Effects of intravenous aminocaproic acid on exercise-induced pulmonary haemorrhage (EIPH)


email: bbuchholz@vetmed.wsu.edu


Reasons for performing study: The antifibrinolytic, 6-aminohexanoic acid, also named aminocaproic acid (ACA), has been used empirically as a treatment for exercise-induced pulmonary haemorrhage (EIPH) on the unsubstantiated basis that transient coagulation dysfunction may contribute to its development.

Objective: To assess the effect of ACA on bronchoalveolar lavage fluid (BALF) erythrocyte counts in horses performing treadmill exercise at an intensity greater than that needed to reach maximal oxygen consumption.

Methods: Eight Thoroughbreds were exercised to fatigue 3 times on a 10% inclined treadmill at a speed for which the calculated oxygen requirement was 1.15 timesinline image. Horses were treated with a saline placebo, 2 and 7 g ACA i.v. 4 h before exercise, with a crossover design being used to determine the order of the injections. Exercise-induced pulmonary haemorrhage severity was quantified via the erythrocyte count in BALF. Bronchoalveolar lavage fluid was collected 4 h before and 30–60 min post exercise. Results were expressed as mean ± s.e.m. and analysed by one way repeated measures ANOVA (P<0.05).

Results: Aminocaproic acid administration had no effect on any measured variables (inline image = 148 ± 3.0 [C]; 148 ± 3.0 [2 g ACA]; 145 ± 3.0 [7 g ACA] ml/kg bwt/min, respectively; run time = 77 ± 3 [C]; 75 ± 2 [2 g ACA]; 79 ± 3 [7 g ACA] seconds, respectively). All horses developed EIPH: 1691 ± 690 vs. 9637 ± 3923 (C); 2149 ± 935 vs. 3378 ± 893 (2 g ACA); 1058 ± 340 vs. 4533 ± 791 (7 g ACA) erythrocytes/µl pre- vs. post exercise recovered in BALF, respectively.

Conclusion: Aminocaproic acid was not effective in preventing or reducing the severity of EIPH or improving performance under the exercise conditions of this study.


It appears safe to assume that exercise-induced pulmonary haemorrhage (EIPH) occurs at least intermittently in the majority, if not all, Thoroughbred racehorses (Pascoe et al. 1981; Raphel and Soma 1982; Mason et al. 1983; Sweeney and Soma 1983; Whitwell and Greet 1984; Burrell 1985). Not only is EIPH prevalent in Thoroughbreds, it is also common in racing Quarter Horses and Standardbreds (Hillidge et al. 1984; Birks et al. 2002). Frusemide and Flair™ nasal strip are currently being used to treat EIPH, based on evidence that they decrease but do not prevent EIPH (Sweeney et al. 1984a,b; Geor et al. 2001; Kindig et al. 2001; Zawadzkas et al. 2006; Epp et al. 2009; Hinchcliff et al. 2009). Prerace medication with frusemide is also controversial because it has been shown to positively affect athletic performance as reflected by racing times, mass-specific maximal oxygen consumption and increased time to fatigue during high intensity treadmill exercise (Sweeney et al. 1990; Bayly et al. 1999; Gross et al. 1999; Zawadzkas et al. 2006).

The aetiology of EIPH remains unclear, but the prevailing hypothesis is that EIPH is induced by the combined effects of high pulmonary capillary blood pressure and very negative peak inspiratory pressures during intense exercise that together cause temporary stress failure of pulmonary capillary walls (Manohar et al. 1993; West et al. 1993; West and Mathieu-Costello 1995; Birks et al. 1997). Multiple occurrences of EIPH result in angiogenesis and fibrosis of the lung that ultimately leads to reduced elasticity of the lung and decreased capacity for oxygen uptake (McKane and Slocombe 2002; Williams et al. 2008; Derksen et al. 2009). Post mortem histological studies of affected areas of lungs of horses that had repeated epistaxis, or were known by endoscopic examination to have numerous episodes of EIPH, revealed increased vascularity of bronchial origin. It was hypothesised that these lesions could cause EIPH to occur more frequently if episodes of strenuous exercise were to continue (O'Callaghan et al. 1987). Since the respiratory system has been shown to be the limiting factor in the exercise tolerance of equine athletes (Ainsworth 2008), vascular propagation in the lung and morphological changes to the blood-gas barrier due to inflammation in response to repeated episodes of EIPH could cause poor performance and early retirement for competing horses (O'Callaghan et al. 1987; Derksen et al. 2009).

