• Open Access

The Effects of Pentoxifylline on Equine Platelet Aggregation


  • This work was performed at the College of Veterinary Medicine, Cornell University, Ithaca, NY.

Corresponding author: Bruce Kornreich, T7012B Veterinary Research Tower, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853; e-mail: bgk2@cornell.edu.


Background: Pentoxifylline (PTX) possesses a number of vasomotor, immunomodulatory, and hemorheologic properties. Based upon the hypothesis that equine laminitis and navicular disease result from microthrombosis, the inhibitory effects of PTX on inflammatory cytokines, and its inhibitory effects on human platelet aggregation, PTX has been widely used to treat equine endotoxemia, navicular disease, and laminitis. Despite this, the effects of PTX on equine platelet aggregation have not been investigated previously.

Hypothesis: PTX decreases platelet aggregation in equine whole blood at concentrations approximating those achieved in horses given clinically relevant doses of PTX.

Animals: Seven healthy adult horses from a research herd.

Methods: Whole blood impedance aggregometry using whole equine blood incubated with varying concentrations of PTX. Adenosine diphosphate (ADP) and collagen were used to initiate aggregation.

Results: The onset time of collagen-induced equine platelet aggregation was significantly shortened by PTX. The maximum slope of resistance change (dR/dt) and total resistance change of collagen-induced platelet aggregation were unaffected by PTX. No effects of PTX on ADP-induced onset time of aggregation, dR/dt, or total resistance change were observed.

Conclusions and Clinical Importance: Our hypothesis is not supported by the results. PTX hastens the onset of collagen-induced platelet aggregation in equine whole blood, but has no effect on the rate of collagen-induced aggregation. PTX does not affect ADP-dependent equine platelet aggregation. Given these findings, PTX may not be a reasonable therapeutic option to decrease platelet aggregation in horses.


adenosine diphosphate








RBC deformability


tumor necrosis factor

Pentoxifylline (PTX) (1-(5-oxohexyl)-3,7-dimethylxanthine) is a xanthine derivative that possesses a number of vasomotor, immunomodulatory, and RBC membrane-altering properties that have been demonstrated in both in vivo and in vitro studies in horses, forming the basis for its use in equine clinical patients.1–6 PTX has been shown to be a potent phosphodiesterase inhibitor in in vitro studies of mouse macrophages and endothelial cells,7,8 and studies in humans, dogs, and cats have demonstrated its usefulness in the treatment of a number of disorders for which vasodilatation is a primary treatment goal.9–13 The in vitro immunomodulatory effects of PTX and those of its active metabolites 3,7-dimetyl-1(5´-hydroxyhexyl)xanthine (R-M1 and S-M1), 3,7-dimetyl -1(4-carboxybutyl)xanthine (M4), and 3,7-dimetyl -1(3-carboxypropyl)xanthine (M5), including attenuation of tumor necrosis factor (TNF) α, interleukin (IL)-12, and interferon (IFN) γ synthesis, and inhibition of T cell and NK cell cytotoxicity in a variety of species have sparked interest in the potential application of PTX in diseases characterized by inflammation.14–18 Some studies also have shown that PTX increases RBC deformability (RBCD) in normal human patients and that pretreatment with PTX can prevent the decrease in RBCD observed in blood circulated through bypass machines during open heart surgery.19 The results of these latter studies are controversial, however, in that other studies have shown no effect of PTX on human RBCD.20,21

PTX is commonly used in horses, primarily in the treatment of endotoxemia, navicular disease, and laminitis.22–25 These applications are based largely upon the known effects of PTX on inflammatory cytokines,15–17,26,27 and upon the fact that PTX may be an efficacious therapy for intermittent claudication in humans.28,29 Although the benefit demonstrated in humans with claudication prompted speculation that PTX may improve microcirculation in the equine foot, studies have shown that PTX has no effect on peripheral microcirculation in the digit or laminae of horses.23 Another rationale for the use of PTX in the treatment of navicular disease and laminitis in the horse is the fact that PTX has been shown to inhibit platelet aggregation in in vitro human whole blood studies and in in vivo studies in humans,30 and microthrombosis has been proposed as an etiology of equine navicular disease and laminitis.25,31 Despite this, the effect of PTX on platelet aggregation in horses is unknown.

