tests (CBC, plasma biochemistry, blood gas analysis, abdominal ultrasonography, and peritoneal fluid analysis). Findings of abdominal radiology, laparotomy, or postmortem examination were used for this classification whenever it was performed.
The obstructive group included all horses with nonstrangulating, noninflammatory disorders such as impactions and large colon displacements without signs of intestinal devitalization,34 which resolved with medical therapy (fluid therapy, laxatives, or both). The enteritis group included horses with acute duodenojejunitis and typhlocolitis. The ischemic group included horses with surgical lesions, such as intestinal volvulus or torsion, epiploic entrapment, inguinal hernias, and intussusceptions. The peritonitis group included horses with gastric or intestinal ruptures, as well as septic peritonitis (presence of intra- and extracellular bacteria in the peritoneal fluid cytology) caused by bowel devitalization but without rupture, intra-abdominal abscesses and other causes. The mixed/other processes group included horses with ≥2 disorders of similar severity (such as ischemic plus inflammatory disorders) and horses with other diseases such as malignancy.
Horses also were grouped according to peritoneal fluid analysis, as previously reported35,36: transudate (nucleated cell count [NCC] ≤ 5000 cells/μL and total protein [TP] ≤ 2.5 g/dL), modified transudate (NCC ≤ 5000 cells/μL and TP > 2.5 g/dL, or NCC > 5000 cells/μL and TP ≤ 2.5 g/dL, with a normal peritoneal fluid cytology), and exudate (NCC > 5000 cells/μL and TP > 2.5 g/dL, or NCC > 5000 cells/μL and TP ≤ 2.5 g/dL with inflammatory peritoneal fluid cytology).
To assess the prognostic value of peritoneal d-dimer concentration of horses upon admission, patients were grouped according to the outcome: survivors (horses that were discharged from the hospital), nonsurvivors (horses that died during hospitalization because of progressive worsening of the disease or were euthanized based on grave medical prognosis), and the financial restraint group (horses that did not receive the suitable treatment for financial or personal reasons, and horses that were discharged without the clinicians' consent).
Blood was collected by direct jugular venipuncture into 5-mL tubes containing 3.8% sodium citrate by with a vacutainer system.c Peritoneal fluid was collected aseptically, with a sterile blunt teat cannula, approximately 2 cm to the right of the midline at the most dependent area of the ventral abdomen, by a standard technique.37 Peritoneal fluid also was collected in 5-mL tubes containing 3.8% sodium citrate for d-dimer, t-PA and PAI-1 analysis, and in 1 mL tubes containing K3EDTA for total NCC, protein concentration measurement, and cytologic evaluation. Blood and peritoneal citrated samples were immediately centrifuged at 1000 ×g for 15 minutes, separated from the sediment, and frozen at −70 °C until assayed.
Peritoneal Fluid Analysis
Analysis was performed on K3EDTA samples by a specialized laboratory within 12 hours after collection or by one of the authors (MAD) when the laboratory was not available (during weekend and night emergencies). For this reason, 2 automated blood cell countersd,e were used for NCC determination of peritoneal fluid. TP concentration was measured in both cases by refractometry. Cytospin preparationsf and Diff-Quick staing were used for microscopic evaluation of all peritoneal cytologies.
Peritoneal and Plasma d-Dimer Concentration (ng/mL)
Both parameters were determined in duplicate with commercial reagents and controlsh by a quantitative immunoturbidimetric latex agglutination assay,i according to the instructions provided by the manufacturer. This assay has been used in previous studies for equine plasma and peritoneal samples.28,a,b,j
Peritoneal t-PA and PAI-1 (IU/mL) Analysis
t-PA and PAI-1 activities were determined by ELISAs with the reagents and instructions provided by the manufacturer.k,l The t-PA assay had a sensitivity of 0.006 IU/mL, with an intra-assay precision of variation for low (9.8%), moderate (4.0%), and high (3.8%) values; whereas the PAI-1 assay only measured activity in the 0–112.5 IU/mL range.
