Cardiovascular Toxicities of Immunosuppressive Agents

Authors


Abstract

Cardiovascular disease is one of the major causes of morbidity and mortality following solid organ transplantation. Many of the current immunosuppressive drugs are associated with an increase of one or more risk factors for the development of atherosclerosis. This review compares the mechanism by which individual immunosuppressive agents may impact on these risk factors and the differential contributionof cyclosporine, tacrolimus, mycophenolate, azathioprine, and Rapamycin to these individual risk factors. Attention to the potential cardiovascular toxicities of individual immunosuppressive agents may help design strategies for maintenance of immunosuppression tailored to individual patients.

Introduction

Over the past 20 years, solid organ transplantation has progressed from an experimental procedure to perhaps the optimal treatment for end-stage disease of several organs. Advances in tissue preservation, surgical techniques, immunosuppression, recipient selection, and infection control have contributed to the improved short- and long-term survival rates for both grafts and patients following transplantation, as well as rates of acute rejection. Now that acute rejection is less of a concern than in the early years of transplantation, outcome measures are focusing on long-term morbidity and survival. Death with a functioning graft due to cardiovascular disease is the leading cause of death following renal and cardiac transplantation, and cardiovascular disease is one of the major causes of morbidity and mortality in all solid organ recipients (1–6).

Despite improvements in survival and acute rejection rates, long-term toxicity associated with immunosuppressive agents remains a concern. Most nonimmune toxicities of immunosuppressive therapy are dose related and cumulative. Long-term toxicity from immunosuppressive therapy increases morbidity, may affect patient and allograft survival, and adversely affects the recipient's quality of life. It may also decrease patient compliance with therapeutic regimens, increase the potential for adverse drug interactions if other medications are added to treat the toxicity, and represents significant added costs.

Donor and recipient age and genetic factors in graft survival are beyond the influence of physicians, but factors such as long-term toxic side-effects of some immunosuppressants can be ameliorated or avoided. Consequently, outcome measures now need to focus on the long-term toxicity and morbidity caused by some immunosuppressants.

An accelerated form of coronary disease, previously referred to as chronic rejection, is the leading cause of death in heart-transplant recipients > 1 year post transplantation (7). Several immunosuppressants are associated with well-documented risk factors for atherosclerosis, such as hypertension, dyslipidemia, and hyperglycemia.

This review assesses the contribution of the currently available immunosuppressive agents – cyclosporine (CsA), tacrolimus (FK506), steroids, azathioprine, mycophenolate mofetil (MMF), and sirolimus – to known risk factors for cardiovascular disease.

Hypertension

Hypertension occurs in 50% to 100% of transplant recipients (8–11). The prevalence of hypertension increases with time after transplantation, from 52% at 1 year to 77% at 4 years (10). Characteristics of patients more likely to develop post-transplantation hypertension are the following: age > 21 years, male gender, family history of hypertension, family history of cardiovascular complications, myocardial infarction, stroke, ischemic cause of heart failure, renal insufficiency, and large body size (10). Chronic hypertension is important, because it is a risk factor for decreased allograft survival, progressive renal failure, left ventricular hypertrophy, coronary allograft and peripheral vascular angiopathy, and stroke (12,13).

Cyclosporine

The principal cause of hypertension in post-transplantation patients is treatment with a calcineurin inhibitor: either CsA or FK506. Before the introduction of CsA with azathioprine-based immunosuppression, the incidence of post-transplantation hypertension was approximately 20%; now it is 60% to 90% (12,14). Patients receiving these drugs for nontransplant indications also have a high incidence of hypertension. There are multiple mechanisms by which CsA may induce hypertension (Table 1). CsA-associated hypertension (CAH), which may resolve when the drug is discontinued, generally develops within weeks to months after transplantation and is independent of nephrotoxicity. However, it may develop years after transplantation in association with renal dysfunction (8,12).

Table 1. : Potential mechanisms for cyclosporine-induced post-transplantation hypertension (8,9,14)
Activation of neurohormonal vasoconstrictors
Alterations in vascular reactivity
Renal tubular reabsorption of sodium in association with volume expansion
Alterations in regulation of intracellular calcium ions
Excess production of vasoconstrictors (prostaglandins, thromboxane, endothelin)
Decreased production of vasodilatory prostaglandins
Stimulation of the renin-angiotensin system

Mechanisms. Despite the prevalence and adverse clinical consequences of CAH, the precise mechanism(s) are unknown; however, several potential mechanisms have been proposed (Table 1) (8,9,14). Twenty-four-hour recordings after heart transplantation have demonstrated that, unlike essential hypertension, CAH may be common at night due to cardiac denervation, a persisting increase in systemic vascular resistance, and an absence of the normal decrease in heart rate and cardiac output seen in nontransplantation patients. This phenomenon is not observed in recipients of other organs. A second mechanism for CAH is vasoconstriction caused by an increased release of endothelin. CsA-induced nephrotoxicity has also been proposed as a mechanism of post-transplantation hypertension, but the rise in blood pressure does not correlate with the decline in renal function (8).