Platelet aggregation is an important mechanism by which hemorrhage is minimized in mammals.32,33 Aggregation of platelets is initiated by damage to the endothelial lining of blood vessels, with subsequent exposure of a number of proteins below the endothelial cells.34 Collagen, the most thrombogenic of these proteins, is composed of a triple helix of peptide chains.34 The activation of platelets by collagen involves the interaction of specific collagen receptors on platelets, such as glycoprotein VI and α2β1, with collagen domains possessing specific tertiary and quaternary structural motifs.35,36 These platelet-reactive domains reside on the α 1 chains of both type I and type III collagen. Adenosine diphosphate (ADP), another well known activator of platelets, is stored in dense granules within platelets.37,38 ADP is released from the platelet after primary platelet activation by collagen and other proteins including thrombin. This secreted ADP then binds to P2Y1 and P2Y12 receptors on other platelets, activating them via several heterotrimeric G protein-mediated pathways that ultimately result in changes in cell shape, phospholipase C activity, and intracellular calcium homeostasis.39–41 These changes in platelet morphology and biochemistry ultimately promote the recruitment of additional platelets to the site of injury, resulting in the formation of a primary hemostatic thrombus.

Whole blood impedance aggregometry is a technique that utilizes changes in impedance to electrical current flow between 2 electrodes immersed in a sample of citrated whole blood to monitor platelet aggregation after the addition of an agonist that promotes platelet aggregation.42–44 This technique has been used to investigate the effects of a variety of compounds on platelet aggregation in a number of species.45 Given the widespread use of PTX in equine veterinary patients, the variable effects of PTX on platelet function in other species, and the lack of information regarding the effects of PTX on equine platelet aggregation, we designed a series of aggregometric experiments to test the hypothesis that PTX decreases platelet aggregation in equine whole blood at concentrations approximating serum concentrations observed in horses given clinically relevant PO and IV doses of PTX.

Materials and Methods

Horses were housed in an AALAC-approved facility and in accordance with protocols prescribed by the Institutional Animal Care and Use Committee. Subjects were healthy adult horses owned by Cornell University and housed in the equine research facility of the College of Veterinary Medicine, Cornell University. Two female warm bloods, 1 female thoroughbred, 2 thoroughbred geldings, 1 standardbred gelding, and 1 quarterhorse gelding comprised the study group (age range, 6–20 years). Subjects had not received any medications for a minimum of 1 week before study, were fed 4 kg of hay and 1.8 kg of grain daily, and were allowed access to water ad libitum.

Whole blood samples were obtained by venipuncture of either jugular vein with an 18-G hypodermic needlea attached to a 20-cm3 syringe,a immediately transferred to citrated vacuum containers (1 : 9 trisodium citrate : blood ratio)b and placed in a portable 37°C water bath for immediate transport to the laboratory. All blood samples were obtained between 9 and 11 am on the day of study to control for diurnal variation in platelet aggregation. Upon arrival in the laboratory, the blood samples were placed in a 37°C bench top water bathc and allowed to equilibrate. Aggregometry studies were started within 30 minutes of sampling and within 20 minutes of samples being placed in the bench top water bath.

Stock solutions of 50, 100, and 200 μg/mL PTX in sterile water were prepared by the dilution of a 40 mg/mL stock solution of PTX in sterile water.d For dose response studies in each subject, 5 μL of the 50, 100, or 200 μg/mL PTX stock solutions was added to 445 μL citrated equine whole blood in a 1.5 mL Eppendorf tubee to achieve final concentrations of 0.5, 1, or 2 μg/mL PTX, respectively. These concentrations were chosen to approximate serum concentrations of PTX demonstrated previously by either IV or PO administration to be clinically relevant doses of PTX for horses.46 For the control sample, 5 μL of sterile water alone was added to 445 μL citrated whole blood. The degree of dilution caused by this volume of sterile water in citrated whole blood (approximately 1%) is expected to result in a very minor alteration in osmolality that is within the physiologic range of variation in whole blood osmolality observed after exercise and rehydration in normal horses.47 After addition of either drug or water alone, the sample was inverted 3 times and placed in a 37°C water bath for 150 seconds. The sample then was inverted another 3 times and placed back into the water bath and incubated for an additional 150 seconds before aggregometry. Increasing and decreasing concentrations of PTX were used in serial aggregometry experiments in alternate subjects (ie, lowest to highest PTX concentrations in 1 horse, and then highest to lowest PTX concentrations in the next horse) to control for the effects on platelet aggregation of elapsed time during completion of studies on all PTX dilutions.

