Pulmonary Arteriography and Hemodynamics During Feline Heartworm Disease

Pulmonary Arteriography and Hemodynamics During Feline Heartworm Disease Effect of Aspirin Clarence A. Rawlings, DVM, PhD The ability of aspirin to block arterial disease and thromboembolism of the pulmonary arteries was studied in heartworm-infected cats. Three groups of cats were transplanted with four heartworms per cat and studied. One group of eight cats (aspirin group) received aspirin (97.5 mg, twice a week) for the five-month infection and another group of eight cats served as the nontreated control group (nontreated group). Based upon the results of the first two groups, the third group (adjusted aspirin group) of six cats was studied in which the aspirin dosage was adjusted in order to maintain an inhibition of in vitro platelet aggregation. Cats were studied by nonselective pulmonary arteriograms before heartworm transplantation and by selective arteriograms, aortograms, and pulmonary hemodynamics five months after heartworm transplant. Pulmonary hypertension, (mean pulmonary artery pressures > 16 mmhg), was discovered in three cats with one cat in each group. There were no differences in the mean pulmonary artery pressure or vascular resistance between the groups. Many of the arterial diameters for the nontreated and aspirin groups were greater after the five-month infection than before heartworm infection. All of the postinfection caudal arteries were tortuous and had aneurysms. Some of the caudal lung lobes had perfused areas that appeared to have a hypervascular microvasculature. The proportion of obstructed right and left distal caudal pulmonary arteries and the resulting nonperfused area of the caudal lung lobe in the nontreated and aspirin treated groups were each greater than in the adjusted aspirin group. Prominent bronchial arteries and arterial circulation were present in three of six of the nontreated cats, five of seven of the aspirin-treated cats, and two of five adjusted aspirin cats, that had diagnostic aortograms. Aspirin given at a standard dosage of 97.5 mg twice a week did not reduce the arteriographically demonstrated arterial disease, the pulmonary hypertension, or the obstruction of caudal lobar arteries by thromboembolism and exuberant villus proliferation. The adjusted aspirin dosage, which required daily administration in some cats, appeared to have limited benefits. Since this dosage was individualized for each cat and near the toxic dosage for cats, we are reluctant to recommend aspirin in cats with heartworm disease. Nonselective pulmonary arteriograms, as done in this study, can be useful in cats with suspected heartworm infection and radiographic signs, but which have negative test results for microfilaria and ELISA-antigen. (Journal of Veterinary Internal Medicine 1990; 4:285-291) THE domestic cat can provide an adequate host environment for the canine heartworm, DiroJlaria immitis, From the University of Georgia, Department of Small Animal Me&- cine and the Department of Physiology and Pharmacology, College of Veterinary Medicine, Athens, Georgia. Supported by the University of Georgia Veterinary Experimental Station, Athens, Georgia. The author sincerely appreciates the technical assistance of Mr. Richard M. Mahood. Reprint requests: Clarence A. Rawlings, DVM, PhD, University of and cats living in heartworm endemic areas are commonly diagnosed as having heartworm disease.'-3 In 11 different surveys that compared the heartworm infection in dogs and cats at necropsy, the infection rate in cats ranged - from 0% to 8.5% as comdared with 5.4% to 59.1% in dogs in the same area. The average frequency in these studies was 1.9% in cats versus 34.6% in dogs.4 Since there are fewer heartworms uer cat and these heartworms have a shorter life than those in the dog, the Georgia, Department of Small Animal Medicine, College of Veteri- heartworms may die and be absent in nary Medicine, Athens, GA 30602. many necropsied cats. Clinical signs of heartworm dis- 28 5