Evidence indicates that the immunosuppressive mechanism of CsA may also mediate its hypertensive effect (12,13,15,16). CsA inhibits T-cell activation by binding to a cytoplasmic-soluble protein, cyclophilin. Cyclophilin belongs to a group of isomerases known as immunophilins. Immunophilins are abundant in lymphoid tissues as well as other mammalian tissues (e.g. kidney, muscle, nervous system). The other important immunophilin is called FK-binding protein, which binds to both FK506 and sirolimus. The target of inhibition of CsA and FK506 is calcineurin, which has a pivotal role in preventing T-cell activation. Animal model studies have shown that CsA- and FK506-induced hypertension are a consequence of inhibition of calcineurin, which has a contributory role in mediating renal (vasoconstriction), vascular (inhibition of nitric oxide-induced vasodilation), and neural (increased glutamate release resulting in increased intracellular Ca2+) mechanisms of hypertension (12,13). The potential effect of these cellular mechanisms is suggested by the finding in one study that CsA stimulated vascular response to nerve stimulation in rats, which may increase renal and systemic vascular resistance and lead to hypertension (17). A study in humans, however, suggests that CAH is more likely due to an alteration in peripheral vascular function than to sympathetic neural activation, because baroreflexes were maintained in solid-organ transplant recipients (18). There are reports that both FK506 and CsA affect the contractility of peripheral vasculature, possibly suggesting that the well-known renal vascular toxicities of CsA and FK506 also manifest in the periphery. When human or rat resistance arteries are exposed to FK506 in vitro (at 1000 ng/mL), there is an increased sensitivity to norepinephrine and a decreased response to acetylcholine. Rats given FK506 at 6 mg/kg/day for 21 days had similar blood pressure to controls, but the response of the resistance artery to acetylcholine was significantly reduced. It was concluded that FK506 affects vascular hemodynamics by influencing smooth muscle relaxation (19). A study in dogs showed that CsA alters the balance of vasomotor tone by reducing the release of nitric oxide and increasing the release of endothelin in conductance and resistance coronary arteries (20). In rat and primate models of CsA-induced hypertension, it has been shown that blood pressure can be significantly lowered by administration of l-arginine or an endothelin receptor antagonist (21).

However, one study that examined forearm vascular tone in patients treated with CsA after heart transplantation has cast doubt on whether the hypertension is due to disturbances in nitric oxide production and release or changes in vasoconstrictor responses in the peripheral vasculature (22).

FK506

FK506 has also been shown to cause hypertension (Table 1) via mechanisms similar to those for CsA, but data suggest that fewer patients develop hypertension while taking FK506 than CsA (23,24). Approximately 50% of renal-transplant recipients receiving either FK506 or CsA reported hypertension post transplantation; the incidence was independent of race (25). However, in a study of renal-transplant recipients by Radermacher et al. (26), the renal arterial resistance index (a measure of renal vascular impedance in the vascular bed distal to the measured artery), serum creatinine, and need for additional antihypertensive therapy were significantly higher with CsA than with FK506 during the first 60 days of treatment. For the remainder of the 12-month follow-up period, renal effects were similar, because lower CsA trough values were associated with improved renal function. During episodes of rejection, renal resistance values increased with both drugs, but the increase was significant with FK506. In contrast, the recent trial comparing FK506 and CsA in de novo heart-transplant recipients demonstrated a significantly lower incidence of hypertension in patients receiving FK506 (71% vs. 48%, p = 0.05) (23).

Long-term follow-up of 1000 primary liver-transplant patients receiving FK506 as primary immunosuppression (27) demonstrated that hypertension occurred in approximately one-third of patients, and a retrospective study of 302 adult recipients of orthotopic liver transplants, with a median follow-up of 18 months, showed that nearly 45% of the patients met the criteria for arterial hypertension (28).

Initial experience has suggested that the incidence of hypertension is lower in liver-transplant recipients treated with FK506 than in those treated with CsA (28,29). The Guckelberger study (28) demonstrated a lower incidence of hypertension in patients treated with FK506 compared with those treated with CsA; however, this difference was not significant. Patients treated with FK506 in the Canzanello study (29) had significantly reduced rates of hypertension at 24 months post transplantation, and the onset of hypertension was delayed in those patients at both 1 and 12 months compared with patients receiving CsA (p < 0.05) (29).

Treatment with FK506 and MMF has also been shown to significantly reduce the rates of hypertension and to allow the withdrawal of steroids 14 days after liver transplantation. The incidence of hypertension in patients treated with FK506 and MMF in this prospective, randomized, open-label study was 12% compared with 30.3% in patients treated with CsA and MMF (p < 0.01) (24).

It should be noted, however, that it is difficult to compare the hypertensive effects of FK506 and CsA because of the apparent differences in their immunosuppressive abilities.

Azathioprine and mycophenolate mofetil

There are no data suggesting that either azathioprine or MMF contributes to the development of clinically significant increases in hypertension.