Quantitative results are expressed as median and interquartile range for principal variables, which did not reach a normal distribution (peritoneal and plasma d-dimer concentrations, and peritoneal t-PA and PAI-1 activities), and rank analysis of variance (ANOVA) models with Bonferroni adjustment for multiple comparisons among groups were used for inferential analysis. Additionally, to compare d-dimer concentration between peritoneal fluid and plasma samples, a Wilcoxon test for paired measures was used. In other quantitative variables normally distributed, such as age of horses, the mean ± standard deviation (SD) was used for descriptive purposes. In the case of qualitative variables, such as sex of horses, diagnosis, type of peritoneal fluid and outcome, the description of them was based in absolute frequencies and percentages. For inferential analysis, the Fisher exact test was used and the significance level was adjusted by means of Bonferroni's method when necessary.
In order to assess the global relationship among the variables d-dimer, t-PA, and PAI-1 in peritoneal fluid, principal component analysis (PCA) with ranks values was performed. After discarding sphericity of the variables, the principal components with eigenvalues >1 were extracted and a new punctuation was performed and subsequently analyzed with an ANOVA model and the Ryan-Einot-Gabriel-Welsch range posthoc method for generated groups for the classification of disease.
A commercial software (SPSS for Windows, version 15.0)m was used for all the statistical analyses. Two-tailed type I error of 0.05 was used in all statistical inference.
During the study, 221 horses with colic and 15 control horses were included. Although peritoneal fluid collection was attempted in 31 control horses, sampling was achieved only in 15 animals. Seventy-one of the 236 animals included were stallions (30.1%), 78/236 were geldings (33.1%) and 87/236 were mares (36.9%), aged between 0 and 26 years (8.6 ± 5.6 years). The breed distribution reflected the hospital's referral population, with 83 Andalusians (35.2%), 26 Warmbloods (11%), 11 Arabians (4.7%), 5 Thoroughbreds (2.1%), 9 Draft horses (3.8%), 9 ponies (3.8%), 2 donkeys (0.8%), 23 horses representing other breeds (9.7%), and 68 cross-breed horses (28.8%).
Horses with colic were grouped according to the diagnosis as follows: the obstructive group consisted of 68 horses (30.8%), and included 53 horses with large colon impaction, 6 horses with enteroliths/fecoliths, 5 horses with small bowel obstruction, and 4 horses with nephrosplenic entrapment. The enteritis group consisted of 45 horses (20.4%) and included 18 horses with proximal enteritis and 27 horses with enterocolitis. The ischemic group consisted of 44 horses (19.9%), and included 14 horses with small bowel volvulus, 7 horses with inguinal hernia, 7 horses with epiploic entrapment, 3 horses with cecum and colon volvulus, 2 horses with intussusceptions, and 5 horses with other ischemic lesions, such as strangulating lipomas and others. The peritonitis group consisted of 38 horses (17.2%), and included 15 horses with bowel rupture, 12 horses with bowel devitalization signs because of a previous inflammatory or ischemic disease, 6 horses with intra-abdominal abscesses, 3 mares with peritonitis caused by severe metritis, and 2 horses with peritonitis because of an unknown cause. The mixed group/other processes group consisted of 26 horses (11.8%), and included 6 horses with intra-abdominal malignancies, 5 horses with severe intestinal obstruction and inflammatory lesion, 3 horses with severe sand-induced colitis and colonic torsion, 3 horses with a chronic enteropathy, and 9 horses with different abdominal diseases, such as hepatopathy.
Results of the peritoneal fluid analysis were available for 203 horses (unfortunately some cytologies were missed) and were classified as follows: 94/203 transudates (46.3%), 47/203 modified transudates (23.2%), and 62/203 exudates (30.5%). According to the outcome, 140 out of 221 colic horses (63.3%) were included in the survivor group, 64/221 (29%) in the nonsurvivor group, and 17/221 (7.7%) in the group with financial restraints. Results of peritoneal fluid analysis and outcome according to the diagnosis, as well as outcome according to the peritoneal fluid analysis, are shown in Tables 1–3.
