Abstract
Although heparins are usually injected intravenously or subcutaneously, antithrombotic activity is observed in rat models following single oral heparin doses. Since repetitive dosing is usually needed for thromboprophylaxis, study objectives were to determine whether repetitive oral heparin prevented arterial thrombosis and to compare effectiveness to subcutaneous administration. Wistar rats were given subcutaneous or oral unfractionated heparin ([UFH] 1 mg/kg per 48 h), low-molecular-weight heparin ([LMWH] tinzaparin, 0.1 mg/kg per 12 h), or saline for 30 days. On the last day, thrombosis was initiated by placing 30% FeCl3-soaked filter paper on the distal carotid. Subsequent flow measurements, for a 60-minute period, included recorded time of initial thrombus formation (time till thrombus begins [TTB]), and time until carotid occlusion (time till occlusion [TTO]). The formed thrombus was dried and weighed. The activated partial thromboplastin time (aPTT), anti-factor Xa, and antithrombin activity were determined from the plasma. Both oral and subcutaneous heparins significantly increased TTB and TTO. Time of initial thrombus formations were 12.6 ± 1.1, 21.2 ± 2.2, 25.3 ± 3.9, 21.7 ± 3.1, and 21.3 ± 1.7 minutes and TTOs were 29.3 ± 3.6, 54.8 ± 4.0, 60.0 ± 0.3, 56.7 ± 3.3, and 58.3 ± 1.7 minutes (mean ± SEM) for control, subcutaneous UFH, oral UFH, subcutaneous LMWH, and oral LMWH, respectively. Thrombus weight was 2.52 ± 0.29 g in control and was reduced to 43%, 23%, 33%, and 28% of control weight for subcutaneous UFH, oral UFH, subcutaneous LMWH, and oral LMWH, respectively. Thrombus weight was significantly less for oral compared to subcutaneous UFH. The aPTT for oral UFH, and anti-factor Xa activity in the LMWH-treated groups were significantly greater than control (two-tailed t tests). These findings confirm that orally administered heparins are absorbed. Repeated treatment with oral heparin showed similar antithrombotic activity compared to subcutaneous heparin. Oral heparin use for arterial thromboprophylaxis should be further investigated.
Introduction
Heparins are extensively used antithrombotic drugs and are currently used for the treatment of cerebral, coronary, and retinal vessel thrombosis to prevent clot extension and embolus,1,2 in vascular surgery and orthopedic surgery, and for the treatment of acute coronary syndrome. 3 For the treatment and prevention of thromboembolism, unfractionated heparins (UFHs) are administered by the intravenous route while low-molecular-weight heparins (LMWHs) are primarily given subcutaneously. 4 Our laboratory and others have obtained considerable evidence that heparins are indeed effective when given orally.5–7 Potential advantages of orally administered heparin include a decrease in pain and bruising experienced by the patients associated with heparin injection, more convenience, and a reduction in health care costs.
In support of the effectiveness of orally administered heparin, we have observed antithrombotic activity in a rat jugular vein, 5 venous stasis, 8 and carotid arterial model, 6 following single-dose oral heparin administration. In the rat carotid arterial model, a single oral dose of UFH (7.5 mg/kg) or the LMWH tinzaparin (0.1 mg/kg), increased the time until the beginning of thrombus formation (time till thrombus begins [TTB]) and time to occlusion (time till occlusion [TTO]) following the application of FeCl3 to the carotid artery, as well as decreasing the thrombus weight. 6 Heparins are found associated with endothelium following oral administration, although there is little evidence of heparin in plasma. 7
Repetitive dosing is required for heparin use as a thromboprophylactic agent as well as for numerous proposed uses. In a previous subacute study, using a rat venous thrombosis model, we determined that a dose regimen of 1 mg/kg per 48 h of UFH or 0.1 mg/kg per 12 h of LMWH significantly decreased thrombotic incidence when given for 30 days prior to thrombus initiation. 5 It is unclear whether this oral dose would be effective in arterial thrombosis. Arterial thrombosis differs from venous thrombosis in the hemodynamics that create it and its subsequent constituents, such that the arterial thrombus has more platelets and less fibrin than a venous thrombus. 9 Our objectives were to determine whether repetitive oral dosing with UFH and LMWH was effective in preventing thrombosis in an arterial model. We also compared the antithrombotic effects following oral heparin administration to that following subcutaneous administration.