Treatment

Nonpharmacologic measures for lowering blood pressure in patients with essential hypertension, which include weight loss, exercise, cessation of smoking, limiting alcohol intake, and control of hyperlipidemia and diabetes mellitus, if present, may help limit development and/or progression of vascular disease. In heart- and renal-transplant recipients, there is evidence that CAH is salt-sensitive, particularly in the first month post transplantation. Hypertensive patients therefore should be instructed to limit dietary salt intake (10). Although few data support the use of one class of antihypertensive over another to treat CAH, the preferred agents are the calcium channel blockers and angiotensin-converting enzyme (ACE) inhibitors. A prospective, randomized, multicenter trial has compared the effectiveness of a calcium channel blocker vs. an ACE inhibitor to treat hypertension in heart-transplant patients (30). It was found that each agent controlled hypertension in less than 50% of patients when given as monotherapy, highlighting the fact that more than one type of drug is often required to treat the hypertension that develops after transplantation (30). There is some speculation that ACE inhibitors may exacerbate the CsA-induced decrease in glomerular filtration rate but may be preferred in patients with diabetes. In contrast, calcium antagonists have been shown to have a beneficial effect on renal blood flow and glomerular filtration rate (12,13). Beta-blockers also have been used; however, diuretics should be used with caution because their volume depletion, vasoconstriction, and neurohormone activation effects may adversely affect renal function and cardiac hemodynamics (8).

Dyslipidemia

Dyslipidemia, a common finding following transplantation, occurs in more than 80% of heart-, 60% to 70% of renal-, and 45% of liver-transplant recipients receiving immunosuppressive therapy (7–9,31–33). The specific lipid fractions that are abnormal vary among organ recipients, but increased low-density lipoprotein (LDL) cholesterol, decreased high-density lipoprotein (HDL) cholesterol, and increased triglyceride levels are common after heart transplantation. The magnitude of the increase in the total cholesterol concentration over pretransplantation levels may be significant, particularly in more critically ill patients, and may rise much higher, especially in the first 3–6 months post transplantation, if untreated (8). The greatest increases are usually noted in the first year post transplantation.

An adverse consequence of dyslipidemia in the nontransplant population is atherosclerosis, involving coronary, cerebral, and peripheral arteries. In the transplant population, dyslipidemia correlates with development of cardiac allograft vasculopathy, an aggressive form of atherosclerosis, as well as atherosclerotic vascular disease in nontransplant vessels (32). In addition, a 7-year post-transplantation study has correlated renal allograft failure with dyslipidemia (32,34).

Risk factors reported to contribute to the development of dyslipidemia in transplant recipients are age, proteinuria, obesity, antihypertensive therapy, prednisone dosage, pretransplantation hyperlipidemia, treatment with CsA, male gender, renal dysfunction, diabetes mellitus, and treatment with sirolimus (8,31–33,35–38), but the cumulative dose of corticosteroid appears to be the most significant risk factor. The combined use of corticosteroids and CsA appears to be an additive risk factor (31,33,39–41). Studies in kidney-, heart-, and liver-transplant recipients have consistently shown a reduced incidence of hyperlipidemia with FK506 (23,29,42–45). Sirolimus, however, has been found to be associated with significant increases in triglyceride and cholesterol levels (38,46). MMF, in contrast, is the only available immunosuppressive agent with no adverse effects on lipids.

Corticosteroids

Because of the strong association between corticosteroid immunosuppressive therapy and hyperlipidemia in post-transplantation patients, various corticosteroid regimens, including alternate-day regimens, early withdrawal, and steroid-free maintenance therapy, have been evaluated. A small study by Turgan et al. concluded that abnormal serum lipid levels can return to normal post transplantation, possibly due to alternate-day steroids (47). A study by Curtis et al. found that patients randomized to alternate-day steroids, as opposed to daily steroids, had a significant decrease in both serum triglyceride and serum cholesterol levels (48). The effects of alternate-day therapy on other outcome variables (e.g. rejection, cardiovascular disease, death) were not reported. Withdrawal of prednisone in kidney and kidney-pancreas transplant recipients treated with CsA and azathioprine (AZA) has been associated with a significant decline in total and LDL cholesterol, and there was a significant decline in HDL cholesterol as well, suggesting that cardiovascular risk related to dyslipidemia may not be reduced (49).

Keogh and other investigators evaluated maintenance regimens with and without a steroid component after heart transplantation (50–53). Patients receiving steroid-based immunosuppressive therapy (prednisolone, CsA, AZA) consistently had higher mean blood pressure readings and significantly higher total cholesterol levels over the 5-year observation period compared with patients who received only CsA and AZA (50). Despite these differences, steroid-related morbidity in the triple-therapy group was low. Other studies have shown that absolute increases in total and LDL cholesterol are greater in heart-transplant recipients who receive a steroid for maintenance therapy than in those who do not, but more importantly, lipid values post-transplantation correlate with pretransplantation values (54).

Mechanisms of steroid-induced dyslipidemia. It has been suggested that corticosteroid-induced dyslipidemia is a result of weight gain, which leads to insulin resistance, increased hepatic secretion of very low-density lipoprotein (VLDL), and increases in total cholesterol and triglyceride levels (9,31,39). Increases in VLDL, total cholesterol, and triglyceride levels and a decrease in HDL cholesterol level with corticosteroid immunosuppression are a result of several effects of corticosteroids: enhanced activity of acetyl-coenzyme A carboxylase and free fatty acid synthetase, increased hepatic synthesis of VLDL, down-regulation of LDL receptor activity, increased activity of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, and inhibition of lipoprotein lipase (32).