Table 1. Number and percentages of horses classified according to the diagnosis and the type of peritoneal fluid analysis.
Type of Peritoneal Fluid
Data expressed as n (%).
Table 2. Number and percentages of horses classified according to the diagnosis and the outcome.
Data expressed as n (%).
Financial restraint group
Table 3. Number and percentages of horses classified according to the type of peritoneal fluid analysis and the outcome.
Data expressed as n (%).
Financial restraint group
Significant relationships were confirmed between the type of peritoneal fluid and the diagnosis (Fisher's exact test, P < .001), the outcome and the diagnosis (Fisher's exact test, P < .001), and the outcome and the type of peritoneal fluid (Fisher's exact test, P < .001). When comparing the type of peritoneal fluid and the diagnosis, significant differences (error type I adjusted by means of Bonferroni's for posthoc comparison was 0.003) were found in the obstruction, peritonitis, and control groups. Mainly, the obstructive group showed a statistically higher (P < .00001) number of horses with a transudate compared with enteritis, ischemic, and peritonitis groups; and the peritonitis group showed a statistically higher (P < .00001) number of horses with an exudate compared with all other colic groups. When comparing the diagnosis and the outcome, significant differences (type I error adjusted by means of Bonferroni for posthoc comparison was 0.005) were found in the obstructive group and the peritonitis group. The obstructive group showed a statistically higher number of survivors, whereas the peritonitis group showed a statistically higher number of nonsurvivors, compared with all other colic groups. When comparing the type of peritoneal fluid and the outcome, pairwise comparisons showed significant differences among all groups (type I error adjusted by means of Bonferroni for posthoc comparison was 0.017).
Peritoneal and Plasma d-Dimer Concentrations According to the Diagnosis
Peritoneal d-dimer concentration in control horses was very low (36 ng/mL, 4.0–88.0), whereas it was significantly (P < .001) higher in all colic groups (Table 4). When the diagnosis was considered, peritoneal d-dimer concentration was significantly lower in the obstructive group compared with the enteritis (P= .016), ischemic (P < .001), and peritonitis (P < .001) groups (Table 4).
Table 4. Peritoneal and plasma d-dimer concentrations in control horses and in horses with colic grouped by the diagnosis.
Control (n = 15)
Obstructions (n = 68)
Enteritis (n = 45)
Ischemia (n = 44)
Peritonitis (n = 38)
Mixed/Others (n = 26)
Data expressed as median (interquartile range).
Statistical differences within the same row, by the rank-ANOVA model with Bonferroni's adjustment.
Different from the control group.
b Different from the obstructive group.
c Different from the enteritis group.
d Different from the ischemic group.
Different from the mixed/other processes group.
* Statistical differences when compared to peritoneal fluid samples within the same diagnostic group, using the Wilcoxon test.
On the other hand, plasma d-dimer results were clearly different when compared with peritoneal fluid analysis results, which were statistically lower in control horses and extremely high in colic horses (Table 4). Moreover, plasma d-dimer concentration was significantly (P < .001) higher in the peritonitis group when compared with the control and all colic groups. This increase in plasma d-dimer concentration was proportional to the large increase observed in peritoneal fluid of those horses with peritonitis. Significant (P= .018) differences also were found in plasma d-dimer results between the enteritis and the obstructive groups, being markedly higher in the enteritis group.
Peritoneal and Plasma d-Dimer Concentrations According to the Peritoneal Fluid Analysis
When peritoneal d-dimer concentrations were compared according to the type of peritoneal fluid analysis, d-dimer concentration was markedly higher in exudates and modified transudates, and both concentrations were significantly (P < .001) higher than that found in pure transudates (Fig 1). Plasma d-dimer concentration also was significantly higher in the group of exudates compared with the groups of transudates (P < .001) and modified transudates (P= .01), as well as in the group of modified transudates compared with transudates (P= .001) (Fig 1). Moreover, peritoneal d-dimer results in altered peritoneal fluids were extremely higher than plasma concentrations in the same animals. Significant differences were found between peritoneal and plasma d-dimer concentrations in transudates, modified transudates, and exudates (P < .001).