The clotting of blood is an essential part of tissue repair when an injury occurs to a vessel wall. Clotting is a complex process involving a balance between pro- and anticoagulant proteins. Coagulation culminates in the formation of fibrin monomers which together form the fibrin filaments needed to initiate repair of the damaged vessel (Davie and Ratnoff 1964; Esmon 1989; Mosesson et al. 2001; Morris 2006). The most critical anticoagulants are antithrombin and plasminogen. Antithrombin prevents the conversion of fibrinogen into its active form, fibrin, and plasminogen in its active form (plasmin), degrades fibrin filaments. Plasminogen is activated into plasmin through the binding of several different enzymes such as the activation enzymes, urokinase plasminogen activator and tissue plasminogen activating factor, found in body fluids (Alkjaersig et al. 1959; Suenson et al. 1984; Vassalli et al. 1991). The drug under examination in this study, 6-aminohexanoic acid, more commonly known as aminocaproic acid (ACA), is an antifibrinolytic that prevents clot lysis by inhibiting the activation of plasminogen but does not affect plasmin activity (Alkjaersig et al. 1959). Aminocaproic acid has also been shown to alter clot stasis by inhibiting plasminogen from binding to fibrin (Thorsen 1975).

Although they have never been demonstrated, exercise-associated clotting defects continue to be proposed as a cause of EIPH and treatments aimed at altering various aspects of coagulation and fibrinolysis are still common treatments for the condition. Our objective was to examine the effects of ACA on EIPH under intense but controlled treadmill exercise conditions by quantifying the number of red blood cells present in bronchoalveolar lavage fluid (BALF) collected soon after exercise.

Materials and methods

This study was approved by the Institutional Animal Care and Use Committee at Washington State University.


Eight fit and sound Thoroughbred horses (6 geldings and 2 mares; age 8–13 years; mean weight 497 kg ± 5.4 [s.e.m.]) were used. Each of the horses had a carotid artery translocation surgery performed to elevate their left carotid artery to a subcutaneous position along the jugular groove to facilitate arterial sampling from a catheter during exercise (Tavernor 1969).

Exercise protocol

One week before each of the experimental exercise trials, each horse performed an incremental exercise test in which it initially warmed-up at 4 m/s for 4 min. The speed was increased to 6 m/s for 1 min and subsequently quickened by 1 m/s every 60 s until the horse could not keep pace with the treadmill. The regression equation derived from the linear portion of the inline image-speed relationship for each horse was used to calculate individual speeds to be used with an oxygen requirement that was 1.15 times each inline image(Rose et al. 1990). These were the speeds at which the experiment was subsequently conducted.

During their experimental exercise trials, the horses warmed-up for 4 min at 40% inline imageon a 10% inclined treadmill and then ran until they could no longer keep pace with the treadmill at the predetermined speeds for which the calculated oxygen requirement was 1.15 times inline image. This protocol was followed 3 times for each horse with at least 2 weeks elapsing between each run. Each horse was treated either with a saline placebo, 2 or 7 g ACA1 i.v. 4 h before each of the 3 test runs. A crossover design was used to determine the order of the injections. Personnel involved in implementation of the exercise test were blind to the treatment in each case.