Cyclosporine

CsA also has been implicated as a causative factor in post-transplantation dyslipidemia. Studies evaluating the impact of immunosuppressive regimens including CsA have reported mixed findings. Some report that the magnitude of dyslipidemia is not as great with CsA-containing regimens as with regimens of prednisone and AZA, but the difference was due largely to significantly lower doses of steroids in CsA-containing regimens (33). One study, however, that eliminated the confounding effect of CsA dosage adjustments based on blood levels reported that serum cholesterol and triglyceride levels were higher with CsA-containing regimens and that CsA has an independent adverse effect on total cholesterol, LDL cholesterol, and triglyceride levels (55). Monotherapy with CsA for amyotrophic lateral sclerosis also revealed an adverse effect on lipids. During 2 months of CsA monotherapy, highly significant increases in serum cholesterol and LDL levels were noted compared with placebo. The effect was reversed on discontinuation of the drug (56).

Mechanisms of CsA-induced dyslipidemia. Several mechanisms for the metabolic effects of CsA have been postulated. Among these are inhibition of bile acid synthesis from cholesterol and transport of cholesterol to the intestines; binding of CsA to the LDL receptor, which increases LDL cholesterol levels; and marked reduction in post-heparin lipolytic activity with increased hepatic lipase activity and decreased lipoprotein lipase activity, which results in impaired clearance of VLDL and LDL cholesterol (32,57).

Binding of CsA to the LDL receptor via CsA-containing LDL cholesterol particles has been proposed as the mechanism of cellular uptake of CsA, since there is no specific CsA cell membrane receptor (58). Low cholesterol concentrations lead to increased concentrations of CsA in LDL cholesterol, which increases its uptake into cells that express LDL receptors. The increased intracellular concentrations of CsA are thought to interfere with DNA and RNA synthesis and thus cause adverse effects. CsA may also have a pro-oxidant effect on plasma LDL cholesterol, which may increase the risk of coronary artery disease (CAD), including the accelerated atherosclerosis seen in transplant recipients. In renal-transplant recipients, CsA concentrations in LDL cholesterol were shown to correlate with the oxidative susceptibility of LDL cholesterol; that is, higher CsA concentrations were associated with increased susceptibility of LDL cholesterol to oxidation (59).

FK506

Combination therapy with a corticosteroid and FK506 is also associated with an increase in total cholesterol and triglycerides (31). In liver-transplant recipients with pretransplantation serum lipid values within the normal range, significant increases in total cholesterol and triglycerides, as well as a marked increase in LDL cholesterol, were observed with immunosuppressive regimens that included FK506 and a corticosteroid or a corticosteroid, CsA, and AZA (60). Another study in liver-transplant recipients found that lipid parameters with FK506 treatment did not differ significantly from pretransplantation values and that increases in lipid parameters were significantly greater with CsA than with FK506 treatment (61). However, the differences in treatment effects may be accounted for by the significantly lower corticosteroid doses in patients receiving FK506 than in those receiving CsA. Similarly, in pancreas-transplant recipients, there were no significant changes in lipid parameters with FK506 treatment (62). Cholesterol and LDL cholesterol levels fell when these patients were switched from CsA to FK506 treatment, and the dose of corticosteroid fell as well. In heart-transplant patients, however, a randomized multicenter comparison of FK506 and CsA regimens has shown that patients on an FK506-based protocol had significantly lower serum LDL, HDL, and triglyceride levels (63). The effect of FK506 on the oxidizability of LDL cholesterol was similar to that associated with low CsA blood levels in renal-transplant recipients (59).

An analysis of the effects of immunosuppressive therapy in renal-transplant recipients found no differences in lipid parameters in FK506-treated patients, based on race (25). Total cholesterol and LDL cholesterol levels were significantly lower at 1 year in patients treated with FK506 than in those with CsA, regardless of race. In African Americans, triglycerides were significantly lower in FK506-treated patients than in those receiving CsA. In pediatric heart-transplant recipients, lipid parameters decreased in patients who were switched from CsA to FK506 (64).

Sirolimus

Hypertriglyceridemia and hypercholesterolemia have been noted following renal-transplantation in patients receiving sirolimus (38,65–67). In a pilot study of 11 patients, sirolimus dosing caused a rise in serum triglycerides and cholesterol. Following a reduction in dose or discontinuation of sirolimus, triglyceride and cholesterol levels decreased over 1–2 months. A weak but significant correlation was noted between high serum triglyceride levels and sirolimus whole blood trough concentrations (38). A dose-ranging study of sirolimus for treatment of acute rejection in 60 heart-transplant recipients showed significant dose-related increases in total cholesterol and triglyceride levels after only 2 weeks of drug exposure; levels did not return to normal until 4 weeks following discontinuation of the drug (46).

Azathioprine and mycophenolate mofetil

There are no data suggesting that either AZA or MMF causes clinically significant increases in any lipid fraction.

Treatment

Based on the time over which dyslipidemia develops, it is recommended that a lipid profile be performed at 3 and 6 months post transplantation (35). Treatment options for post-transplantation dyslipidemia include lowering the doses of corticosteroid and CsA (with a possible increase in risk of allograft rejection), conversion from CsA to FK506, and specific lipid-lowering therapy, usually a statin. Diet alone generally does not lower lipid levels sufficiently (9,31,35,68). Pharmacologic therapy is instituted by risk stratification with goals of < 200 mg/dL for younger, nondiabetic patients and < 180 mg/dL for patients with diabetes and LDL < 100 mg/dL. Kobashigawa et al. (69) have shown that empiric, pre-emptive lipid-lowering therapy with pravastatin has a significant beneficial effect on the incidence of allograft CAD, as measured by intravascular sound myocardial events, graft failure, or survival (7,69–71). The design and results of the study by Kobashigawa et al. were duplicated recently using simvastatin (72).