Methods and Materials
Animals
Thirty-five male Wistar rats weighing 184.5 ± 37.5 (SD) g were obtained from the Charles River Canada Company, St Constant, Quebec, Canada, or from Animal Resources Centre, University of Saskatchewan, Saskatoon, Canada. The rats were handled in accordance with the Principles of Animal Care set out by the Canadian Federation of Biological Societies. Experimental procedures were approved by the Animal Care Committee of the University of Saskatchewan. Rats were housed as 2 rats/cage, except on day 29 when the rats were kept in individual metabolic cages for 24 hours. Rats were anesthetized with equithesin (chloral hydrate 4.2% w/v, sodium pentobarbital 0.98% w/v, magnesium sulfate 2.12% w/v, ethanol 10% v/v, propylene glycol 40% v/v, sterile water to a volume of 100 mL:1 mL/250 g rat intraperitoneally [ip]) immediately prior to thrombus formation.
Drug Administration
The rats were divided into 5 experimental groups treated with saline (10 rats), subcutaneous UFH (6 rats), oral UFH (7 rats), subcutaneous LMWH (6 rats), and oral LMWH (6 rats). Unfractionated bovine lung heparin (157 IU/mg, Lot No. 901S1580 from Scientific Protein Laboratories Waunakee, Wisconsin) was dissolved in double-distilled water at 1 mg/mL and administered at a dosage of 1 mg/kg per 48 h for a 30-day period. The LMWH, sodium tinzaparin (anti-Xa activity, 90.7 IU/mg, peak maximum molecular mass 5 600 Da) was obtained from Novo Nordisk, Denmark. Tinzaparin was dissolved in double-distilled water (0.1 mg/mL) and given at a dosage of 0.1 mg/kg per 12 h for a 30-day period. For oral administration, heparins were introduced into the stomach by gastric lavage in a calculated volume, depending on rat weight, of approximately 0.1 to 0.4 mL followed by 0.2 mL of saline. Subcutaneous injections were given in the scruff using a 25-gauge needle. Control rats were given saline (0.3 mL) by gastric lavage (5 rats) or subcutaneous injection (5 rats).
Arterial Thrombosis Test
The thrombosis test was performed according to Kurz et al 10 and Leadley et al 11 with minor modifications. After the induction of surgical anesthesia and a 10- to 15-minute stabilization period, a medial ventral incision was made extending from the base of the mandible to the suprasternal notch. The carotid sheath was exposed and the carotid artery isolated from the vagus nerve by careful blunt dissection. Loosely tied ligatures were placed on the proximal and distal ends of the vessel. Parafilm was placed under the vessel to prevent the FeCl3 from damaging the surrounding tissue. Flow through the carotid was measured using a velocimeter (Statham Blood Flow Meter). The flow measurements from the velocimeter were routed through an attenuator on a GRASS polygraph and recorded with a chart speed of 0.25 cm/s for later interpretation. After all machines were calibrated, the velocimeter probe was placed around the proximal portion of the exposed carotid and the ground wire was sewn into the hypodermal side of the skin. Filling the periarterial space of the probe with electroconductive Cardio-Cream (Ingram and Bell Medical, Don Mills, Ontario, Canada) ensured a more accurate reading. After a satisfactory baseline reading was established, thrombosis was initiated by placing a 2 × 3 mm piece of Whatmann #1 filter paper soaked in 30% FeCl3 (w/v) on the distal carotid. Flow readings were measured continuously for a 60-minute period beginning immediately after the addition of the FeCl3-soaked filter paper. The time when the flow through the vessel was reduced was recorded as the time when the thrombus formation began (TTB). The time when the flow in the carotid artery ceased was recorded as the TTO. If the vessel was not occluded at 60 minutes, the value of 60 minutes was used for TTO. The investigators were blinded when the carotid flow was measured and when the tracings were interpreted for TTB and TTO.
