Heart-transplant recipients develop an accelerated form of coronary atheroslerosis with a high incidence of sudden death, acute myocardial infarction, and systemic arterial and venous thromboembolic complications (1,2). The disease is associated with several biologic abnormalities, which include dyslipidemias (3), increased platelet reactivity (3), resistance to the inhibitory effect of aspirin (4), increased lipid peroxidation (5), and reduced prostacyclin production (6). Although platelet aggregation and high-density lipoprotein (HDL) cholesterol have been shown to be predictive of acute coronary events (7), we do not know which of these biologic abnormalities is most influential nor which treatment could be protective. In this setting, it is clear that antiplatelet compounds that do not directly interfere with endothelial generation of prostacyclin may be of interest, as even very low dose aspirin has been shown to reduce the generation of this major endogenous, antithrombotic substance (8).
Ticlopidine is thought to be antithrombotic because of its property selectively to inhibit adenosine diphosphate (ADP)-induced platelet activation by antagonizing the binding of ADP to platelet ADP receptors (9). However, no study has been carried out in humans or animals to establish unambiguously whether ticlopidine may directly or indirectly interfere with the generation of prostacyclin. It is also unclear whether ticlopidine may affect thromboxane (TBX), which is essentially produced by activated platelets. In fact, cyclosporin A, the major immunosuppressive agent, has been claimed to injure endothelial cells, to interfere with prostacyclin generation (10,11), and to increase platelet reactivity (12). It is therefore important to clarify whether, in this specific setting, ticlopidine may directly or indirectly decrease prostacyclin generation, for instance by the induction of cytochrome P450 (13) or by an effect on cyclosporin A metabolism. Because of ex vivo platelet activation during blood sampling, measurements of plasma TBX concentrations are unlikely to provide valid evidence of in vivo ticlopidine interaction with TBX metabolism. We therefore elected to measure two stable urinary indices of TBX and prostacyclin generation in vivo, 11-dehydro-TBX-B2 and 2,3-dinor-6-oxoprostaglandin F1α (PGF1α), which have often been used as indicators of platelet-endothelial cell interaction in vivo (14).
Finally, we examined whether ticlopidine may interfere with the generation of nitric oxide, another major product of the vascular endothelium (15). Recent report indeed suggested that, in specific animal models, certain antiplatelet agents may affect nitric oxide production by neutrophils (16). Not to increase the risk of graft rejection in these fragile patients by increasing cyclosporine catabolism (17), we decided to use half the classic dose of ticlopidine in this study. It has actually been shown that for that dosage, there is no interaction between ticlopidine and cyclosporine (18).
MATERIALS AND METHODS
Study design and patients
Twenty clinically stable heart-transplant recipients (all were men and >6 months after transplant) were randomized to receive either ticlopidine, 250 mg, or placebo (blinded capsules), administered orally once daily with the evening meal between 7 and 9 p.m., 1 h after cyclosporine intake. All patients were receiving stable immunosuppressive treatment comprising cyclosporine (dosage adapted to obtain trough cyclosporine blood level >100 ng/ml with respect to kidney function), azathioprine, and prednisolone. The length of time since transplant was similar in the two groups: 35 ± 18 and 33 ± 19 months in the placebo and treated groups, respectively. Ticlopidine treatment lasted 14 days, and 24-h urinary samples were collected for eicosanoid and nitric oxide metabolites analyses on day 0 (before treatment) and day 14. All samples were frozen immediately after collection and stored at −70°C until analyzed. Health status of each subject was determined by physical examination, electrocardiogram, echocardiography, and laboratory screening tests. Blood samples were drawn in the fasting state before the treatment period and 2 weeks later for platelet aggregation (see later) and routine laboratory (hematology and biochemistry). The first dose of study medication was taken on day 0 after urines have been collected and the last dose on day 13 in the evening. Patients were requested to return all packaging and unused medication on day 14, and the investigators recorded the number of tablets taken to check compliance. Patients were not eligible to enter the study if they met any of the following criteria: unstable cardiac condition, heart failure (NYHA class >II), history of recent (<6 months) graft rejection, treated diabetes, hepatic dysfunction, hepatitis B surface (HBs) antigen or HIV antibodies, history of hemostatic disorder or systemic bleeding, history of thrombocytopenia (<100,000/mm3) or neutropenia (<1,500/mm3), treatment with antiplatelet or nonsteroidal antiinflammatory drugs during the last 4 weeks, or current treatment with investigational drug. Whereas the study was conducted on an outpatient basis, patients attended the study center for a 24-h hospitalization on days 1 and 13 to permit 24-h urine collection and morning platelet-aggregation study. For safety reasons (detection of rejection), trough and peak cyclosporine levels were monitored regularly during the study, by asking the patients to attend the research center on days 3, 5, and 10 and by using a radioimmunoassay (Cyclo-Trac SP whole blood; INCSTAR, Stillwater, MN, U.S.A.).
