Materials and Methods

Animals—Horses were categorized on the basis of age (adult [≥ 5 years of age] or foal [≤ 6 months of age]) and the presence or absence of acute gastrointestinal disease (colic or noncolic). Foals were included in the noncolic group if they had been admitted to the teaching hospital for routine, simple herniorrhaphy and had no evidence of omphalophlebitis or other abnormalities. Foals were also included in this group if they were to be euthanized because of a noninflammatory, non-gastrointestinal disease process (eg, severe angular limb deformity or fracture) or for research purposes unrelated to the abdomen. Adult horses were included in the noncolic group if they were to be euthanized for research purposes unrelated to the gastrointestinal system. Foals and adult horses were included in the colic group if they were admitted to the hospital for colic and had abdominocentesis as part of their initial clinical examination or underwent exploratory laparotomy within 6 hours after admission. They were excluded if they had received antiplatelet (other than NSAIDs), thrombolytic, or antithrombotic medications or had been hospitalized at another clinic prior to admission to the teaching hospital. All research horses and foals used were systemically healthy and had not received medications other than α-2 receptor agonists for sedation in the 2 weeks prior to euthanasia. When research procedures had been performed, the procedures happened at least 8 weeks prior to study inclusion, and all were related to the musculoskeletal system.

Age, sex, problem, and diagnosis were recorded. For colic, type of treatment (medical or surgical), survival to discharge (yes or no), and characterization of colic (strangulating or nonstrangulating) were recorded. For horses with ischemia, diagnosis was made at laparotomy or necropsy. This study was performed in compliance with institutional guidelines for research on animals.

Sample collection and processing—Sample collection occurred from August 2006 to September 2009. Blood samples (3 mL) were collected via jugular venipuncture into evacuated tubes containing a one-tenth volume of 3.8% sodium citrate.^a Abdominal fluid was collected into similarly filled but open tubes by means of abdominocentesis or at laparotomy by use of flexible catheters to aspirate dorsally dependent fluid immediately upon abdominal entry. Blood and abdominal fluid samples were centrifuged at 4,200 × g for 5 minutes within 1 hour after collection. The supernatant plasma or abdominal fluid was immediately harvested and frozen at −80°C for batch analyses of coagulation parameters at 3-month intervals, so that samples were never stored for more than 3 months prior to analysis.

Laboratory assays—All assays were performed with an automated coagulation analyzer.^b Fibrinogen concentration was measured via the Clauss method by use of a human thrombin reagent.^c Pooled equine plasma (prepared from 12 healthy adult horses not included in this study) was used as the fibrinogen assay standard for all samples. This pooled plasma contained 300 mg of fibrinogen/dL as measured by means of a gravimetric method.²³ Plasminogen activity was measured by use of a chromogenic substrate kit.^d The plasminogen assay was modified to measure equine plasminogen activity by substitution of urokinase for streptokinase, activation, and acidification of the test and standard plasmas, as described.²⁴ Sample plasminogen activity was reported as a percentage of the pooled equine plasminogen activity, which had an assigned value of 100%. Antiplasmin activity was measured by use of a chromogenic substrate kit^e used in another study of fibrinolysis in horses.²⁵ Sample antiplasmin activity was reported as a percentage of the activity of the kit manufacturer's human standard.^f D-dimer concentration was measured via a quantitative, immunoturbidimetric method by use of a commercial kit^g as described elsewhere²⁶ and the manufacturer's human D-dimer calibration standards. The ratio of the plasma value to the abdominal fluid value was calculated to convey the difference between blood and abdominal values for each analyte.