Collection of Blood and Blood Vessels
At 60 minutes, immediately after the FeCl3-soaked filter paper was removed, the abdominal cavity was opened and a blood sample of approximately 8 mL (9 parts blood to 1 part 3.8% sodium citrate) was obtained from the abdominal aorta. Plasma was prepared by centrifugation at 3000g for 15 minutes. As a source of endothelium, the thoracic aorta and the caudal vena cava were removed and placed in saline. Finally, the ligatures around the carotid were tightened, and the vessel was removed and kept moist with a drop of saline.
Examination of the Formed Thrombus
The treated section of carotid artery was opened under a dissecting microscope using microsurgical scissors and forceps. The vessel was cut longitudinally and rinsed with a gentle stream of saline to remove the blood. The thrombus was scraped out and placed on a small square of pre-weighed wax paper and was air-dried for at least 24 hours. The paper and thrombus were then weighed, in a blinded fashion, and the thrombus weight determined by subtracting the paper weight.
Examination of Animals for Bleeding
Animals were examined for any evidence of external bleeding every 48 hours when the rats were administered with heparin or saline. This included examination of the nose and fur of the rat and for presence of blood in the bedding of the cage. After 30 days of treatment, when tissues were collected from the rats, internal organs, including the lung, gastrointestinal (GI) tract, and kidney were examined for signs of hemorrhage and the thoracic and abdominal cavity were examined for the presence of stale pooled blood.
Endothelial Extracts
Endothelium from the thoracic aorta and caudal vena cava was collected as described by Hiebert and Jaques. 12 The vessels were slit open and pinned on dental wax with their lumen sides facing upward. The endothelial surface was then gently rinsed with saline. Cellulose acetate paper (Gelman Sciences, Inc, Ann Arbor, Michigan) was placed on the endothelial surface, patted down using curved forceps to assure adequate contact, and when lifted, the endothelium was removed from the vessel. The length and width of the imprint were measured to the nearest millimeters. Mean areas for aortic and vena caval endothelium were 2.71 ± 0.35 and 0.39 ± 0.06 (SD) cm 2 , respectively. Cellulose acetate paper was removed from the endothelium by dissolving in cold acetone followed by centrifugation and disposal of the supernatant. This process was repeated twice and then dried, resulting in an endothelial extract. Glycosaminoglycans (GAGs) were extracted from the endothelial extract by digestion with pronase (2 mg/mL, Sigma, St Louis, Missouri). Samples were then centrifuged at 10 000 rpm for 10 minutes, supernatant was collected, and the precipitate was washed twice with 250 μL 26.8% NaCl. The washings were added to the original supernatant. Glycosaminoglycans were precipitated from the supernatant with 5 volumes of methanol and the precipitate dried.
Urine and Feces
On day 29, the day before the thrombosis test, the rats were placed in metabolic cages and urine and feces were collected for a 24-hour period. Urine was dialyzed against water using 1000 molecular weight cutoff (MWCO) dialysis tubing (Spectrum Laboratories, Inc, Rancho Dominguez, California). Glycosaminoglycans were extracted from feces by a modified method of Jaques. 13 Feces were defatted with acetone and isopropanol/petroleum ether (1:1) and digested at 37°C by pronase, 40 mg/kg in 1 mol/L Tris buffer in 0.1 mol/L CaCl2 at pH 8. Digests were purified by precipitating with 1% NaCl in acetone and then methanol. Precipitates were dried, dissolved in water, and analyzed.