Laboratory data evaluation
For platelet-aggregation studies, blood samples were drawn without stasis from an antecubital vein through a 21-gauge butterfly needle. The blood was anticoagulated with 3.8% trisodium citrate (9:1, vol/vol). In addition, a 5-ml sample was collected in a tube containing EDTA for hematologic determinations and 2 × 10-ml samples were taken without anticoagulant for biochemical and lipid determinations in serum, including creatinine, glucose, bilirubin, γ-glutamyl transferase, and alkaline phosphatase. Serum triglycerides and total cholesterol were measured with an enzymatic assay kit (Biomérieux, Lyon, France).
Platelet-aggregation studies were performed as previously described (3,4). In brief, blood was immediately centrifuged to obtain platelet-rich (PRP) and platelet-poor (PPP) plasma. Aggregation tests were performed on a recording aggregometer (Rubel-Renaud, Lyon, France) in PRP with a platelet count adjusted to 300,000/μl by dilution with PPP. For aggregation tests, 500 μl of PRP was warmed at 37°C for 2 min with stirring at 1,100 rev/min. Then 100 μl of the aggregating agent diluted in complete Tyrode's solution (pH 7.4) was added. We used thrombin (Sigma) at final concentration in PRP of 0.04 NIH units/ml and ADP (Sigma) at final concentration of 9.2 × 10−5M. The aggregometer was adjusted for each sample so that PRP gave no light transmittance and PPP gave 100% light transmittance. The extent of aggregation was recorded as a percentage of the maximal difference between PRP and PPP. The value retained was the mean of three measurements made with each agent. The delay between blood sampling and the platelet-aggregation test was constant for each agent (70-90 min). The tracings were analyzed by two independent observers.
Urinary excretion of TBX and prostacyclin metabolites as indicators of their in vivo systemic production was measured as previously reported (6). Thromboxane production was evaluated by measuring urinary TBX-B2, and the measured metabolite of prostacyclin was 2,3-dinor-6-oxo-PGF1α. Thromboxane and prostacyclin metabolites were extracted on reverse-phase C18 Sep-Pak cartridges (Waters, Milford, MA, U.S.A.) from urine samples in the presence of 2 nCi each of [3H]2,3-dinor-6-oxo-PGF1α (synthesized by incubation of [3H]2,3-dinor-6-oxo-PGF1α with isolated rat hepatocytes, as described by Balazy et al. (19) and of [3H]11-dehydro-TBX2 (Amersham, Les Ulis, France) for recovery correction. These prostaglandins were resolved by thin-layer chromatography on silica gel 60 (Merck-Clevenot, Nogent-sur-Marne, France) by using the organic phase of ethyl acetate/isooctane/acetic acid/H2O (130:30:20:100, vol/vol) as eluant. The 2,3-dinor-6-oxo-PGF1α was quantified by radioimmunoassay by using cross-reactivity of an antibody to 6-oxo-PGF1α (Biosys, Compiègne, France): standard curves were obtained with authentic 2,3-dinor-6-oxo-PGF1α. TBX2 was quantified by an enzyme immunoassay. All kits were purchased from Cayman Chemical (Ann Arbor, MI, U.S.A.).