Instrumentation and analysis

Prior to beginning each exercise test, an 18 gauge, 5 cm catheter was inserted into translocated left carotid artery with an extension set for blood collection during exercise. Blood samples were collected in heparinised syringes before exercise, every 15 s during exercise at the calculated test speed and immediately post exercise. Oxygen consumption was also recorded every 15 s (Rose et al. 1990; Bayly et al. 1999). Haematocrit (Hct) and temperature-corrected blood gases were measured on each sample using a blood gas analysing machine2. Blood temperatures were recorded via a thermistor tipped Swan-Ganz catheter3 passed into the right heart via the right jugular vein. Heart rate was recorded during the run using a cardiotachometer4. inline imageand run time to fatigue were recorded for each test.

The severity of EIPH was quantified via the erythrocyte count in BALF. This was collected 4 h before and 30–60 min post exercise. Endoscope5-guided bronchoalveolar lavage was performed on the first 6 horses to ensure that the same segment of lung was lavaged on each occasion. Endoscopy was not utilised with the last 2 horses as experience with the first 6 horses indicated that the endoscope passed to the same lung segment in the right caudodorsal region without the benefit of guidance or use of the endoscope controls and a previous study had made the same observation with blindly passed catheters (McKane and Rose 1993; Couëtil and Hinchcliff 2004). Prior to the bronchoalveolar lavage procedure, each horse was sedated with xylazine6 (0.5–0.6 mg/kg) i.v. The endoscope (outside diameter = 9.6 mm) or bronchoalveolar lavage (BAL) catheter (outside diameter = 8 mm)7 was passed via the right ventral nasal meatus and advanced until wedged in the bronchus. Due to slight differences in the cuffs of catheters used, the cuff was inflated with 6–8 ml of air depending on the catheter used. Three-hundred ml of buffered saline was instilled into the lung through the BAL catheter. The luminal volume of the lavage catheter was measured prior to performing the procedure. This volume was collected and discarded before collecting the BALF. Lavage fluid was immediately aspirated using several 60 ml syringes and then placed slowly into an Erlenmeyer flask to avoid lysing the red blood cells. Fluid was continually aspirated from the BAL catheter until negative pressure on the syringe was significant. Red blood cells were counted using a haemocytometer8, immediately following completion of the BAL. The amount of fluid retrieved from the lung following instillation was recorded and the number of erythrocytes in the BALF was expressed in red blood cells per µl.

The horses were regularly exercised on the treadmill between test runs and remained well conditioned during the entire experimental period.

Statistical analysis

All results were expressed as mean ± standard error of the mean (s.e.m.). BALF red blood cell number, inline imageand run time to fatigue were analysed for the effect of treatment using a one-way repeated measures analysis of variance. Blood gases were tested for effects of sampling time and treatment using a 2-way repeated measures analysis of variance. In all cases significance was set at P<0.05. Statistical analysis was performed with the software Sigma Stat9 (2.03 version).


There was no significant difference in pre-exercise erythrocyte counts from BALF between treatment groups (P = 0.50). The number of red blood cells counted post exercise was not significantly altered from the control values following either dose of ACA although it approached significance (P = 0.09; Fig 1). Horse No. 4 experienced much more severe EIPH than the other 7 horses during the control run (31.0 × 106/µl vs. 6.6 ± 2.4 × 106/µl). If the results for Horse No. 4 were omitted from the data set, the effects of ACA were clearly not significant (P = 0.24). Post exercise Hct (P = 0.20), blood pH (P = 0.97), PaCO2 (P = 0.15), and PaO2 (P = 0.47) were also not significantly altered by treatment with ACA (Table 1). Aminocaproic acid had no significant effect on mass-specific inline image(P = 0.36), run time to fatigue (P = 0.53), or total distance run on the treadmill (P = 0.56). The mean speeds at which the horses performed the 3 tests were 13.5 ± 0.53, 13.6 ± 0.42, and 13.4 ± 0.56 m/s for placebo, 2 g ACA, and 7 g ACA, respectively.

Figure 1.