286 RAWLINGS Journal of Veterinary Internal Medicine TABLE 1. Arteriographic Changes Nontreated Aspirin Adjusted Aspirin Main PA* Right PA* Left PA* Right caudal PA* Left caudal PA* Right cranial PA* Right distal caudal pulmonary artery (proportion that were obstructed) Left distal caudal pulmonary artery (proportion that were obstructed) Percentage of nonperfused caudal lobe 10.64 f 1.70 11.29 f 1.07 6.00M f 1.00 6.36'd f 0.63 5.29a f 0.39 7.21' f 0.95 3.14 f 0.38 3.57 f 1.88 2.93' f 0.45 3.64' f 1.73 2.07' f 0.19 2.5OhC f 0.41 4/7".h 417" 3.5" f 1.8 27.1h f 8.6 10.14 f 1.95 11.29 f 1.47 5.21" f 1.11 6.34cd f 0.80 5.14" f 1.14 6.29' f 0.16 2.36 k 1.03 3.36? 1.77 2.07" f 0.53 3.50' f 0.87 1.50" f 0.29 2.35h f 0.80 617" 517" 6.Yh f 3.1 41.0a f 11.2 11.00 f 1.28 9.90 -+ 1.47 5.8Oahe f 0.57 6.10" f 0.96 6.50d f 0.35 6.50kd? 1.41 2.20 * 0.57 2.90 f 0.96 2.50"" -+ 0.6 I 3.50Cd k 1.12 1.70ad k 0.27 1.90d f 0.42 215' 0/6b 7.2' f 3.3 15.6' f 6.0 * PA is the diameter (mm) of pulmonary arteries. Means for each of the arterial diameters within a row and column (top 7 rows) that were not different by a two factor analysis of variance are noted by the same superscript. There are no differences between means of either the main pulmonary arterial diameter or the right caudal pulmonary arterial diameter. Means of proportion of obstructed arteries and of percentage of nonperfused caudal lung lobe are also not different if their superscripts are the same. ease in cats include coughing, vomiting, dyspnea, lethargy, anorexia, weight loss, syncope, and blindness. Many signs are chronic, but the most dramatic appear to be associated with heartworm death and thromboembolism. Some cats die so acutely that premortem disease signs are not Heartworms within the pulmonary arteries produce dilation, tortuosity, and peripheral arterial pr~ning.~,~ Villous myointimal proliferation in the larger pulmonary arteries develop in a fashion similar to their development in dog^.^.^.' Thromboembolism has been assumed based upon the abrupt pruning and obstruction of flow to large and small arteries and upon the associated parenchymal disease. Arterial enlargement can be seen radiographically within a few weeks of heartworm transplantation and is a useful diagnostic sign for feline heartworm infe~tion.~,~,~ Nonselective pulmonary arte- riograms can be useful to diagnose intra-arterial heartworm silouhettes and the pathognomonic arteriographic features of heartworm disease. In dogs with heartworm disease, aspirin impedes platelet adhesion to damaged arterial surface^.^ Aspirin has arrested arterial disease development during heartworm infection and permitted resolution in dogs with arterial disease already produced by live heart worm^^-'^ and has reduced the exuberant villus proliferation and thromboembolism produced by dead heartworms. I Since the arterial response in cats appears to combine thromboembolism and severe villus proliferation with obstruction of arteries, aspirin was evaluated in cats with a controlled heartworm infection. Materials and Methods Cats for this study were healthy based upon physical examination and a nonselective pulmonary arteriogram. The cats had been quarantined for at least two months before the study and had been vaccinated for feline panleukopenia, calicivirus, and rhinotracheitis. Three groups of cats were transplanted with four heartworms per cat and studied. One group of eight cats (aspirin group) received aspirin* (1.5 grains (97.5 mg) twice a week) for the five-month infection and another group of eight cats were the nontreated control group (nontreated group). Based upon the negative results in the first two groups, a third group (adjusted aspirin group) of six cats was studied in which the aspirin dosage was adjusted in order to maintain an inhibition of in vitro platelet aggregation. The aspirin dosage of 97.5 mg twice a week given to the aspirin group was based upon suppression of in vitro aggregation of platelets obtained from normal cats.'* Platelet aggregation studies were not done in these heartworm-infected cats. The adjusted aspirin dosage group had platelet aggregation studies * St. Joseph Aspirin for Children, Plough Inc. Memphis, TN.