Urinary nitrite plus nitrate excretion as an indicator of systemic nitric oxide generation was measured with the use of the Greiss reagent by a technique adapted from Conrad et al. (20) and Rosselli et al. (21). Aliquots of urine were diluted (1:20 and 1:40) and incubated in (final concentrations) 25 mM imidazole buffer pH 7.6 with 100 μM NADPH, 2 μM flavin adenine dinucleotide (FAD), and 35 mU/ml of nitrate reductase (Aspergillus niger, Boehringer Mannheim) in a final volume of 100 μl for 2 h at room temperature to convert NO3− to NO2−. Total NO2− (NO2− + NO3−) was analyzed by reacting the samples with the Greiss reagent (+50 μl of 2% sulfanilamide in 10% H3PO4 and +50 μl of 0.2% naphthylethylenediamine). Absorbance was measured at 520 nm in a 96-well microplate reader (Molecular Devices, Paris, France). Amounts of NO2− were interpolated from a standard curve of NaNO3 (0-200 μM; Sigma) enzymatically converted into NaNO2. Urinary nitrite and nitrate excretion was expressed in μmol per 24 h and in mmol per mg of urinary creatinine.
In a previous study conducted in apparently healthy, non-transplanted, young and middle-aged subjects, the mean values recorded for urinary nitrite and nitrate were 2,832 ± 215 μmol/24 h (range, 2,035-4057 μmol/24 h) and 2,131 ± 309 mmol/mg creatinine (range, 960-3,660 mmol/mg creatinine).
Homogeneity between the two groups was assessed for platelet aggregation and for eicosanoid and nitric oxide metabolites on day 0 by using an unpaired Student's t test.
The effects of ticlopidine on hematologic and biochemical parameters, platelet aggregation, and eicosanoid and nitric oxide metabolites were analyzed by comparing the means at days 0 and 14 by analysis of variance (ANOVA) and a paired Student's t test. The variations (D14 - D0) between the two groups for each parameter were compared by unpaired Student's t test. Two-way ANOVA (Group × Time) with repeated measures in time was used to analyze the effects of ticlopidine treatment on cyclosporine and ticlopidine blood levels. Relations between variations of nitric oxide metabolites, TBX metabolite, and platelet aggregation were evaluated by linear-regression analysis. The allowed type I error was fixed at 5%.
The 20 patients completed the study. There was no significant change in the associated treatments between days 0 and 14, in particular for immunosuppressive treatment. ANOVA for repeated measures with time for whole-blood cyclosporin A levels indicated that there was a steady state in both groups for cyclosporin A at day 0. No major influence of ticlopidine could be detected on cyclosporin A metabolism, as shown by the lack of significant change in cyclosporin A trough levels and peak concentrations 2 h after cyclosporine administration. For cyclosporine trough and peak levels, there was no Day effect, no Group effect, and no Day × Group interaction.
Results of platelet function are indicated in Table 1 for thrombin- and ADP-induced platelet aggregation. With thrombin, as expected, there was no effect of ticlopidine. With ADP, there was a borderline nonsignificant decrease of the first wave at day 14 compared with day 0 in the ticlopidine group and no change in the placebo group. Regarding the second wave, there was a significant treatment effect with a significant decrease in the ticlopidine group (from 40.2% to 14.7%) but not in the placebo group. Because the reduction in one ticlopidine patient was unusual (from 84.3 to 31.2%), we repeated the analysis without this patient. The treatment effect remained significant (from 35.3 to 13.4%; p = 0.004).