After equilibration with drug or water, 450 μL of the citrated whole blood sample was loaded into a 1 mL polystyrene cuvette,f a siliconized metallic stir barf was added, and the cuvette was placed into the test chamber of a model 500-V5 impedance aggregometer.f Four hundred and fifty microliters of heparinizedg normal Tyrode solution containing (in mM): NaCl 136, KCl 4, NaH2PO4-H2O 0.9, MgCl2-6H2O 0.5, CaCl2-2H2O 2, Glucose 5.5, and NaHCO3 24 then was added to the cuvette and the sample was stirred at 1,000 revolutions per minute. Tyrode solution of similar composition has been used previously to store platelets and study platelet function. The sensing electrode then was placed in the sample and aggregometry was performed. Recordings were obtained at a paper speed of 2 cm/min and were calibrated to a 20 Ω calibration pulse. Samples were stirred until a stable baseline was obtained for each sample before addition of either 10 μM ADPf or 10 μg/mL collagenf to initiate platelet aggregation. These concentrations of ADP and collagen have been used previously in whole blood impedance aggregometry in a number of species,45 and separate dose response experiments with 5, 10, and 20 μM ADP and 5, 10, and 20 mg/mL collagen, respectively, added to blood collected from healthy horses suggested that the aggregation response was saturated at 10 μg/mL collagen and 10 μM ADP, respectively (data not shown). Onset time was defined as the time from the addition of the aggregation-inducing agonist (ADP or collagen) to the point at which a slope >0 Ω/s was inscribed. Maximum slope (dR/dt; Ω/min) was defined as the slope of a tangent drawn to the steepest portion of the aggregometry tracing. The total resistance change was defined as the change in resistance (Ω) from the time of addition of aggregation-inducing agonist to the 6-minute point of the study.

Statistical Analysis

Onset time, maximum slope, and total resistance change after addition of ADP or collagen (separate, parallel experiments) were measured in blood obtained from each subject after incubation with either water (control), 0.5 μg/mL PTX, 1 μg/mL PTX, or 2 μg/mL PTX. The values of each of these continuous outcomes for each treatment condition (control or concentration of PTX) were compared nonparametrically by the Wilcoxon signed rank test contained in the Statistix 8 software packageh because they did not follow a Gaussian distribution. Descriptive data are presented as median and 1st and 3rd interquartiles. Statistical comparison was performed for both ADP- and collagen-dependent platelet aggregation. A P value of <.05 was considered significant.


Collagen induced platelet aggregation in whole blood from all subjects, whereas 1 subject showed a lack of ADP-induced platelet aggregation. No difference between the onset time (in seconds) of ADP-induced platelet aggregation after incubation with either 0.5 μg/mL (60, 30–136.5, P= .11), 1 μg/mL (76, 39–114, P= .22), or 2 μg/mL (69, 33–144, P= .38) PTX was observed when compared with control onset time (92 seconds, 60–123). The onset time of collagen-induced platelet aggregation after incubation with 0.5 μg/mL PTX (42, 36–60) was shorter than control onset time, but this difference was not significant (54, 48–61, P= .08). The onset time for collagen-induced aggregation after incubation with both 1 and 2 μg/mL PTX was shorter than control collagen-induced onset time (42, 36–54, P= .03; 47, 36−62, P= .04 for 1 and 2 μg/mL PTX, respectively). Examples of aggregometry experiments are shown in Figure 1, and onset time data are summarized in Figure 2.

Figure 1.

 Whole blood aggregograms with 10 μM adenosine diphosphate (ADP) (A) and 10 μg/mL collagen (B) to initiate platelet aggregation following incubation with either water alone (control) or water with varying concentrations of pentoxifylline. Note dose-dependent decrease in onset time for collagen induced aggregation.

Figure 2.