VOI. 4. NO. 6 ARTERIOGRAPHY IN HEARTWORM-INFECTED CATS 287 FIG. 1A. Nonselective pulmonary arteriogram of a cat in the nontreated heartworm-infected group before heartworm infection. Arteries gradually taper with normal arborization of the arterial tree with small arteries being perfused uniformly. Reprinted with permission from the Proceedings of the American Heartworm Society. 1989. done at two- to three-week intervals and the dosage was progressively increased in order to block aggregation. The aspirin dosage for individual cats at the end of the study ranged from 25 mg/kg twice a week to 15 mg/kg daily. The cats were each studied by a nonselective arteriogram before heartworm transplantation. The cats were sedated with 0.2 mg of acepromazine, 0.14 mg of atropine, and 50 mg of ketamine IM, and then anesthetized with 33 mg of pentobarbital sodium IV. A jugular venotomy was aseptically performed to place an 18 gauge catheter.? This catheter tip was within the cranial vena cava. During inflation of the lungs (1 5 to 20 cm H20 airway pressure), 1 ml/kg of meglumine iothalamate was injected using hand pressure. Four radiographs were exposed at 0.6 second intervals.' Four heartworms, which had been obtained from dogs superinfected with infective larvae, l3 were then transplanted via a similar jugular catheter with an amputated tip. Each heartworm was loaded longitudinally into an intravenous extension tube and then rapidly injected through the catheter and into the central venous system. Five months after heartworm transplantation, the cats were reanesthetized in similar fashion for pulmonary pressure measurement, selective pulmonary arteriography, and aortography. Polyethylene catheters (PE- 190) were placed into the main pulmonary artery and ascending aorta via the external jugular vein and the carotid artery, respectively. Pressures within the arteries were transmitted through the catheters to strain gauges.14.15$ t Monoject Sovereign Indwelling Catheter, Shenvood Medical, St. Louis, MO. 4 Statham P23b, Statham Instruments Inc, Oxnard, CA. Following pressure and flow measurements, meglumine iothalamate was injected into each catheter and pulmonary arteriograms and aortograms performed. The aortograms were performed to characterize the bronchial artery flow. Pulmonary arteries were measured at sites listed below at the widest diameter visualized on successive artenograms, i.e., during systole. The pre- and postartenograms were measured at the same time in order to compare the same diameters for each cat. The sites measured included 1) main pulmonary artery, just distal to the postvalvular dilation; 2) left pulmonary artery, distal to its branching from the main pulmonary artery and proximal to the branching of the left cranial lobar artery; 3) right pulmonary artery, distal to its branching from the main pulmonary artery and proximal to the branching of the right cranial lobar artery; 4) left caudal pulmonary artery, just cranial to the ninth rib; 5) right caudal pulmonary artery, just cranial to the ninth rib; 6) left distal caudal pulmonary artery, just cranial to the 1 lth rib; 7) right distal caudal pulmonary artery, just cranial to the 1 lth rib; and 8) right cranial pulmonary artery, at the level of the fourth rib. Since many cats had obstructions in their caudal pulmonary arteries, this level of obstruction was identified and the percent of the caudal lung lobes that were not perfused was measured. The area of the caudal lung lobes as viewed with a laterally projected radiograph was defined as being bordered by the ventrum of the thoracic vertebra, the caudal margin of the most caudally viewed seventh rib with a line drawn from its ventral portion to the junction of the diaphragm with the sternum, and the cranial margin of the most caudal crus of the diaphragm. The nonperfused region was defined as the radiolucent area of the caudal lung lobe not receiving contrast perfused arterial or microcirculatory flow. The area of each lung lobe and the nonperfused area was measured with computerized data collecting system for geometric measurements using a digitizing tablet. Means and standard deviations were calculated for the mean pulmonary arterial pressure, cardiac output, total pulmonary resistance, the arterial diameters, and percent of nonperfused caudal lung lobe. Arteriographic data were analyzed using a two-factor analysis of variance (ANOVA) and when there were significant differences (P < 0.05), means were then compared using a Tukey test. The hemodynamic data were analyzed using a one-factor ANOVA. The proportions of obstructed distal caudal pulmonary arteries were compared by a z-statistic for binomial parameters. Statistical significance was defined as P < 0.05.16 the final studies of pulmonary hemodynamics and arteriography and aortography, the cats were euth- 0 Micro-comp Data Acquisition System, Southern Micro Instruments, Inc. Atlanta, GA.