Between days 0 and 14, diuresis was similar in the two groups: 2,115 ± 576 and 2,537 ± 754 ml/24 h in the placebo and ticlopidine groups, respectively at day 14, and 2,440 ± 606 and 2,632 ± 850 ml/24 h at day 0. Urinary excretion of creatinine (1,053 ± 532 and 1,134 ± 577 mg/L at day 14 and 741 ± 349 and 1,016 ± 481 mg/L, respectively at day 0) also did not change significantly. Urinary excretion of TBX and prostacyclin metabolites expressed in picograms per milligram of creatinine were not significantly different. However, when expressed in nanograms per 24 h (as in Table 2) a statistically non-significant (p = 0.10) tendency to a higher excretion of TBX-B2 at day 14 compared with day 0 was seen in the ticlopidine group. Also, when we examined the relation between the variations in ADP-induced platelet aggregation and the variations of TBX between days 0 and 14, we found a borderline nonsignificant inverse correlation (r = −0.44; p = 0.06), which suggested that ticlopidine treatment could be involved in the increased TBX production.
At day 0, there was no significant difference between groups (Table 3) regarding the excretion of nitric oxide metabolites. Also, there was no difference with values observed in healthy nontransplanted subjects (see Methods for these values). Between day 0 and day 14, no significant change occurred in the placebo group. In the ticlopidine group, however, there was a significant increase in the 24-h urinary excretion of nitrite and nitrate when expressed as μmol/24 h (p < 0.005) or per mg of urinary creatinine (p = 0.05) (Table 3). When we compared the mean changes between days 0 and 14 in the two groups, there were also significant (p < 0.005 and p = 0.05) differences between the two groups.
Finally, a trend toward an inverse correlation was found between the variations in nitric oxide metabolites and the variations of the second wave of ADP-induced platelet aggregation (r = −0.40; p = 0.08), which suggested that the largest reductions in platelet aggregation in the patients receiving ticlopidine tended to be associated with the highest increases in nitric oxide generation. Also, the absence of modification of platelet function in the placebo group was associated with a lack of variation of nitric oxide. On the other hand, we found no relation (even no tendency) between the excretion of TBX metabolites and nitric oxide generation.
By using half the recommended daily dosage of ticlopidine (to reduce the risk of graft rejection), we observed in this study a significant reduction of platelet aggregation, which confirmed the efficacy of the drug at that dosage. In contrast, there was no significant effect on eicosanoids.
The major finding, however, is that the generation of nitric oxide was augmented in response to ticlopidine. It is noteworthy that, taking into account the kidney function, the data do not support the hypothesis that systemic nitric oxide production is altered in clinically stable cyclosporine-treated heart-transplant recipients, as their values were similar to those recorded in healthy subjects. The fact that the ticlopidine-induced changes in platelet function and nitric oxide production tended to be correlated suggests that both resulted from ticlopidine treatment. This is the first description of a possible novel property of this class of drugs. A dual beneficial effect on platelets and presumably on the endothelium may confer a unique advantage to ticlopidine as an antithrombotic agent, at least in transplant recipients.
Ticlopidine and the generation of endothelial and platelet eicosanoids
Cyclosporine is known to be toxic for endothelial cells (11,12) and, in heart-transplant recipients, this results in reduction of the generation of prostacyclin (6). Although dyslipidemia and increased platelet reactivity seem to be partly responsible for the accelerated atherosclerosis that occurs after transplantation and for the high prevalence of thromboembolic complications in these patients (7,22), deficient prostacyclin production may also play a role. In addition, even low-dose aspirin (which is often given to patients with transplants) was shown to reduce prostacyclin generation in patients with various arterial diseases (8). On the other hand, antiplatelet treatments have been ineffective in the prevention of experimental and clinical graft atherosclerosis (1,23), and platelets of heart-transplant recipients have been shown to be resistant to aspirin (4). Thus although further comparative studies are warranted, the lack of effect of ticlopidine on prostacyclin production and the lack of effect of aspirin on platelets suggest that ticlopidine should be preferred to aspirin, at least in these patients. However, the potential myelotoxic effect of ticlopidine should also be considered.