Statistical analysis—Data were generally not normally distributed. Several transformations were attempted to achieve a normal distribution; however, none were successful. Therefore, nonparametric statistics were used, and an interaction of disease status and age could not be assessed. Analyte data are reported as median and quartiles. Potential differences attributable to the effect of disease status (noncolic or colic) and age (foal or adult) were evaluated with a Wilcoxon rank sum test. Differences in nominal data (yes vs no) were tested with a Fisher exact test. All tests were 2 tailed, and values of P < 0.05 were considered significant. Statistical testing was performed with commercially available software.^h

Results

Animals—Paired blood and abdominal fluid samples were collected from 25 foals (7 males and 18 females) and 20 adults (12 males and 8 females) without colic and 16 foals (5 males and 11 females) and 19 adults (10 males and 9 females) with colic. In the noncolic group, the median (interquartile range) age of foals was 143 (56 to 210) days, and that of adults was 6.5 (5 to 17) years. None of these horses had ischemic gastrointestinal disease. Reasons for foal inclusion were omphalectomy (n = 15), euthanasia for research (4), and euthanasia for angular limb deformity or fracture (3 each). All adults in the noncolic group had been used in other research and euthanized.

In the colic group, the median age of foals was 52.5 (30 to 146) days, and that of adults was 16 (9 to 18) years. The definitive cause of colic was determined by exploratory laparotomy (n = 10) or at necropsy evaluation (3). Only a presumptive diagnosis was available when medical treatment was attempted and successful (n = 3). Diagnoses made at exploratory laparotomy included enteritis (n = 1), nonreducible hernia (3), fecalith (2), meconium impaction (1), and volvulus nodosus (3). Presumptive diagnosis made for medically treated colic included ascarid impaction (2) and no diagnosis (1). Diagnoses made at postmortem examination included intra-abdominal adhesions (2) and atresia coli (1).

The definitive cause of colic in the adults was determined by exploratory laparotomy (n = 9) or at necropsy (4). Only a presumptive diagnosis was available when medical treatment was attempted and successful (n = 6). Diagnoses made at exploratory laparotomy included strangulating lipoma (n = 3), infarcted large colon (1), large colon volvulus (2), volvulus nodosus (1), nephrosplenic entrapment (1), and nonstrangulating lipoma (1). Strangulating lipoma was diagnosed in 4 horses via necropsy. Diagnoses made for horses treated medically included a presumptive diagnosis of anterior enteritis (n = 2), gas colic (2), and pelvic flexure impaction (2).

In the colic group, there was no difference between the proportion of foals (7/16) and the proportion of adults (10/19) that had strangulating disease (ischemia) identified at postmortem examination or at surgery (P = 0.738). Foals and adults also did not differ in the proportion that survived to discharge (11/16 and 11/19, respectively; P = 0.727). Of the horses with colic that did not undergo surgery and that did not survive to discharge, euthanasia was performed within 6 hours after evaluation or sample collection.

Values for abdominal fluid and plasma fibrinogen, plasminogen, antiplasmin, and D-dimer were summarized by age (foal vs adult) and disease status (colic vs no colic; Table 1). Colic in both age groups was typically associated with significant increases in values for all abdominal fluid analytes (disease status effect; Table 2). However, there were no differences in values of abdominal fluid analytes when age groups were compared within disease status (foals without colic vs adults without colic and foals with colic vs adults with colic; age effect). In plasma, there were no significant differences for foals (with colic vs without colic) for all analytes, and only D-dimer concentration (adults without colic had lower values than adults with colic; P < 0.001) and plasminogen activity (adults without colic had greater values than adults with colic; P = 0.004) were significantly different in adults (disease status effect). In comparisons across age groups, foals without colic had a significantly higher fibrinogen concentration than adults without colic and foals with colic had a significantly higher fibrinogen concentration and plasminogen activity than adults with colic (age effect).

Discussion

The hypothesis in the present study was that comparison of hemostatic balance between foals with colic and adults with colic would reveal conditions favorable to adhesion formation, such as relative increases in peritoneal fibrinogen concentration or antiplasmin activity or relative decreases in peritoneal D-dimer concentration or plasminogen activity. Although several age-dependent (foal vs adult) and disease status–dependent (noncolic vs colic) differences in plasma and abdominal fluid analytes were identified, none supported this hypothesis.