Inhibitors of HMG-CoA reductase (statins) are the preferred lipid-lowering therapy in transplant recipients, because they are highly effective, well tolerated (although rhabdomyolysis may occur), and may protect graft function and/or slow the atherosclerotic process (31,68,70,73). In nontransplant patients, the marked favorable effects of statins on plasma lipid concentrations are associated with significant reductions in the incidence of myocardial infarction and death from cardiovascular causes (74). In several studies of small numbers of heart- or renal-transplant recipients, statins had beneficial effects on the lipid profile and generally were well tolerated (75–79).

Importantly, statins appear to also have an immunosuppressive effect in transplant recipients, as evidenced by reductions in the incidence of acute and recurrent rejection episodes (78–82). Moreover, in heart-transplant patients, 1-year survival and the incidence of coronary vasculopathy were significantly lower in statin-treated patients (78). The favorable effects of statin treatment on rejection did not correlate with cholesterol levels (78–82).

Mechanisms by which statins may enhance immunosuppression include inhibition of chemotaxis by monocytes, inhibition of vascular smooth muscle cell proliferation, and decreased natural killer cell cytotoxicity (73,78–80). The atherosclerotic process may be slowed by statins via reduced proliferation of endothelial myocytes, as evidenced by slower intimal thickening in treated vs. control patients and in an animal model of severe vascular immune injury (73,78,81,82).

Bile acid sequestrants and nicotinic acid generally are not considered first-line cholesterol-lowering therapy in transplant recipients because of side-effects and drug interactions, and gemfibrozil and probucol are not preferred because their lipid-lowering effects are modest (31,35,73). Some investigators have reported significant reductions in total cholesterol, LDL cholesterol, and triglyceride levels in renal-transplant recipients receiving gemfibrozil therapy (83). Fenofibrate lowered total cholesterol, LDL cholesterol, and triglyceride levels by about 20% in one study of heart-transplant recipients, but was associated with a 67% withdrawal rate because of side-effects (71).

Drug interactions with immunosuppressive agents and lipid-lowering drugs

Although several agents are effective in reducing lipids, most have some serious interactions, and this may compromise their use in transplant recipients. CsA, FK506, and, to an as-yet-undefined extent, sirolimus are metabolized via the action of cytochrome P450 (CYP450) isozymes. Plasma concentrations of these drugs may be affected by the concomitant administration of medications that inhibit CYP450 metabolism.

Bile acid sequestrants may interfere with the absorption of CsA, and the side-effects of nicotinic acid may be exacerbated by immunosuppressive drugs (35). Coadministration of probucol and CsA in heart-transplant recipients significantly increased the clearance of CsA, with corresponding decreases in CsA plasma concentrations, area under the plasma concentration-time curve (AUC), and half-life (84). Whether the probucol-induced alterations in CsA pharmacokinetics were caused by gastrointestinal drug loss or altered tissue distribution of CsA was not evaluated.

The safe use of statins requires agent-specific modification of dose. The risk of rhabdomyolysis with statins may be increased by concomitant therapy with CsA or gemfibrozil and nicotinic acid, the risk varying with the lipophilicity of the statin (70,73,85,86). Early estimates of the incidence of myopathy with lovastatin were 0.2% if neither immunosuppressive nor fibric acid derivative therapy was given, 5% with concomitant fibric acid derivative therapy, and 30% with concomitant immunosuppressive therapy (including CsA) (85). The increased risk of rhabdomyolysis is probably related to altered disposition of statins, with the resultant increase in systemic exposure. In heart-transplant recipients receiving CsA and pravastatin, the half-life of pravastatin was four times longer and the pravastatin AUC was 20 times higher in transplant than in nontransplant patients (87). Similar alterations in lovastatin pharmacokinetic parameters have been described in heart-transplant patients (85). One of the newest statins, cerivastatin, has a very long half-life and has been associated with very severe rhabdomyolysis when given with CsA, especially in combination with gemfibrozil (88,89). Its use should be avoided in transplant recipients.

Diabetes

Prior to the introduction of CsA, when high doses of steroids were the principal immunosuppressive agents, permanent post-transplantation diabetes mellitus (PTDM) developed in up to 25% of renal-transplant recipients, and temporary diabetes in 22% (90). Although the doses of steroids presently used are much lower, the use of pulsed steroids to treat episodes of rejection has been identified as a critical factor in the onset of PTDM, and the risk of developing the condition has been calculated at 5% per 0.01 mg/kg/day of increase in prednisolone dose (91,92). In vitro and in vivo preclinical data have shown that CsA and FK506 negatively impact pancreatic function, and this has been shown to have a clinical impact, with FK506 having the more pronounced diabetogenic effect (93–95). Recently, Drachenberg et al. conducted a study of 26 allograft biopsies, performed 1–8 months post transplantation (96). The patients (20 simultaneous kidney-pancreas-transplant recipients) were randomized to receive either FK506 or CsA. A microscopic analysis of islet morphology was correlated with mean and peak serum levels of CsA and FK506, corticosteroid administration, and glycemia. The investigators found that toxic levels of CsA or FK506 coupled with higher doses of steroids potentiate their overall diabetogenic effects.