Chemical Determination of Heparin in Endothelium, Feces, and Urine
Agarose gel electrophoresis was used to identify and measure heparin in endothelial extracts, feces, and urine. 14 The dried powders, dissolved in suitable volumes of water, were applied to agarose gel slides along with the administered heparin used as a reference. Following electrophoresis, gels were fixed in 0.1% hexadecyltrimethylammonium bromide (Sigma) and air-dried. Slides were stained with 0.04% toluidine blue in 80% acetone and background color was removed with 1% acetic acid. Heparin was identified by electrophoretic migration as compared to reference heparin, at known concentrations, and amounts determined by densitometry. Heparin was differentiated from other GAGs using migration distance and staining characteristics with toluidine blue. The limit of detection was 0.02 and 0.01 μg/cm2 for aortic and vena caval endothelium, respectively. The investigators were blinded when the endothelial concentrations were measured.
Indicators of Plasma Heparin
The activated partial thromboplastin time (aPTT kit, Sigma) and the anti-factor Xa chromogenic assay (Accucolor, Sigma) were used to determine whether heparin was present in the plasma from all rats. In addition, the antithrombin activity (anti-factor IIa activity, Dia Pharma Group Inc,West Chester, Ohio) was measured in the plasma from controls and UFH-treated animals.
Analyses
Data are expressed as mean ± standard error of the mean (SEM) or standard deviation (SD) in the case of rat weight or area of harvested endothelium. Student’s two-tailed t tests were used to determine the differences between treated and control groups for TTO, TTB, thrombus weight, heparin concentrations in aortic and vena caval endothelium, aPTT, and anti-factor Xa and antithrombin activity. The χ2 test for differences between proportions was used to compare, between groups, the percentage of endothelial samples that were positive for heparin.
Results
Thrombosis
Thrombus formation in the carotid artery was determined by measuring the TTB and TTO over a 60-minute period. As shown in Figure 1 , TTB was significantly increased in rats treated with subcutaneous or oral UFH given at a dose of 1 mg/kg per 48 h, where the TTB was 21.2 ± 2.2 and 25.0 ± 3.9 minutes, respectively, compared to control rats where the TTB was 12.6 ± 1.1 minutes (P = .0013 and .0027, respectively, two-tailed t test). Similarly, TTB was increased with subcutaneous or oral LMWH given at a dose of 0.1 mg/kg per 12 h where TTB was 21.7 ± 3.1 and 21.3 ± 1.7 (P = .0047 and .0004, respectively, when compared to control). The TTO was significantly greater in all heparin-treated groups compared to control (29.1 ± 3.6 minutes) and was 54.8 ± 4.0, 60.0 ± 0.3, 56.7 ± 3.3, and 58.3 ± 1.7 minutes for subcutaneous UFH, oral UFH, subcutaneous LMWH, and oral LMWH, respectively (P = .0004, <.0001, .0002, and <.0001, respectively). As shown in Table 1 , the dry thrombus weight and percentage change in thrombus weight were significantly decreased in all heparin-treated rats compared to control (P = .0020, <.0001, .0006, and .0003 for subcutaneous UFH, oral UFH, subcutaneous LMWH, and oral LMWH, respectively, for both parameters). Thrombus weight and percentage change in thrombus weight for rats given oral UFH was significantly less than subcutaneous UFH (P = .0002 for both parameters).
Time till thrombus begins (TTB) and time till occlusion (TTO) in the carotid artery of rats treated with oral or subcutaneous heparin. Rats were treated with subcutaneous (sc) or oral unfractionated heparin (UFH, 1 mg/kg per 48 h) or sc or oral low-molecular-weight heparin (LMWH, 0.1 mg/kg per 12 h) for a 30-day period. On the last day of heparin administration, a thrombus was created by the addition of FeCl3 to the exposed carotid artery. The TTB (open circles) and TTO (filled circles) were recorded by a flow meter placed around the carotid artery for 60 minutes. A value of >60 is shown when occlusion did not occur in the 60-minute period. Individual values are shown with the solid line indicating the calculated mean value where 60 was used when TTO was greater than 60 minutes. aSignificantly different than control, two-tailed t test.