Intra-abdominal adhesion formation is a complex process initiated by abdominal injury secondary to inflammatory disease or surgery. Changes in vascular permeability at the sites of injury and an influx of inflammatory cells interact with mesothelial cells to produce a procoagulant environment that fosters fibrin deposition and fibroblast proliferation. The concomitant release of tPA from injured tissues generates the active serine protease plasmin, which under physiologic conditions lyses fibrin matrices and prevents the formation of permanent fibrous bands (adhesions). Plasminogen activator inhibitor is the major inhibitor of tPA and, along with other inflammatory cytokines, acts to retard plasmin-mediated fibrinolysis and promote the deposition of extracelluar matrix. The balance between tPA and PAI-1 is believed to play a pivotal role in the rate of fibrinolysis and ultimately pathological adhesion formation.^17,19,20,27

Studies²⁸ in horses reveal that, as in other species, tPA and PAI-1 are the major regulators of plasmin-mediated fibrinolysis and that experimentally induced endotoxemia and naturally occurring gastrointestinal disorders result in increases in plasma and abdominal fluid PAI-1 activity. Severe colic and low survival rates are associated with dysregulated fibrinolysis in horses. Evidence for dysregulated fibrinolysis in horses includes sustained high plasma PAI-1 and high abdominal fluid and plasma D-dimer concentration.^28,29 Whereas previous studies involved only adult horses, the purpose of the present study was to identify differences in fibrinolysis between adult horses and foals with and without colic at the time of evaluation that might foster adhesion formation in foals with colic. The analytes fibrinogen, plasminogen, antiplasmin, and D-dimer were selected on the basis of their role in fibrinolysis and the availability of validated equine assays for these analytes at the time of study inception. A study²⁶ of equine colic revealed that abdominal fluid D-dimer concentration is significantly correlated with increased abdominal fluid PAI-1 activity and decreased tPA activity. Our finding of high abdominal D-dimer concentration in horses with colic reflected similar changes in these plasminogen regulatory proteins.

We found no significant effects of age (foal vs adult) on any of the abdominal fluid parameters, regardless of disease status. Foals and adult horses with and without colic had apparently similar potential to produce adhesions within the abdomen because of similar clot substrate concentration (fibrinogen) and evidence of ongoing fibrinolysis within the abdomen manifested by similar plasminogen activities and D-dimer concentrations.

The effect of age on plasma variables revealed significantly higher fibrinogen concentrations in foals with and without colic than in adults with the same disease status. Although the difference in median concentrations for foals without colic versus adults without colic (404 vs 354 mg/dL) was not clinically relevant, the foals with colic had a higher median fibrinogen concentration (close to 200 mg/dL more) than adults with colic. Higher plasma fibrinogen concentration in the foals with colic may have been due to differences in the underlying disease or duration of illness prior to evaluation. Regardless, higher plasma fibrinogen concentration was not associated with higher abdominal fluid fibrinogen concentration in foals with colic, and we found no significant age effect on plasma-to-abdominal fluid ratios for fibrinogen concentration for either disease status. Importantly, the foals with colic had a marked increase in abdominal D-dimer concentration (compared with foals without colic) that was not significantly different from that of adult horses with colic. This finding suggests an active abdominal fibrinolytic process in foals with colic, with no apparent difference from that of adults with colic.

When the effect of disease status was evaluated, colic was associated with significant increases in nearly all analytes measured in the abdominal fluid, in adults and foals, in contrast to the plasma values. The increase in D-dimer concentration in the abdominal fluid of adults with colic in the study reported here was in agreement with previous findings.²⁶ Although the other analytes measured (antiplasmin, fibrinogen, and plasminogen) were not evaluated in that study,²⁶ concomitant increases in these analytes secondary to colic in the horses in our study were not surprising and confirm that perturbations of the hemostatic system exist in the abdomen of horses and foals with colic that are not reflected in the plasma compartment.^26,28,29 Fibrinogen and antiplasmin are synthesized by the liver, and the increased concentrations of these analytes within the abdomen of foals with colic and adults with colic result primarily from increased vascular permeability and vascular injury that occurs with abdominal disease. In contrast, plasminogen activity within the abdomen is primarily generated by abdominal mesothelial cells and activated monocytes. Similarly, high abdominal D-dimer reflects the in situ action of plasmin on cross-linked fibrin.^19,20,27