Reports on the incidence of post-transplantation de novo diabetes in liver-transplant recipients are inconsistent. Long-term data reported by Sheiner et al. (97) suggest a greater prevalence of medication-dependent diabetes post transplantation than pre transplantation, while other studies have shown no significant differences in de novo diabetes (98).

Coronary Artery Disease

All the risk factors previously mentioned are independent risk factors for coronary artery disease (CAD), and the composite risk profile is shown in Table 2. These are probably additive as well as independent risk factors. The incidence of CAD in heart-transplant recipients, which ranges from 1% to 18% at 1 year and from 20% to 50% at 3 years (32), has not decreased in the past 25 years, despite improvements in immunosuppressive regimens (99).

Table 2. : Summary of impact of immunosuppressive agents on cardiovascular risk factors
 HypertensionLipidsDiabetes mellitus
  1. NC, no contribution.

Cyclosporine↑↑↑↑↑
FK506↑↑↑↑
Steroids↑↑↑↑↑
SirolimusNC↑↑↑NC
Mycophenolate mofetilNCNCNC

An unusually accelerated form of CAD, cardiac allograft vasculopathy (CAV), is a long-term complication following cardiac transplantation. This disease affects both intramural and epicardial coronary arteries and veins. Vascular injury seen in CAV is proposed to be induced by a number of stimuli, including immune response to the allograft, ischemia reperfusion injury, viral infection, immunosuppressive drugs, and the risk factors previously described in this review (100). CAV is a major cause of graft failure in long-term transplant survivors and of overall morbidity and mortality.

Allograft vascular disease occurs with approximately the same frequency in renal allografts and is less common following liver and lung transplantation (99). In renal-transplant recipients, cardiovascular disease events develop in about 20% of patients by 15 years post transplantation (33). In heart and renal allografts, the arteriopathy is an endothelial process, whereas in hepatic and lung allografts, it is epithelial (99). The pathophysiologic mechanisms underlying transplant vascular disease and arteriopathy have not been elucidated, but include both antigen-dependent and antigen-independent factors (99,101,102). Common to the various mechanistic hypotheses is arterial injury (nondenuding), to which arterial-wall components respond (101). Cell-mediated immunity, humoral immunity with antibody-mediated injury, and metabolic abnormalities such as dyslipidemia that may be induced by immunosuppressive therapy have been proposed as response mechanisms.

Intramural coronary artery vasculitis and vascular- or humoral-mediated rejection with hemodynamic compromise are predictive of poor cardiac graft survival (103,104). The role of cellular mechanisms was suggested in a study of failed human heart allografts, which found that arteriopathy began as early as 2 weeks after cardiac allograft implantation as a T-cell predominant process with progressive intimal thickening and narrowing of the lumen (103). Smooth muscle cells, CD4 lymphocytes, and macrophages that accumulate below the endothelial cell layer are thought to stimulate release of growth factors and cytokines to produce allograft vascular disease (99). Other investigators have shown that coronary segments with progressive intimal thickening may not show compensatory enlargement, but actual shrinkage of lumenal area (remodeling) over time, which is similar to findings following mechanical injury from percutaneous transluminal coronary angioplasty (PTCA) (105). In fact, lesions in patients who develop restenosis after PTCA are nearly identical histologically to allograft vascular lesions (99). Ischemic damage to vessels during harvesting and transplantation procedures and aggravation of the ischemic injury by CsA have also been thought to contribute to the intimal thickening and vasculopathy observed after transplantation (106).

Risk factors for atherosclerotic disease in heart-transplant recipients, particularly those with a history of CAD pre transplantation, are the development of hypertension and hypertriglyceridemia, low HDL and high LDL cholesterol, obesity, and ischemic etiology pre transplantation (107). If only one of these factors was present, the increase in risk of peripheral vascular disease was 14%, whereas the presence of all three increased the risk by 60%. All three contribute to CAD in transplant recipients, particularly hyperlipidemia (32,68,101,103). Several mechanisms may account for the increased accumulation of lipids in allograft recipients (Table 3) (102).

Table 3. : Potential mechanisms for increased accumulation of lipids in allograft recipients (102)
  1. LDL, low-density lipoprotein.

Endothelial injury/disturbances in endothelial regulation of LDL cholesterol transport
 Accelerated transcytosis of LDL cholesterol and other molecules
 Altered lipid permeability with increased lipid transport across
 arterial wall
Impaired ability of allograft arterial cell walls to remove LDL cholesterol
 May increase LDL cholesterol interactions with extracellular
 matrix of arterial wall
Increased lipid synthesis within atherosclerotic lesion
Dyslipidemia induced by immunosuppressive therapy
 Corticosteroids
Cyclosporine

Evidence supporting the contributory role of dyslipidemia in allograft vascular disease includes a strong association between low HDL cholesterol and high triglyceride levels and intimal thickening/CAD at 1 year in heart-transplant recipients (101). Other studies in heart-allograft recipients have demonstrated that although serum lipid levels were not markedly elevated, coronary tissue lipid levels were 10–20 times higher than normal native artery limits in age-comparable donor vessels (102). In addition, there was marked intracellular and extracellular deposition of lipids within the superficial and deep intima and media of narrowed epicardial coronary arteries in these patients. The extent of luminal narrowing strongly correlated with the concentration of lipids in the arterial wall.