In this study, a nonsignificant tendency to a higher generation of TBX was observed in the ticlopidine group, suggesting that ticlopidine may stimulate platelet TBX production, as reported in the rabbit (24,25). It may be speculated that at standard (500 mg/day) dosage of ticlopidine, the TBX pathway could be more stimulated and more TBX generated. Clearly, this point should be reexamined in a further study in humans by using the standard dosage of ticlopidine. In addition to its well-known effects on platelet function, TBX is actually a powerful vasoconstrictor. In conditions such as heart transplantation or heart failure, in which high peripheral resistances are detrimental, a well-balanced production of TBX and prostacyclin may be critical. The use of aspirin in patients with heart failure has been criticized because of its possible deleterious effect on peripheral resistances by reducing the generation of the vasodilative prostacyclin (26). Recent data actually demonstrated that local generation of prostanoids in human vessels in vivo alters vascular tone (27). The fact that prostacyclin generation is not modified by ticlopidine is thus reassuring regarding its possible use in such fragile patients.
Ticlopidine and nitric oxide generation
Although confirmative studies in nonimmunosuppressed patients are warranted, our results suggest that ticlopidine actually increases the systemic production of nitric oxide. This is observed in the absence of any sign of subclinical infection or allograft rejection, known to induce nitric oxide synthase (28,29). In contrast to its effect on endothelial prostacyclin generation, cyclosporine was not shown to impair nitric oxide production in the rat (30-32). Also, O'Neil et al. (33) showed that cyclosporine treatment does not impair the release of nitric oxide in human coronary artery. Our data in humans indicate (for the first time, to our knowledge) that the systemic generation of nitric oxide is likely not altered by cyclosporine. Treated patients actually had similar values to those of healthy subjects not treated with cyclosporine. In addition, in this study, cyclosporine blood levels did not change after treatment, and the various biologic side effects induced by cyclosporine were not modified. Thus increased nitric oxide production in this study probably did not result from an interaction between ticlopidine and cyclosporine metabolisms. It is possible that ticlopidine per se (i.e., by acting directly on the endothelial cells, platelets, or monocytes) was responsible for the increased nitric oxide production in these patients. Further basic studies are needed to test this hypothesis.
Previous studies showed epicardial coronary endothelial dysfunction (34,35) or loss of myocardial microvascular response to acetylcholine (36) after cardiac transplantation and early endothelial dysfunction may predict the development of transplant coronary artery disease and acute cardiac complications (34-36). On the other hand, impaired local endothelial function has been related to abnormalities in the L-arginine-nitric oxide pathway, and recently, Drexler et al. (37) demonstrated that intravenous infusion of L-arginine improves endothelial function. Thus whereas it is not clear from our data whether ticlopidine improved the generation of nitric oxide by stimulating the endothelial cells, platelets, or monocytes, the drug may be useful in these patients by increasing systemic nitric oxide generation in the long term. Further studies are, however, needed to evaluate whether ticlopidine may affect coronary endothelial function in cardiac-transplant recipients.
Limitations of the study and clinical implications
No side effect was observed in this study. We indeed used half the usual dosage of ticlopidine so as not to interfere with the metabolism of cyclosporine (17,18). Thus future investigations should theoretically test the normal dosage. On the other hand, ticlopidine seems to have major antiplatelet effects in these patients who are almost constantly at risk of developing acute rejection. This may be a major problem for clinicians who have to decide to perform myocardial biopsy to ascertain the diagnosis of acute rejection before initiating the specific antirejection treatment. There is indeed a small but actual risk of myocardial wall perforation and subsequent tamponnade. In such a context, the use of a potent antiplatelet agent, such as ticlopidine at normal dosage, may be questionable, as no emergency antidote is available to neutralize the effect of ticlopidine. This difficulty likely explains why many physicians involved in the field of transplantation still prescribe low-dose aspirin with the hope of interfering with the pathophysiologic process of accelerated atherosclerosis, despite the apparent lack of efficiency of this treatment (1).