288 RAWLINGS Journal of Veterinary Internal Medicine FIG. I B. Selective pulmonary arteriogram of the same cat as in Fig. 1A after a five-month heartworm infection. There is marked dilation, tortuosity, and obstruction of the arteries. Most of both caudal lung lobes are not perfused and the microcirculation of the cranial portion of the caudal lobes have a blush appearance. This cat was hypertensive with a mean pulmonary arterial pressure of 28 mm Hg. Reprinted with permission from the Proceedings of the American Heartworm Society, 1989. anized and their lungs perfusion fixed for morphologic studies. The lungs were dissected and processed for both light and scanning electron microscopy. The typical heartworm induced lesions of the caudal pulmonary arteries were confirmed. Since the heartworms were fixed in situ, their survival could not be confirmed, but several of the heartworms, as observed grossly and with light microscopy, were either dead or were in the process of dying. All cats had heartworms that appeared to be alive at the time of necropsy based upon previously reported criteria of normal reflective cuticle, normal shape, and lack of uptake of Evan s blue dye. Results One cat from each group died. The nontreated cat died acutely; the aspirin-treated cat was diagnosed and treated several days before its death; and the adjusted aspirin cat died during five-month angiography study. Necropsies of the two sick cats (1 from the nontreated group and 1 from the aspirin group) revealed signs of bronchopneumonia complicated by heartworm thromboembolism. The remaining cats of each group completed all procedures. Only three of the 19 cats studied had pulmonary hypertension with a mean pulmonary artery pressure higher than 16 mm Hg. The aspirin-treated cat with a mean pulmonary arterial pressure of 65 mm Hg had 85% of its arterial surface involved with villus proliferation and 52.4% of its caudal lung not perfused. The non-aspirin-treated cat with a pressure of 28 mm Hg had only 17.1% involvement of the arterial surface, but FIG. IC. Aortogram ofthe cat in Fig. 1 B after a five-month heartworm infection. Contrast medium outlines a well-developed bronchial arterial circulation. Medium also appears in the aorta and the branches of the celiac artery. 40.2% of the caudal lung lobe was not perfused. The adjusted aspirin-treated cat with a pressure of 22.5 mm Hg had 81.9% of its pulmonary artery involved, but 25.8% of the caudal lung lobe was not perfused. In each heartworm group, the cat with the highest percentage of the caudal lung lobe not perfused was the only cat in that group with pulmonary hypertension. The means of the mean pulmonary arterial pressure of the groups (nontreated 14.3 f 7.0 (SEM); aspirin, 20.4 f 19.7; and adjusted aspirin, 12.3 f 5.9) were not different. The higher mean for the aspirin-treated group was a reflection of the one markedly hypertensive cat. Many of the arterial diameters for the nontreated and aspirin groups were greater after the five-month infec- FIG. 2. Selective pulmonary arteriogram of a nontreated cat after heartworm infection. Arterial dilation, an intraluminal negative contrast object, and obstruction of distal arterial flow are present.