To the extent that dyslipidemia has a role, the development of allograft vascular disease may be slowed or prevented by use of calcium antagonists or lipid-lowering therapy, which seem to have protective effects (78,79,101,106). As noted previously, statins significantly lowered the incidence of coronary vasculopathy in heart-transplant recipients (78,108). In renal-transplant recipients, diltiazem prevented the increase in arteriole wall thickness noted at 3 months in patients receiving only standard immunosuppressive therapy (106). The dose of CsA in diltiazem-treated recipients was significantly lower than that in patients not receiving the calcium antagonist (3.5 ± 0.9 vs. 6.0 ± 1.1 mg/kg/24 h), which may reflect the ability of diltiazem to inhibit the metabolism of CsA.

Corticosteroids

High levels of plasminogen activator inhibitor (PAI-1), which increase the risk of thrombosis, have been found in atherosclerotic plaques (109,110). Those high levels are an independent risk factor for cardiovascular disease and are predictive of recurrent myocardial infarction and sudden death. Patients with hypertriglyceridemia, diabetes mellitus, hypertension, or insulin resistance syndrome or patients who are overweight often have high PAI-1 levels. Transplant recipients often have one or more of these characteristics and are at risk for thrombotic complications following transplantation. In a study of heart-transplant recipients, basal and stimulated fibrinolytic capacity was more reduced (and thus thrombotic risk increased) in patients who received a corticosteroid as part of a three-drug immunosuppressive regimen than in those receiving only CsA and AZA (109). Significantly higher PAI-1 antigen and activity levels were noted in corticosteroid-treated patients vs. patients not receiving a corticosteroid. This finding suggests a greater risk of thrombotic complications in transplant recipients who have unstable atherosclerotic plaques. These effects are in addition to the primary role of hyperlipidemia.

Cyclosporine

Although the addition of CsA to immunosuppressive regimens has resulted in a significant decrease in the incidence of acute rejection in heart-transplant recipients, the incidence of allograft CAD in CsA- vs. AZA-treated recipients was not shown to be significantly decreased during a 5-year follow-up period (111). Similarly, a stepwise logistic regression analysis of 163 consecutive heart-transplant patients found that an average daily CsA dose > 4.5 mg/kg/day appeared to protect the recipient from the development of comorbid factors resulting from cardiac allograft vasculopathy (112). In another study, multivariate analysis performed on 163 heart-transplant recipients found that a low mean CsA dose (< 4 mg/kg/day) significantly predicted the development of intimal thickening and CAD (113).

Studies in animals have attempted to elucidate the role of CsA in allograft vascular disease. A study in rabbits suggested CsA has direct effects on the development of proliferative vasculopathy, not wholly as a result of lipid deposition (114). CsA and hypercholesterolemia promoted coronary vasculopathy in native or graft coronary arteries in rabbits that underwent heterotopic heart transplantation; CsA promoted fibrous intimal hyperplasia, and hypercholesterolemia promoted fatty proliferative lesions. CsA administration also led to hypercholesterolemia in rabbits fed a normal diet and may contribute to hypercholesterolemia in cholesterol-fed rabbits. However, marked reductions in plasma cholesterol levels did not prevent the development of fibrous intimal hyperplastic vasculopathy.

Endothelialitis, a subendothelial inflammatory response, is an early feature in the pathogenesis of accelerated arteriosclerosis. In a rat aortic allograft model, 1–2 months of treatment with CsA, but not AZA or corticosteroids, induced endothelialitis in the intima (115). Based on these findings, accelerated atherosclerosis is thought to be due, in part, to CsA's effect on the endothelium, causing endothelial cell damage that leads to the accumulation of inflammatory cells and release of endothelin and other growth factors, which in turn mobilize smooth muscle cells to proliferate (115,116). An in vitro study suggests that the cellular effects of CsA may be dose-dependent. A CsA dose corresponding to the trough concentration, which has the greater potential for inducing vascular damage, was shown to induce DNA synthesis and proliferation of smooth muscle cells, whereas a dose corresponding to the peak concentration induced DNA synthesis but not cell proliferation (117).

CsA has also been shown to increase the plasma levels of PAI-1, leading to decreased fibrinolytic activity (118). In a small, prospective, controlled study of renal transplantation patients initially receiving steroids and CsA or AZA, a substantial increase in the activity of the fibrinolytic system was observed in patients who were randomized to convert from CsA to AZA at 6 months post transplantation. This increase in fibrinolytic activity at 12 weeks after conversion to AZA was accompanied by a significant decrease in the plasma concentrations and activity of PAI-1 (p = 0.016 and 0.009, respectively) (119).

FK506

Vasculitis has been reported with FK506 in renal transplantation animal models (120). Coexisting histologic changes included hypertrophy of the vessel wall as a result of edema or fibrosis, cell infiltration, and fibrinoid necrosis of the wall. The vascular changes were not found in all vessels of the heart or throughout the total area of any artery; rather, they were focal and spotty. Vasculitis has also been reported in humans receiving FK506 (121). Autopsy findings from a pediatric liver-bowel-transplant recipient who died 3 weeks after transplantation showed arteritis and extensive calcification of cardiac tissue. These findings may be related to effects by FK506 on calcium channels in cardiac and striated muscle. Significant cardiomyopathy with progression to severe hypertrophic obstructive cardiomyopathy and congestive heart failure in some patients has also been noted in liver and/or bowel-transplant recipients receiving FK506 (120). The hypertrophy was partially reversed by lowering the dose or discontinuing FK506.