Regarding clinical implications, it may be of great interest in future studies to examine in parallel the effect of ticlopidine on endothelial function and on the generation of nitric oxide. It could indicate whether the effect of ticlopidine on the generation of nitric oxide is clinically relevant.
Finally, the interest for ticlopidine should not be restricted to heart-transplant recipients. Indeed, if the ability of ticlopidine to augment nitric oxide production is confirmed in nonimmunosuppressed patients, the drug could be very useful in conditions other than transplantation, for instance, in moderate congestive heart failure, in which a marked reduction in nitric oxide production was reported in a dog model (38), This means that increasing nitric oxide generation in patients with moderate congestive heart failure could actually result in a therapeutic advantage (39). Because the use of aspirin has been criticized in these patients (26), whereas antithrombotic treatment is necessary, it seems important to test whether ticlopidine may improve the overall prognosis in patients with heart failure and after heart transplantation.
1. Billingham ME. Cardiac transplant atherosclerosis. Transplant Proc
2. Forrat R, Ferrera R, Boissonnat P, et al. High prevalence of thromboembolic complications in heart transplant recipients. Transplantation
3. de Lorgeril M, Dureau G, Boissonnat P, et al. Platelet function and composition in heart transplant recipients compared with nontransplanted coronary patients. Arterioscler Thromb
4. de Lorgeril M, Dureau G, Boissonnat P, et al. Increased platelet aggregation after heart transplantation: influence of aspirin. J Heart Lung Transplant
5. de Lorgeril M, Richard MJ, Arnaud J, et al. Lipid peroxides and antioxidant defenses in the accelerated transplantation-associated coronary arteriosclerosis. Am Heart J
6. Bordet JC, de Lorgeril M, Durbin S, et al. Systemic but not renal production of prostacyclin is highly reduced in cyclosporin-treated heart transplant recipients. Am J Cardiol
7. de Lorgeril M, Boissonnat P, Mamelle N, et al. Platelet aggregation and HDL cholesterol are predictive of acute coronary events in heart transplant recipients. Circulation
8. Knapp HR, Healy C, Lawson J, FitzGerald GA. Effects of low-dose aspirin on endogenous eicosanoid formation in normal and atherosclerotic men. Thromb Res
9. Savi P, Herbert JM. ADP receptors on platelets and ADP-selective antiaggregating agents. Med Res Rev
10. Voss BL, Hamilton KK, Samara S, McKee PA. Cyclosporin suppression of endothelial prostacyclin generation: a possible mechanism for nephrotoxicity. Transplantation
11. Hattori Y, Kasai K, Emoto T, Hiraiwa M, Shimoda S. The inhibitory effect of cyclosporin on prostacyclin production by cultured endothelial cells from porcine aorta. Transplant Proc
12. Grace AA, Barradas MA, Mikhailidis DP, et al. Cyclosporin A enhances platelet aggregation. Kidney Int
13. Batt AM, Siest G, Magdalou J, Galteau MM. Enzyme induction by drugs and toxins. Clin Chim Acta
14. Fitzgerald GA, Pedersen AK, Patrono C. Analysis of prostacyclin and thromboxane biosynthesis in cardiovascular diseases. Circulation
15. Moncada S, Higgs A. The L/-arginine-nitric oxide pathway. N Engl J Med
16. Lopez-Farré A, Riesco A, Digiuni E, et al. Aspirin-stimulated nitric oxide production by neutrophils after acute myocardial ischemia in rabbits. Circulation
17. de Lorgeril M, Boissonnat P, Dureau G, Guidollet J, Renaud S. Evaluation of ticlopidine, a novel inhibitor of platelet aggregation, in heart transplant recipients. Transplantation