Vol. 4 * NO. 6 ARTERIOGRAPHY IN HEARTWORM-INFECTED CATS 289 FIG. 3. Selective pulmonary arteriogram of a nontreated cat after heartworm infection. The left caudal lobar artery is abruptly obstructed and the right has minimal flow. Dilation, tortuosity, and altered microcirculation are present. tion than before heartworm infection (Table 1). The distal caudal pulmonary arteries were frequently not measured after infection since many were obstructed. None of the cranial arteries were obstructed. All of the postinfection caudal arteries were tortuous and had aneurysms (Figs. 1 to 4). In the affected caudal lung lobes, some perfused areas appeared to have a hypervascular microvasculature, in that the contrast medium produced a transient blush in the parenchyma just cranial to the obstructed artery. In retrospect, arteriograms taken before heartworm transplantation of one of the aspirin-treated cats had at least one heartworm silhouette and sluggish flow to the caudal lung lobes. For appropriate statistical comparisons, the premeasurements were not used for this cat. The proportion of obstructed right and left distal caudal pulmonary arteries in the nontreated and aspirin treated groups was each greater than in the adjusted aspirin group (Table 1). This observation was similar to the mean percentage of caudal lung lobe that was not perfused in that the lowest percentage of nonperfusion was in the adjusted aspirin group and the highest was in the aspirin group (Table 1, Fig. 5). Aortograms were subjectively evaluated. Prominent bronchial arteries and arterial circulation were present in three of six of the nontreated cats, five of seven of the aspirin-treated cats, and two of five adjusted aspirintreated cats, that had diagnostic aortograms. In addition to prominent arteries, the injection also produced a persistent blush of the microvasculature about the caudal bronchi (Fig. 4B and 4C). The two cats with the highest pulmonary artery pressure had prominent bronchial arterial circulation. FIG. 4A. Selective pulmonary arteriogram of a cat after a five-month infection with heartworms and aspirin treatment. Both caudal lobar arteries are obstructed and the intralobar arteries are dilated. Discussion Two clinical relevant conclusions could be derived from this study. First, the main hypothesis of aspirin s value was not demonstrated and we do not recommend aspirin in cats with heartworm disease. Second, nonselective pulmonary arteriograms can be done on clinical patients and can be useful in those cats with suspected heartworm infection and radiographic signs, but that have negative test results for microfilaria and ELISAantigen. Most cats seen clinically do not have circulating microfilaria and the ELISA-antigen tests results are frequently falsely negative due to the small number of heartworms in most cats. The hypothesis is that by reducing platelet adhesion, aspirin could suppress pulmonary arterial disease and FIG. 4B. Aortogram of the same cat in Fig. 4A after a five-month heartworm infection and aspirin treatment. Contrast medium is outlines a well-developed bronchial arterial circulation.

290 RAWLINGS Journal of Veterinary Internal Medicine ADJUSTED ASPIRIN Pre (4.9%) Post (1 5.7%) FIG. 5B. Traces as presented in Figure 5A ofan adjusted aspirin group. The postarteriogram is typical in that the mean area not perfused at the postarteriogram for this group was 15.6%. FIG. 4C. The radiographic exposure, approximately 0.6 s. after Fig. 4B. Contrast medium is moving slowly through the bronchial microcirculatory system and there is a blush in the caudal lung lobes. hypertension in heartworm-infected cats. This hypothesis was based upon prior research that heartworms produce endothelial injury, which leads to platelet adhesion and release of platelet derived growth factor, which induces myointimal proliferation of the pulmonary arteries. Activated platelets, which are responsive in the cat, also produce thromboemboli, and this produces focal alveolar disease in the caudal lung lobes. In this study, aspirin