Peritoneal and Plasma d-Dimer Concentrations According to the Outcome
Plasma and peritoneal d-dimer concentrations showed a significant relationship with outcome. Their median values in the nonsurvivor group in peritoneal fluid and plasma (14,399 ng/mL [1,875–60,680] in peritoneal fluid and 2,499.5 ng/mL [918–4,535.5] in plasma, respectively) were significantly higher (rank ANOVA; P < .001) when compared with the survivor group (3,551.5 ng/mL [847–14,610] in peritoneal fluid, and 544.75 ng/mL [168.5–1,642.5] in plasma, respectively). Plasma d-dimer concentration also was significantly higher (rank ANOVA; P= .019) in the nonsurvivor group when compared with the group with financial restraints (521 ng/mL [305.5–2,091.5]).
Relationship Among d-Dimer Concentration and t-PA and PAI-1 Activities in Peritoneal Fluid
Peritoneal t-PA and PAI-1 activities in control horses were 0.4 IU/mL (0.27–0.45) and 0.0 IU/mL (0.0–2.37), respectively; whereas peritoneal t-PA and PAI-1 activities in colic horses were 0.5 IU/mL (0.23–1.33) and 0.0 IU/mL (0.0–1.63), respectively. Although t-PA activity was measured in 92.3% control horses and 96.4% colic horses, PAI-1 activity only was detected in 23.1% control horses and 34.3% colic horses.
The PCA with rank approach of d-dimer concentration, t-PA, and PAI-1 activities in peritoneal fluid showed that there was dependence among the 3 variables (P= .001). To address this codependence, a new variable was constructed after the results of the component matrix:
This new variable explained that d-dimer and t-PA worked opposite of each other, whereas PAI-1 had similar behavior to d-dimer. This means that horses with a high d-dimer concentration in peritoneal fluid also had decreased t-PA activity and mildly increased PAI-1 activity, which is consistent with an increase of peritoneal fibrinolysis activity and t-PA consumption.
In a 2nd step, a 1-way ANOVA was performed with this new combined variable for diagnosis groups (data not shown). Two possible groups were found: one for controls, obstructive, and enteritis, and another for enteritis, ischemia, and peritonitis. In fact, this approach indicates that in the future a 1st screening for diagnosis is possible with external validation and a quantitative approach to the results.
The results of the present study show that horses with severe GI disorders suffer an important increase in peritoneal d-dimer concentration, thus representing a strong activation of the peritoneal fibrinolysis activity as a consequence of increased peritoneal fibrin formation.