Dry Thrombus Weight 60 Minutes After Application of FeCl3 to the Carotid Artery Following Oral or Subcutaneous (sc) Heparin Administration in Repetitive Doses for 30 Days a
Abbreviations: UFH, unfractionated heparin; LMWH, low-molecular-weight heparin; SEM, standard error of the mean.
aMean ± SEM
bSignificantly different than control.
cSignificantly different than sc UFH, two-tailed t test.
Anticoagulant Activity
The aPTT and anti-factor Xa activity were measured in plasma from blood samples obtained from all rats immediately after the thrombosis test. Antithrombin activity was measured in plasma from control rats and those treated with UFH but not LMWH. As shown in Table 2 , aPTT was significantly elevated only in the oral UFH-treated group compared to control (P = .3052, .0359, .2769, and .4562 for subcutaneous UFH, oral UFH, subcutaneous LMWH, and oral LMWH, respectively). Anti-factor Xa activity was significantly elevated in the LMWH-treated groups compared to control, P = .0458 and .0406 for subcutaneous and oral LMWH, respectively. There was no difference in values for anti-factor Xa or antithrombin activity between UFH-treated and control groups (P = .2991, .0805 for anti-factor Xa activity and P = .2226 and .1641 for antithrombin activity for subcutaneous and oral UFH, respectively). There was no evidence of bleeding in the heparin-treated or control animals.
Anticoagulant Activity Following Oral or Subcutaneous (sc) Heparin Administration in Repetitive Doses for 30 Days a
Abbreviation: aPTT, activated partial thromboplastin time; SEM, standard error of the mean.
aMean ± SEM.
bSignificantly different than control, two-tailed t test.
Recovery of Heparin From Tissue
Heparin was found in some endothelial samples in all treatment groups except the controls (Table 3 ). The percentage of samples that were positive was significantly different than control when UFH was given subcutaneously, for both aortic and vena caval endothelium, and when UFH was given orally in aortic endothelial samples. The concentration on both aortic and vena caval endothelium was greatest and significantly different than control when UFH was given subcutaneously (for aortic endothelium P = .0227, .0635, .2071, and .1521 and for vena caval endothelium P = .0201, .1648, .2071, and .2071 for subcutaneous UFH, oral UFH, subcutaneous LMWH, and oral LMWH, respectively). Only small amounts of heparin were recovered from urine and feces. One of 10 animals in the control group and 1 of 7 animals in the group treated with oral UFH had heparin like material in the urine (0.002 and 0.009 μg/mL, respectively). Heparin was found in 1 of 7 animals in the oral UFH-treated group and 4 of 6 animals in the subcutaneous LMWH-treated group (0.6 ± 0.6, and 1.2 ± 0.4 μg/g wet weight, mean ± SEM, respectively).
Heparin Found With Endothelium Following Oral or Subcutaneous (sc) Heparin Administration in Repetitive Doses for 30 Days
Abbreviations: UFH, unfractionated heparin; LMWH, low-molecular weight heparin.
aMean ± SEM.
bSignificantly different than control, χ2 for differences between proportions.
cSignificantly different than control, two-tailed t test.
Discussion
Our laboratory has collected considerable data that support the notion that orally administered heparins are effective antithrombotic drugs.5,7,15,16 Most of our studies have focused on single-dose administration, however prophylactic use of heparin depends on the administration of repeated doses. In this regard, we have previously shown, in a 30-day study, that repetitive orally administered UFH is effective in a rat venous thrombosis model 5 ; however, this is the first report of the effect of repeated doses of oral heparin on arterial thrombosis. Repetitive doses, used in the present study, were those previously found effective in the rat venous thrombosis model, that is, 1 mg/kg per 48 h for UFH 5 and 0.1 mg/kg per 12 h for the LMWH tinzaparin. 17 Based on the results in the present study, repetitive administration of oral heparins for 30 days prolong TTB and TTO and reduce thrombus weight in the carotid arterial model, thus oral heparins are effective antithrombotic agents for both venous and arterial thrombosis.