18. Boissonnat P, de Lorgeril M, Perroux V, et al. A drug interaction study between ticlopidine and cyclosporin in heart transplant recipients. Eur J Clin Pharmacol
19. Balazy M, Brass EP, Gerber JG, Nies AS. Facile method for preparation of 2,3-dinor-6-keto-PGF1α
, the major metabolite of prostacyclin. Prostaglandins
20. Conrad KP, Joffe GM, Kruszyna H, et al. Identification of increased nitric biosynthesis during pregnancy in rats. FASEB J
21. Rosselli M, Imthurn B, Keller PJ, Jackson EK, Dubey RK. Circulating nitric oxide (nitrite/nitrate) levels in postmenopausal women substituted with 17 beta-estradiol and norethisterone acetate: a two-year follow-up study. Hypertension
22. Vanrenterghem Y, Lerut T, Roels L, Gruwez J. Thromboembolic complications and haemostatic changes in cyclosporin-treated cadaveric kidney allograft recipients. Lancet
23. Hoyt G, Collin G, Billingham M, Miller DC, Jamieson SW. Effects of antiplatelet regimens in combination with cyclosporine on heart allograft vessel disease. J Heart Transplant
24. Dalal KB, Ebbe S, Mazoyer E, Carpenter D, Yee T. Biochemical and functional abnormalities in hypercholesterolemic rabbit platelets. Lipids
25. Sweatt DJ, Blair IA, Cragoe EJ, Limbird LE. Inhibitors of Na+
exchange block epinephrine- and ADP-induced stimulation of human platelet phospholipase C by blockade of arachidonic acid release at a prior step. J Biol Chem
26. Hall D, Zeitler H, Rudolph W. Counteraction of the vasodilator effects of enalapril by aspirin in severe heart failure. J Am Coll Cardiol
27. Bhagat K, Collier J, Vallance P. Vasodilatation to arachidonic acid in humans: an insight into endogenous prostanoids and effects of aspirin. Circulation
28. Yang X, Chowdhury N, Cai B, et al. Induction of myocardial nitric oxide
synthase by cardiac allograft rejection. J Clin Invest
29. Lewis NP, Tsao PS, Rickenbacher PR, et al. Nitric oxide synthase in the human cardiac allograft is associated with contractile dysfunction of the left ventricle. Circulation
30. Balligand JL, Godfraind T. Endothelium-derived relaxing factor and muscle-derived relaxing factor in rat aorta: action of cyclosporin A. J Cardiovasc Pharmacol
31. Verbeke M, Van de Voorde J, de Ridder L, Lameire N. Functional analysis of vascular dysfunction in cyclosporin treated rats. Cardiovasc Res
32. Roullet JB, Xue H, McCarron DA, Holcomb S, Bennett WM. Vascular mechanisms of cyclosporin-induced hypertension in the rat. J Clin Invest
33. O'Neil GS, Chester AH, Kushwaha S, Rose M, Tadjkarimi S, Yacoub MH. Cyclosporin treatment does not impair the release of nitric oxide in human coronary arteries. Br Heart J
34. Davis SF, Yeung AC, Meredith IT, et al. Early endothelial dysfunction predicts the development of transplant coronary artery disease at 1 year posttransplant. Circulation
35. Mills RM, Billett JM, Nichols WW. Endothelial dysfunction early after heart transplantation. Circulation
36. Treasure CB, Vita JA, Ganz P, et al. Loss of the coronary microvascular response to acetylcholine in cardiac transplant patients. Circulation
37. Drexler H, Fischell TA, Pinto FJ, et al. Effect of L-arginine on coronary endothelial function in cardiac transplant recipients: relation to vessel wall morphology. Circulation
38. Smith CJ, Sun D, Hoegler C, et al. Reduced gene expression of vascular endothelial NO synthase and cyclooxygenase-1 in heart failure. Circ Res
39. Koifman B, Wollman Y, Gobomolny N, et al. Improvement of cardiac performance by intravenous infusion of L-arginine in patients with moderate congestive heart failure. J Am Coll Cardiol