Because d-dimers are fibrin degradation fragments released exclusively by plasmin-mediated lysis of cross-linked fibrin,20 increases in d-dimer concentration in peritoneal fluid should be a specific indicator of an active peritoneal fibrinolysis. Moreover, the increase in peritoneal d-dimer concentration together with the decrease in peritoneal t-PA activity observed in this study is consistent with peritoneal hyperfibrinolysis and t-PA consumption. However, studies in humans also suggested that a decrease in peritoneal t-PA was associated with inhibition of peritoneal fibrinolysis activity. For example, several studies in humans with peritonitis observed that these patients had higher intraperitoneal production of PAI-1 and decreased t-PA activity during the disease.8–11,13–16 Changes in these markers were associated with decreased peritoneal fibrinolytic activity based on the evidence that cultured human peritoneal mesothelial cells under basal conditions and after exposure to tumor necrosis factor α (TNFα), interleukin-1α, or bacterial lipopolysaccharide, decreased their fibrinolytic activity by decreasing t-PA production and increasing PAI-1 synthesis. The addition of TNFα resulted in activation of the coagulation cascade by the expression of tissue factor.14 Similarly, the exposure of equine peritoneal macrophages to endotoxin resulted in a significant increase in TNFα, tissue factor, and PAI-2 activity in vitro.38 In contrast, the decrease in t-PA activity (and mildly increased PAI-1 in those horses in which it was detected) observed in the present study was consistent with t-PA consumption because of a hyperfibrinolysis. This conclusion was drawn because it was measured together with a sensitive fibrinolysis degradation product, d-dimer, and it was markedly increased. In fact, the few studies conducted in humans measuring peritoneal d-dimer concentration in different abdominal diseases showed that an increase of this parameter was associated with an increase in peritoneal fibrinolysis activity, and a decrease in d-dimer was related to hypofibrinolysis.16,39 In addition, Collatos et al19 observed an increase in peritoneal fibrinolysis activity in horses with severe colic based on an increase in peritoneal FDPs. Thus, comparing results observed in humans, in horses in other studies, and in the present study, it can be concluded that caution must be taken when interpreting changes in fibrinolytic activity based only on t-PA and PAI-1 activities. Including a sensitive product of fibrinolysis activity, such as d-dimer, allows a more accurate interpretation.
Intraperitoneal fibrin formation represents a physiological response of the mesothelial cells to visceral trauma in order to produce a matrix covering the damaged cells before cellular regeneration occurs.1–3 In horses suffering from GI disorders, diseases producing more intestinal wall lesions, like enteritis, ischemic disorders, and peritonitis, are expected to induce greater increases in fibrin formation, thus increasing the risk of intra-abdominal adhesion formation.4,5,19,40 This suspicion correlated well with the results observed in this study, because peritoneal d-dimer concentration was significantly higher in all colic groups compared with control, and significantly higher in horses suffering severe GI diseases, such as enteritis, ischemic disorders, and peritonitis, compared with horses with large intestinal obstructions. These results are in agreement with studies published by Collatos et al,18,19 which demonstrated that horses with colic are prone to an increase of peritoneal fibrinolysis activity, especially those disorders with higher levels of endotoxin in peritoneal fluid, such as inflammatory diseases, strangulating lesions, and visceral ruptures. However, no statistical differences were found in peritoneal d-dimer concentration in the present study among horses having enteritis, peritonitis, and ischemic disorders. This may be because of time before referral to the hospital and the severity of the lesion, which are important bias factors that contribute to the degree of alteration of the peritoneal fluid, and the wide range of d-dimer concentrations observed in this study. However, this can be expected in a prospective clinical study, where horses were grouped by diagnosis, with the disadvantage of being forced to include patients with different clinical situations and lesion severities in the same group.
Severe GI disorders typically are characterized by altered peritoneal fluid analysis that, depending on severity, sampling time, and chronicity of the disease, varies between a modified transudate and an exudate.41,42 This fact is in accordance with our original hypothesis and it has been demonstrated in this study because higher d-dimer concentrations were observed in colic horses with altered peritoneal fluid analysis (modified transudate and exudate) compared with those having normal peritoneal fluid. In addition, the Fisher exact test confirmed that a significant relationship exists between the type of peritoneal fluid and the diagnosis, which can explain the results obtained. Normal peritoneal fluid analysis typically is found in horses with noninflammatory/nonischemic obstructive diseases, whereas severe GI disorders are associated with marked alterations of peritoneal fluid.
When the prognostic value of d-dimers was analyzed, peritoneal d-dimer concentrations were significantly higher in nonsurvivors. There are no previous reports about the relationship of peritoneal d-dimers with outcome in horses with colic, but it is expected that events occurring in the peritoneal cavity should be well correlated to alterations in the peritoneal fluid. Thus, the more severe GI lesion diagnosed, the poorer the prognosis.