administered at a standard dosage of 97.5 mg twice a week did not reduce the arteriographically demonstrated arterial disease, the pulmonary hypertension, or the obstruction of caudal lobar arteries by thromboembolism and exuberant villus proliferation. The adjusted aspirin dosage, which required daily drug administration in some cats, appeared to have limited benefits. CONTROL Pre (2.6%) Post (29.6%) FIG. 5A. Traces from the caudal lung lobe and the area of nonperfusion as viewed from a lateral projection of the pulmonary artenogram of a nontreated heartworm cat. This was typical for the nontreated group in that the mean area not perfused at the postarteriogram for the group was 27.1%. Since this dosage was individualized for each cat and near the toxic dosage for cats, safe clinical dosage recommendations can not be made. The aspirin dosage required to block platelet aggregation in normal cats was 25 mg/kg given every three days. The dosage of aspirin in the aspirin cats averaged 35 mg/kg given twice a week, with the lowest dosage of a 90 mg tablet in a 4.1 kg cat being 22 mg/kg. In the adjusted aspirin group, the aspirin dosage in all cats had to be increased during the five-month heartworm infection period in order to block in vitro aggregation. There were periods when the aspirin dosage was ineffective in blocking in vitro function of platelets and during these times, platelets could have participated in worsening the arterial disease. In our studies of heartworm-infected dogs, we have observed that the mean platelet survival time is 5.2 1 days in normal, uninfected dogs and 4.17 days in heartworm-infected dogs. When heartworminfected dogs were administered a cyclooxygenase inhibitor,ll the dosage required to block in vitro platelet aggregation increased from 1 mg/kg before heartworm infection to 2 mg/kg one month after infection and to 3 mg/kg three months after infection. The chronic endothelial injury and platelet consumption of heartworm disease appears to increase platelet turnover and therefore would require either an increased dosage or frequency of administration of a platelet blocker. Platelet activation by other pathways than arachidonate has been proposed and aspirin may not block these alternate pathways. Despite this rationale, aspirin has blocked platelet involvement in normal cats,12 cats with experimental aortic thrombosis at a dosage of 650 mg/ cat, and cats with lethal endotoxin infusion with dosages of 10 to 100 mg/kg.* The arachidonate pathway has been demonstrated to be the dominant system to stimulate and produce aggregation in vitro in feline IT U-53, 059, Upjohn, Kalamazoo, MI.

VOl. 4. NO. 6 ARTERIOGRAPHY IN HEARTWORM-INFECTED CATS 29 1 platelets in a fashion similar to human platelet^.^'-^^ Finally, it is unlikely that the mechanism for arterial injury and response is different in the cat than in the dog, in that both appear to have similar histologic and arteriographic changes. Most cats had obstructions of the caudal lobar arteries and many had evidence of dead or dying heartworms. This obstructive arterial disease appeared to be more typical of the canine response to dead heartworms rather than that seen with live heartworms. Most of the cats also had evidence of pulmonary hypertension as described histologically, but hypertension was infrequently diagnosed. The severity of the arteriographic lesions, the histologic evidence of pulmonary hypertension, and presence of some dead heartworms support a hypothesis of arterial disease producing an acute and transient hypertension. An acute increase in pulmonary arterial pressure or in vascular permeability, associated with spontaneous heartworm death, is a logical explanation for the acute dyspnea and/or death seen in some cats with heartworm disease. A series of transient pulmonary artery hypertensive episodes could explain the dilation of pulmonary arteries in cats with normal pulmonary artery pressure. The bronchial circulation developed in the presence of decreased pulmonary arterial flow to the caudal lung lobes. This systemic supply may have been beneficial in reducing the onset of ischemia to the parenchyma. References I. Donahoe JMR. Clinical aspects of feline dirofilariasis. Proceedings of the American Heartworm Society 1974; 7459-65. 