Analysis of the effect of disease status on plasma values revealed significant differences in adults only. Unlike foals, the adult horses with colic in our study had an increased D-dimer concentration and decreased plasminogen activity, compared with values in the same-aged noncolic group. In agreement with findings reported here, high plasma fibrin degradation products in horses with colic have been reported.²⁹ The simultaneous findings of low plasma plasminogen activity and high D-dimer concentration are suggestive of a thrombotic process that results in plasminogen depletion as fibrin is degraded within the microvasculature of adults with colic. The reason foals with colic in our study did not have similar evidence of high systemic fibrin deposition is unclear. Although foals with colic had higher median plasma D-dimer concentration than foals without colic, there was wide overlap in the range of data, unlike the adults. Colic severity may have differed for adults and foals, although the proportion of horses with ischemic disease was not different between the 2 age groups. It is possible the duration of illness or presence of endotoxemia (not evaluated in this study) varied between the groups.

Although none of the adults in the present study had adhesions causing colic, 2 foals with colic had adhesions that caused partial (n = 1) or complete (1) jejunal obstruction. The partial obstruction was in a 5.5-month-old foal that had an umbilical resection 2 months prior to its evaluation for colic. The resection had been routine; however, the foal developed an incisional seroma that later became infected and was treated medically within 3 weeks after resection with enterally administered antimicrobials. There were no signs of colic until the foal was evaluated at 5.5 months of age. The complete obstruction was in a 2-week-old foal that had mild colic in the first 1.5 days after birth caused by a presumptive meconium impaction, after which the foal was apparently healthy until evaluated for colic at 2 weeks of age. Although one may speculate that the presence of adhesions in 2 foals supports an increased risk of adhesion in foals with colic versus adults with colic, comparing the incidence of adhesion formation in foals with colic and adults with colic was not the purpose of this study. Both foals had mature adhesions that could not be bluntly reduced and there was no gross evidence of continued adhesion development from the original primary problem (infected umbilical resection site and meconium impaction), yet we considered that the presence of adhesions could have affected the hemostatic balance in these foals. However, censoring these foals during data analysis did not change the results of significance testing except for the fold change in the ratio of plasma fibrinogen to abdominal fluid fibrinogen when foals with colic were compared with foals without colic. In that comparison, the difference became nonsignificant (P = 0.06). Therefore, the presence of adhesions in 2 foals with colic did not change our conclusions.

The major limitations of this study include small numbers of horses, heterogeneous groups, and lack of serial testing to assess changes in hemostatic variables over time. Although the foals with colic did not differ from adults with colic with respect to the proportion of ischemic disease or survival to discharge from the hospital, the underlying diseases that caused colic in the foals varied widely. Additionally, foals up to 6 months of age were included in the study. If a difference exists in foals, it may be in a much younger age group; however, clinical experience suggests there is an increased risk of adhesions in foals up to 6 months of age versus adult horses. Furthermore, the age distribution was different within each age group; foals with colic were generally younger than foals without colic, and adults with colic were generally older than adults without colic. Finally, other analytes may better reflect differences if they indeed do exist.

In the present study, foals did not differ from adult horses in abdominal fluid D-dimer concentration and plasminogen activity, regardless of health status, suggesting that intra-abdominal fibrinolytic activity (plasminogen) or fibrinolytic activity (D-dimer concentration) did not differ on the basis of age. Similarly, there were no differences in abdominal fluid antiplasmin activity and fibrinogen concentration between the age groups. Additional investigation is needed to characterize differences in hemostatic balance that may exist in the postoperative period or later during medical management; such conditions were not assessed in this study. If foals do have an increased incidence of intra-abdominal adhesions relative to adult horses, although intra-abdominal hemostasis may be similarly regulated between the ages, the increased rate of healing in the young foal may outpace the fibrinolytic system, leading to a higher rate of intra-abdominal fibrous adhesion formation in foals.