In this study, peritoneal fluid of healthy horses was found to contain small amounts of d-dimers compared with concentrations on plasma; whereas in horses suffering from GI disorders, peritoneal d-dimer concentrations greatly exceeded the corresponding plasma concentrations. Several studies have shown that different concentrations of coagulation and fibrinolysis proteins are found when comparing those results in peripheral blood and in peritoneal fluid, suggesting that these peritoneal proteins are released from the peritoneum and do not derive from leakage of plasma.15,16,18,19 Furthermore, peritoneal d-dimer concentration is not influenced by plasma concentrations, and increases of d-dimer concentration in peritoneal fluid are not correlated with what is occurring in plasma, as previously cited19 and confirmed in our study. In addition, peritoneal d-dimer concentration is not affected by blood contamination that may occur during peritoneal fluid sampling, except when peritoneal fluid is highly contaminated (≥20%).j
When analyzing data of plasma d-dimer concentrations obtained in this study by diagnosis, results were similar to previous studies.20,24,25,27 However, horses with peritonitis had the highest plasma d-dimer concentrations, which were significantly higher when compared with all the other colic groups. This suggested that these patients had the most severe hypercoagulable state, associated DIC, and the poorest prognosis. Plasma d-dimer concentrations also were increased with altered peritoneal fluid results. Horses having a peritoneal exudate also showed the highest plasma d-dimer concentrations; but these results could be influenced by the diagnosis of the GI disorder. Finally, when the outcome was considered, plasma d-dimer concentrations were found to be significantly higher in nonsurvivors, although these results also could be influenced by the diagnosis of the GI problem. Sandholm et al22 also concluded that plasma d-dimer concentrations were valuable as a predictor for outcome in equine GI colic cases. However, other studies have shown that plasma d-dimer concentrations were not correlated with the likelihood of euthanasia.20,25,27
Thus, findings of plasma d-dimer concentrations observed in the present study are in agreement with previous studies, which demonstrated that a systemic coagulopathy consistent with a marked hypercoagulable state frequently can be found in horses with severe colic. This systemic hypercoagulable state can lead to serious complications, such as DIC, thrombophlebitis, and pulmonary thromboembolism, especially in horses with peritonitis, enteritis, and ischemic disorders, as well as in nonsurvivors.19,43–50
In the present study, predictive values, accuracy, sensitivity, and specificity of d-dimer concentration in peritoneal fluid were not determined because of the overlap of values between groups and the difficulty for defining a cut-point value. However, the rapid and cost-affordable semiquantitative latex agglutination test might be used in emergency settings to provide further information for the diagnosis of acute GI disorders.
In conclusion, results of this study confirm that d-dimer concentration is an useful variable to assess fibrinolytic activity in peritoneal fluid, and that horses with severe GI disorders have marked hyperfibrinolysis related to increases in peritoneal fibrin formation and fibrin degradation activities. Thus, determination of this variable could provide further information for the diagnosis of different GI disorders, particularly to differentiate severe disorders from obstructive lesions when very high peritoneal d-dimer concentrations are found. Moreover, the test also may be useful to monitor adhesion risks. Further studies should be conducted to assess changes in fibrinolysis activity by measuring peritoneal d-dimer concentration during progression of severe GI disorders (especially those with high risk for abdominal adhesions such as ischemic lesions, enteritis, and peritonitis).
aArmengou L, Monreal L, Segura D, Navarro M. Plasma d-dimers in horses with colic. Proceedings of the 8th International Equine Colic Research Symposium. Quebec, Canada; 2005:1 (abstract)
bCesarini C, Monreal L, Segura D, et al. Hemostatic follow up of horses with medical and surgical colic. J Vet Intern Med 2009;23:434 (abstract)
jDelgado MA, Monreal L, Tarancón I, et al. Effects of blood contamination on peritoneal d-dimer concentration in horses with colic. Proceedings of the 9th International Equine Colic Research Symposium. Liverpool, UK; 2008:138-139 (abstract)
kHuman t-PA activity assay, Molecular Innovations, Southfield, MO