The mechanism of the immunosuppressive effects of FK506 is similar to that of CsA, but the cardiotoxicity profile of FK506 appears to be more favorable. Greater experience is necessary to fully evaluate the cardiotoxic potential of FK506.

Sirolimus

Whereas CsA causes contraction of small arteries, probably by damaging the endothelium, with subsequent imbalance in the synthesis/release of vasodilatory and vasoconstricting substances, sirolimus has a vasodilatory effect: endothelium-dependent vasodilation is increased (122). Endothelial injury with subsequent alterations in function and impaired endothelium-dependent vasodilation has been shown to be the pathophysiologic mechanism underlying hypertension and the cardiovascular disease process (123). Sirolimus also inhibits the development of intimal hyperplasia and transplant arteriosclerosis in aortic grafts by allowing endogenous gene expression of nitric oxide synthase, the enzyme involved in the synthesis of nitric oxide, a vasodilatory substance that also inhibits proliferation of smooth muscle cells (124). Recently, the use of sirolimus-coated stents in human coronary arteries has been found to result in reduction of neointimal proliferation compared with controls (125). This result needs to be confirmed in a placebo-controlled trial.

Vasculitis of the gastrointestinal tract has been noted in experimental models of renal and heart transplantation with sirolimus (120,126). Vasculitis was also increased by the immunologic events of rejection. However, the combination of sirolimus and CsA reduced the frequency and severity of vasculitis compared with sirolimus alone (126). Myocardial necrosis was also reported in a rat model (127). It remains to be seen whether the effect of sirolimus on hyperlipidemia will lead to an increase in CAD when it is used for longer periods.

Azathioprine and mycophenolate mofetil

Azathioprine and MMF (128–130) have the lowest adverse cardiovascular risk/toxicities of the currently available immunosuppressive agents. This has recently been highlighted for MMF in the results of the latest cardiac trial (131). Both AZA and MMF act by inhibiting purine synthesis, although the functional selectivity of MMF on lymphocytes allows the proliferation of these cells to be specifically targeted. There were no significant differences between MMF and AZA treatment with respect to development of new CAD or progression of pre-existing CAD. However, intracoronary ultrasound demonstrated that the mean change in lumen area from baseline to 12 months showed a beneficial increase with MMF treatment and a decrease with AZA treatment. The data at 3 years show no significant differences in any of the intravascular ultrasound parameters between the two groups (change from 2 months vs. 36 months), but MMF-treated patients show a numerical advantage in each parameter compared with AZA-treated patients. In addition, MMF-treated patients had fewer fatal cardiovascular events and less autopsy-proven atherosclerosis. Three-year graft survival was significantly higher in the MMF-treated patients than in the AZA-treated patients (p = 0.029), (132). Clearly, if these findings are sustained over the long-term, MMF represents a potential advantage in the control of cardiac allograft vasculopathy.

In vitro studies and studies in animal allograft models confirm that MMF is not likely to cause the accelerated atherosclerosis/CAD associated with CsA regimens. MMF inhibits the formation of adhesion molecules that facilitate the attachment of leukocytes to endothelial and target cells and the interaction of lymphocytes with other cell types, thereby inhibiting ongoing rejection. It is predicted that this effect will disorganize lymphocyte homing and reduce the movement of lymphocytes and monocytes into the regions of chronic inflammation typically found in organs that are being rejected. In addition, clinically attainable concentrations of mycophenolic acid (1–10 µ&mgr;) have been shown to inhibit the proliferation of human smooth muscle cells in culture. These findings provide evidence that MMF may inhibit proliferative arteriopathy (133).

In a rat aortic allograft model, intimal proliferation was inhibited by MMF, suggesting that it may inhibit vasculopathy. MMF also prevented vascular changes associated with chronic rejection in a hamster-to-rat cardiac xenograft (134).

It has been reported that MMF increases the incidence of cytomegalovirus (CMV) infection (135). CMV has been associated with CAV (136) and is a major risk factor for human heart and kidney graft loss experimentally (137,138). Due to the consequences of CMV infection on both morbidity and graft survival, strategies employing the prophylactic use of oral antiviral agents have been initiated. Giral et al. (139) examined the impact of CMV infection treated with ganciclovir on long-term graft survival in 126 consecutive kidney-transplant recipients who had received antirejection prophylaxis with MMF, compared with 319 consecutive kidney-transplant patients who had received AZA. The incidence and time to development and treatment of CMV disease in patients treated with MMF or AZA were similar in both groups, and all patients were treated with ganciclovir for at least 14 days. However, CMV disease was found to be strongly associated with graft loss only in the AZA group. Graft survival at 1 year was significantly increased in patients treated with MMF, compared with those treated with AZA, but only in patients who developed CMV during follow-up (77% vs. 90%; p < 0.02).

Conclusion

Cardiovascular toxicities such as hypertension and hyperlipidemia are major clinical management issues following transplantation. Over the past 10 years, the approval of FK506, MMF, and sirolimus has expanded the range of immunosuppressive agents available to treat transplantation patients. The toxicity profiles of these very effective agents offer the opportunity for the development of new regimens that may reduce the incidence of cardiovascular events in patients over the long-term.

Ancillary