Graphical representation of the mean numbers ± s.e.m. of pre- and post exercise red blood cell counts via BALF using a haemocytometer with each treatment of ACA. There was no significant difference in post exercise RBC counts between treatments (P = 0.09).

Table 1. Mean values of variables ± s.e.m. measured at rest (pre), during exercise (VO2max, time to fatigue, total distance run, arterial pH, PaO2, and PaCO2), and immediately post exercise (post) for each treatment of ACA
 Control2 g ACA7 g ACA
Pre RBC/µl1691 ± 6902149 ± 9351058 ± 340
Post RBC/µl9637 ± 39233378 ± 8934533 ± 791
Pre Hct (%)35 ± 238 ± 236 ± 1
Post Hct (%)53 ± 156 ± 155 ± 2
inline image(ml/min/kg)148 ± 3148 ± 3145 ± 3
Time to fatigue (s)77 ± 375 ± 279 ± 3
Total distance (m)1041 ± 361011 ± 221043 ± 30
Arterial pH7.2 ± 0.027.2 ± 0.017.2 ± 0.02
Arterial PO2 (mmHg)59 ± 360 ± 361 ± 2
Arterial PCO2 (mmHg)52 ± 253 ± 251 ± 1
Max heart rate (beats/min)205 ± 2.8206 ± 2.8205 ± 3.3


There is no evidence from this or the study by Epp et al. (2008) that treatment with ACA is able to prevent EIPH. This is not surprising given that ACA's known effects are solely on the inhibition of the fibrinolytic, plasmin and are not involved in the formation of a clot (Okamoto et al. 1997). Theoretically, ACA could reduce the amount of haemorrhage by maintaining clot stability, but previous studies have shown that horses undergoing maximal exercise already experience thrombocytosis (Bayly et al. 1983). Decreasing haemorrhage would reduce the inflammatory response caused by erythrocytes entering the pulmonary interstitium and small airways, and potentially promote pulmonary health. Based on the results of this study, it appears unlikely that such a benefit is to be derived from treatment with ACA. However, although highly unlikely, it cannot be completely discounted, because the results from Horse No. 4, which was the only horse to experience a severe bout of EIPH, changed the P value from 0.24–0.09 and prompt speculation that ACA might reduce the volume of haemorrhage in severely affected horses but not those that suffer only mild bouts of EIPH. Such conjecture must remain just that for now, as no runs were repeated with Horse No. 4 and no other horses in the study experienced the same degree of EIPH. It could be argued more strongly that, because EIPH is a highly variable phenomenon in terms of volume of haemorrhage, the impact of the results from Horse No. 4 were spurious in that they were due solely to the large volume of EIPH experienced by this horse during the control run.

Epp et al. (2008) analysed the effects of ACA on various coagulation variables such as prothrombin time, partial thromboplastin time, antithrombin III activity, and plasminogen percentage during maximal exercise using a 5 g dose and found no effects of the agent on any of these variables. The red blood cell counts in BALF were also not significantly altered after treatment with 5 g of ACA (Epp et al. 2008). This is consistent with a maximal and submaximal exercise studies that revealed no significant difference in measures of coagulation variables between resting and exercising horses (Bayly et al. 1983; McKeever et al. 1990).

Aminocaproic acid may not have affected post exercise erythrocyte concentration in BALF because the dosages chosen may have been too low. Heidmann et al. (2005) used dosages of 30 mg/kg bwt and 100 mg/kg bwt in healthy resting horses and ponies to determine the effects of ACA on different measures of haemostasis and found that the high dosage increased α2-antiplasmin activity significantly 1 h after administration. This suggested that ACA may be an effective antifibrinolytic at higher doses. We are unaware of any previous studies using these doses in exercising horses. A consequence of using high doses of ACA is the possibility of a hypercoagulable state in the horse that would result in formation of thrombi and subsequently decrease the size of the capillary lumen. A decrease in the size of the lumen would increase capillary transmural pressure, thereby increasing the severity of EIPH. This may not be relevant, however, as Heidmann et al. (2005) also showed that resting horses tolerate higher dosages of ACA without serious complications.