The ability of oral administration of heparins to prevent arterial thrombosis is similar to that observed for subcutaneous administration where TTB and the TTO were similarly prolonged. This suggests that distribution of heparin to blood and blood vessels is as complete after oral heparin administration as it is after subcutaneous administration. In fact, thrombus weight was significantly reduced in the oral UFH-treated group compared to the subcutaneous UFH-treated group. Additionally the carotid artery was not occluded at 60 minutes (TTO >60 minutes) more in rats treated with oral UFH than those treated with subcutaneous UFH, suggesting that oral UFH may be a more effective antithrombotic agent than subcutaneous UFH when given at 1 mg/kg per 48 h for 30 days. Therefore, orally administered heparin should not be ruled out as an agent for arterial or venous thrombophylaxis.
The increase in TTO and TTB occurred without a substantial increase in anticoagulant activity. The aPTT was significantly elevated compared to control only in the oral UFH-treated group where the mean value was 3.5 seconds greater than the mean control value (Table 1). Anti-factor Xa activity was significantly elevated compared to control only when LMWH was given, evident in both oral and subcutaneous LMWH administration. The observation that thrombosis was prevented without substantial changes in anticoagulant activity supports the idea observed clinically that there is a poor correlation between anticoagulant activity and antithrombotic activity and bleeding tendency.18,19 Heparin was recovered from both aortic and venal caval endothelium in up to 67% of samples, the highest recovery seen when UFH was given subcutaneously although amounts found after subcutaneous administration did not differ significantly from that found after oral administration. This supports our previous hypothesis that the interaction of heparin with endothelium likely contributes to antithrombotic activity. We have previously shown that heparin readily binds to endothelium in in vitro and in vivo studies, regardless of the route of administration. 20 The amounts recovered from endothelium were quite variable for all treatment groups. This is not surprising since only the endothelium from the superior vena cava and thoracic aorta were harvested. Although endothelium can easily be harvested from the aorta and superior vena cava, the endothelial surface area collected is approximately 3 cm2 and therefore represents only a very small portion of the total endothelial surface in the rat which is estimated to be 2 m2 based on the value of 7000 cm2 of vascular surface area/100 g of muscle. 21 The avid distribution to the endothelium, the small amount of endothelium available for sampling, and the small amount of drug remaining in plasma, at the doses used, make it difficult to calculate an oral bioavailability.
It is obvious that the pharmacokinetics of orally administered UFH is quite different than that of the oral LMWH tinzaparin. The present results suggest that the antithrombotic effect of oral UFH is longer lasting than the LMWH tinzaparin since UFH is given every 48 hours and LMWH is given every 12 hours. However, the amount of LMWH tinzaparin given is less (0.1 mg/kg) than that of UFH (1 mg/kg), suggesting a faster absorption of LMWH. 6 The faster absorption of LMWH agrees with our previous observations in single-dose studies where doses of 0.025 and 0.1 mg/kg of the LMWHs reviparin and tinzaparin, respectively, were required to reduce thrombosis by 50% compared to 7.5 mg/kg for UFH when thrombosis in rat jugular vein was examined 4 hours after thrombus initiation and heparin administration. 7 These observations also agree with our studies on the absorption of heparin using gastric mucosa mounted in a Ussing chamber where we show that UFH crosses the membrane best in an acidic environment, while LMWHs are transported better in a basic environment. This suggests that UFH may be best absorbed in the stomach, while LMWH may be best absorbed in the intestine where a much larger absorptive surface is available and LMWH may be absorbed faster or more completely. 22
Thus, the present studies indicate that orally administered heparins, both UFH and LMWH, are effective antithrombotic agents in both rat arterial and venous models and can be given in repetitive doses without bleeding. Oral administration produces a similar antithrombotic effect as seen with subcutaneous administration, supporting the notion that orally administered heparins are readily absorbed. Further studies, including clinical studies, should be performed to further investigate the use of oral heparin which would be of benefit to patients and clinicians.
Footnotes
The author(s) declared no conflicts of interest with respect to the authorship and/or publication of this article.
The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: This work was supported by an operating grant from the Heart & Stroke Foundation of Saskatchewan.