 Box and whisker plots of onset time of platelet aggregation induced by adenosine diphosphate (ADP) (n = 6) (A) and by collagen (n = 7) (B) in equine whole blood following incubation with water alone (control) and with varying concentrations of pentoxifylline. The lower, middle, and upper lines of each box represent the 1st, 2nd, and 3rd quartiles, respectively. The whiskers, where present, delineate the range. Where no whiskers are present, the 1st/3rd quartile represents the lower/upper limit, respectively. *Indicates significantly different from control.

No difference between the dR/dt (Ω/s) induced by ADP-dependent platelet aggregation was seen after incubation of blood with 0.5 μg/mL (0.9, 0.45–1.43, P= .50), 1 μg/mL (0.75, 0.5–1.2, P= .44), or 2 μg/mL (0.75, 0.38–1.5, P= .42) PTX compared with control ADP-induced dR/dt (0.9, 0.5–1.13). dR/dt similarly was not affected when evaluating collagen-induced platelet aggregation after incubation with either 0.5 μg/mL (3.5, 2.0–6.0, P= .47), 1 μg/mL (3.8, 3.0–6.0, P= .34), or 2 μg/mL (4, 3.0–5.8, P= .53) PTX compared with control collagen-induced dR/dt (4.5, 3.3–5.0). dR/dt data are summarized in Figure 3.

Figure 3.

 Box and whisker plots of the maximum slope of platelet aggregation (measured as change in impedance [Ω] per unit time [seconds]) induced by adenosine diphosphate (ADP) (n = 6) (A) and by collagen (n = 7) (B) in equine whole blood following incubation with water alone (control) and with varying concentrations of pentoxifylline. See Figure 2 legend for plot specifics.

The total resistance change for ADP-induced platelet aggregation (measured in Ω/6 min) was not affected by incubation with either 0.5 μg/mL (7.8, 2.6–11.3, P= .22), 1 μg/mL (6.5, 3.6–8.5, P= .50), or 2 μg/mL (8.5, 1.6–14, P= .16) PTX when compared with control (6.5, 3.1–8.8). Although the total resistance change (measured in Ω) for collagen-induced platelet aggregation tended to be higher than that observed with ADP-induced aggregation, neither the 0.5 μg/mL PTX group (31, 16–42, P= .29), the 1 μg/mL PTX group (34, 23–48, P= .47), nor the 2 μg/mL PTX group (32, 23–47, P= .47) was different from the control collagen-induced total resistance change (35, 28.3–40). Total resistance change data are summarized in Figure 4.

Figure 4.

 Box and whisker plots of total resistance change (measured as change in impedance [Ω] over the 6 minute assay interval) resulting from adenosine diphosphate (ADP) induced platelet aggregation (n = 6) (A) and by collagen (n = 7) (B) following incubation with water alone (control) and with varying concentrations of pentoxifylline. See Figure 2 legend for plot specifics.


The results do not support our hypothesis, and show that PTX decreases the onset time of collagen-induced platelet aggregation in equine whole blood while having no effect on the maximum rate of collagen-induced platelet aggregation or on ADP-dependent onset time or maximum rate of platelet aggregation. The fact that 1 subject did not demonstrate ADP-dependent platelet aggregation is not unprecedented, because substantial variation in platelet response to ADP, believed to be mediated at least in part by polymorphisms or mutations in platelet ADP receptors or both, previously has been shown in humans.30,48 These results have potential clinical implications. If the treatment goal in an equine patient is to decrease platelet aggregation, our results suggest that PTX may not be a reasonable therapeutic choice, and that it may, in fact, promote the initiation of platelet aggregation in horses. PTX may increase the likelihood of thrombosis after the exposure of platelets to a given amount of subendothelial collagen after vascular injury. If so, administration of PTX to certain equine patients may be contraindicated, including surgical candidates and those with acute injuries involving vascular trauma. As microthrombus formation has not been verified as the etiology of equine laminitis, the clinical relevance of our findings with respect to the use of PTX in laminitis patients is difficult to ascertain, although increasing the likelihood of thrombosis in a patient with distal limb pathology is intuitively contraindicated. Potentially salutary immunologic effects of PTX, such as its inhibitory effect on inflammatory cytokines, may however outweigh the potentially negative proaggregatory effect of PTX on platelet function in horses. Additional studies are required to more specifically identify subsets of horses in which PTX administration may be of benefit or in which it may be contraindicated.