2. Dillon R. Feline dirofilariasis. In: August JR, Loar AS, ed. Veterinary Clinics of North America: Small Animal Practice, Symposium on Advances in Feline Medicine 11. Philadelphia: WB Saunders, 1984; 1 185- I 199. 3. Rawlings CA. Heartworm Disease in Dogs and Cats. Philadelphia: WB Saunders, 1986. 4. Kume S. Epizootiology of canine heartworm disease in the Tokyo area: Diagnosis and treatment. In: Bradley RW, ed. Canine Heartworm Disease: The Current Knowledge. Gainesville, FL: University of Florida Press, 1970; 38-60. 5. Donahoe JMR, Kneller SK, Lewis RE. In vivo pulmonary arteriography in cats infected with Dirofilaria immitis. J Am Vet Radio1 SOC 1976; 17: 147-1 5 1. 6. Wong MM, Pedersen NC. Cullen J. Dirofilariasis in cats. J Am Anim Hosp Assoc 1983; 19:855-864. 7. Byerly CS, Donahoe MR, Todd KS Jr. Histopathologic changes in cats experimentally infected with Dirofilaria immitis. Proceedings of the Heartworm Symposium 1977; 77:79-8 I. 8. Lewis RE, Losonsky JM, McCall JW, et al. Radiographic changes in cats with transplanted adult Dirofilaria immitis. New Orleans: American Association of Veterinary Parasitology. 12 1 st Annu Meet Am Vet Med Assoc. Abst. 42. July 15-17, 1984. 9. Rawlings CA, Keith JC Jr, Schaub RG. Effect of acetylsalicylic acid on pulmonary arteriosclerosis induced by a I year, low level vascular injury. Arteriosclerosis 1985; 5:355-365. 10. Schaub RG, Keith JC Jr, RawlingsCA. The effect ofacetylsalicylic acid on vascular damage and myointimal proliferation in canine pulmonary arteries subjected to chronic injury by Dirofilaria immitis infection. Am J Vet Res 1983; 44:449-454. 1 1. Rawlings CA, Keith JC Jr, Losonsky JM, et al. Aspirin and prednisolone modification of postadulticide pulmonary arterial disease in heartworm infection: Arteriographic study. Am J Vet Res 1983; 44:821-827. 12. Greene CE. Effects of aspirin and propanolol on feline platelet aggregation. Am J Vet Res 1985; 46: 1820-1823. 13. Rawlings CA, McCall JM. Surgical transplantation of adult Dirofilaria immitis to study heartworm infection and disease in dogs. Am J Vet Res 1985; 46:221-224. 14. Rawlings CA. Cardiopulmonary function in the dog with Dirofilaria immitis infection: During infection and after treatment. Am J Vet Res 1980; 41:319-325. 15. Bjorling DE. Rawlings CA: Relationship of intravenous administration of Ringer s lactate solution to pulmonary edema in halothane-anesthetized cats. Am J Vet Res 1983; 44: 1000-1006. 16. Ott L. An introduction to Statistical Methods and Data Analysis. 2nd ed. Boston: Duxbury Press, 1984; 194-197.636-664. 17. Rawlings CA, Keith JC Jr, Lewis RE, et al. Aspirin and prednisolone modification of radiographic changes caused by adulticide treatment in dogs with heartworm infection. J Am Vet Med Assoc 1983; 182:131-136. 18. Schaub RG, Keith JC Jr. Simmons CA, et al. Smooth muscle cell proliferation in chronically injured canine pulmonary arteries is reduced by a potent platelet inhibitor U-53,059. Thromb Haemost 1985; 53:351-355. 19. Schaub RG, Gates KA, Roberts RE. Effect ofaspirin on collateral blood flow after experimental thrombosis of the feline aorta. Am J Vet Res 1982; 43:1647-1650. 20. Greenway CV, Murthy VS. Mesenteric vasoconstriction after endotoxin administration in cats pretreated with aspirin. Br J Pharmacol 1971; 43:259-269. 2 I. Thomas DP, Niewiarowske S, Ream VJ. Release of adenine nucleotides and platelet factor 4 from platelets and four other species. J Lab Clin Med 1970; 75:607-6 18. 22. Marcinkiewicz E, Grodzinska L, Gryglewski RJ. Platelet aggregation and thromboxane A2 formation in cat platelet rich plasma. Pharm ResComm 1978; 10:1-12. 23. Meyers KM, Katz JB, Clemmons RM, et al. An evaluation of the arachidonate pathway of platelets from companion and foodproducing animals, mink, and man. Thromb Res 1980; 20: 13-24,