It is also possible that administration of ACA should have been closer to the time of exercise in order to have had an effect on post exercise BALF erythrocyte numbers. Ross et al. (2007) used a dosage of 3.5 mg/kg bwt/min for 20 min (total of 35 g for a 500 kg horse) and found an elimination half-life of 2.3 h and ACA concentration remained above the proposed therapeutic concentration (130 µg/ml) for only 1 h. Renal excretion of all quantifiable amounts of ACA administered i.v. or per os occurred within 6 h of administration (Heidmann et al. 2005), again suggesting that after 4 h the plasma concentration of ACA in horses used in this study may have been too low to alter haemostasis at the time of exercise. We elected to inject ACA 4 h before exercise because this is the time by which all medications must be administered in many racing jurisdictions that allow prerace treatments.

Aminocaproic acid administration had no significant effect on performance variables such as blood gases, inline image, the run time to fatigue or the total distance the horse ran during the test run. The latter finding was in contrast to that of Epp et al. (2008) who found that ACA actually decreased the time to fatigue. We cannot rule out that our findings and those of Epp et al. (2008) may have been as much a reflection of the volume of EIPH incurred as an indication that ACA treatment had no effect on performance. The volume of EIPH equated with acute reduction on performance is unknown in horses racing and training under field conditions. Controlled treadmill studies with small numbers of horses have shown that instillation of 100 ml of blood in each lung produces statistically significant reductions in performance, but that smaller volumes do not (Kingston et al. 2002; McKane et al. 2007). Although only the right lung was lavaged, the concentration of erythrocytes in recovered BALF in this study indicates that the volume of EIPH suffered by our horses was considerably less than 100 ml. It is widely recognised that EIPH occurs in both lungs and is volumetrically asymmetrical (O'Callaghan et al. 1987), with the right lung usually being the site of greatest haemorrhage. A recent review of EIPH literature (Couëtil and Hinchcliff 2004) indicates that virtually all studies that have utilised BAL have only collected fluid from the right lung for this reason. Consequently, although BAL was not performed on the left lung in our study, it is likely that the fluid collected from the right side was most representative of the severity of haemorrhage sustained by the horses.

In conclusion, administration of ACA did not affect any variables that would indicate a reduction in the severity of EIPH. Aminocaproic acid did have a profound effect on the BALF erythrocyte number of the only horse of those used for this study that experienced severe pulmonary haemorrhage, indicating that further studies with severe ‘bleeders’ may be warranted. Additional research on the effects of using higher dosages of ACA and administering ACA closer to the time of strenuous exercise may also be justified. Coagulopathies have never been linked to the development of EIPH. However, the possibility that they contribute to EIPH in a small number of horses has not been ruled out. That said, present results indicate that pre-exercise treatment with ACA has no effect on the volume of EIPH.


Support for this study was provided by the Washington State University College of Veterinary Medicine, Equine Research Program. The invaluable assistance of Cassie Dotts, Shane Smith, Danielle Thomas, Krista Morrow, Kristen Kline, Heidi Banse, Laura Bronsart and Kelci Porter is greatly appreciated.

Manufacturers' addresses

1 Hospira, Inc., Lake Forrest, Illinois, USA.

2 Instrumentation Laboratory, Orangeburg, New York, USA.

3 Edwards Lifesciences, Irvine California, USA.

4 Polar Electro Oy, Kempele, Finland.

5 Pentax Medical Company, Montvale, New Jersey, USA.

6 Akorn, Inc., Decatur, Illinois, USA.

7 Smiths Medical North America, Waukesha, Wisconsin, USA.

8 Hausser Scientific, Horsham, Pennsylvania, USA.

9 SPSS Inc, Chicago, Illinois, USA.