In attempting to reconcile the finding of a PTX-induced decrease in onset time of collagen-induced platelet aggregation with no effect on dR/dt, we theorize that onset time is a measure of the initiation of aggregation by the binding of glycoprotein VI and α2β1 to platelet reactive domains of collagen, whereas dR/dt evaluates the ADP/thromboxane dependent recruitment of platelets. If so, 1 hypothesis to explain our results is that PTX increases the binding affinity of glycoprotein VI or α2β1 or both for the platelet-reactive domains found on collagen while having no effect on the affinity of ADP for P2Y1 or P2Y12 receptors or both on equine platelets. If this is true, aggregation onset time would be decreased because of an increased rate of binding of glycoprotein VI and α2β1 with platelet reactive domains, but the rate of platelet aggregation after this primary activation event may be unchanged because the local serum concentration of ADP achieved by the exocytosis of a relatively small number of dense granules exceeds that required to fully saturate P2Y1 and P2Y12 receptors on platelets. Further aggregometry studies using specific agonists and antagonists for each receptor subtype would be required to test this hypothesis.

The mechanism and relevance of the disparity in the effect of PTX on platelet aggregation in horses compared with its effect on human platelet aggregation remains to be determined. Given the fact that platelet aggregation is initiated and amplified by receptor-ligand interactions, it is possible that PTX affects structural interactions between collagen and glycoprotein VI or α2β1 receptors or both or between ADP and P2Y1 or P2Y12 receptors or both differently in horses than in humans by either differential interaction with 1 or more of these receptor-ligand binding sites, or by a more nonspecific mechanism that affects these receptor-ligand systems differently in different species (ie, allosterism). Additional studies are required to determine the basis for the different interspecies effects of PTX on platelet aggregation.

This study has a number of limitations. One limitation is that the effects of the active metabolites of PTX were not investigated in our in vitro studies. These metabolites have been shown to affect platelet aggregation in other species, and they may have clinically important effects on equine platelet function. Equine platelets respond to a number of agonists, including collagen, ADP, and platelet activating factor (PAF). The effects of PTX on PAF-induced platelet aggregation may differ from those identified for collagen and ADP in our studies. An additional limitation is that although whole blood impedance aggregometry is a useful technique for investigation of platelet aggregation, other techniques are available. Flow cytometry, for example, is a highly sensitive means of evaluating platelet aggregation, and follow-up studies by this technique may provide valuable insight into the effects of PTX on equine platelet aggregation. With respect to samples used for analysis of platelet function, some investigators believe that aggregometry by platelet-rich plasma (PRP) is a more sensitive assay of platelet aggregation than whole blood aggregometry, whereas others believe that whole blood aggregometry may be more physiologically relevant. Aside from the fact that the response of platelets under various conditions may be dependent upon interactions with RBCs and leukocytes that are not present in PRP, the effects of PRP preparation (ie, centrifugation) on platelet function are not completely understood, and PRP preparation may alter platelet function.49,50 Despite these potential limitations of aggregometry by PRP, follow-up studies investigating the effects of PTX on equine platelets in PRP may supplement the findings of our study. Another limitation is that our study does not provide definitive information regarding the specific mechanism of the PTX effect on collagen-dependent aggregation onset time. Determination of this mechanism will require additional biochemical, binding, and spectroscopic studies. Finally, although our experiments were designed to measure the effects of PTX on the rate of equine platelet aggregation, additional studies by different techniques such as thromboelastography may provide important supplemental information regarding the effects of PTX on other aspects of thrombus formation, such as clot strength.


aKendall Healthcare Monoject, Mansfield, MA

bBecton Dickinson and Co, Franklin Lakes, NJ

cVWR Scientific Inc, Westchester, PA

dProfessional Compounding Centers of America, Houston, TX

eEppendorf Inc, Westbury, NY

fChrono-Log Corp, Havertown, PA

g2 units/mL; Sigma Aldrich, St Louis, MO

hTallahassee, FL


This study was not supported by grant funds.

The authors thank Dr Stephen Barr and Ms Karen Warner for helpful discussions and technical assistance, and Ms Paula Sharp for assistance with manuscript preparation.