Liver transplantation is the standard care for end-stage liver disease and for fulminant hepatic failure. During recent years, advances in both medical management and surgical techniques have led to an increase in graft and patient survival following orthotopic liver transplantation (OLT). Nowadays, patient survival after OLT is about 85% at 5 years1 and 70% at 10 years.2
Early mortality (within 1 year after OLT) has been reduced. The major causes of early mortality are infections, surgical complications, allograft failure due to primary allograft dysfunction, and allograft rejection. Late complications are increased because of the enhancement of survival after OLT. Major late complications are cardiovascular events, metabolic complications, renal dysfunction, disease recurrence, and malignancy.1
Arterial hypertension is one of the major late complications after liver transplantation and is an important risk factor for cardiovascular events. De novo arterial hypertension is defined as the development of arterial blood pressure > 140/90 mm Hg in patients that were normotensive before OLT.3De novo arterial hypertension has a prevalence exceeding 50% after liver transplantation, and several reports have suggested that the real prevalence of this complication is about 75%–85%.1, 4
Mechanisms underlying de novo arterial hypertension are several and complex. The main cause is treatment with a calcineurin inhibitor (CNI), either cyclosporine A or tacrolimus. The major cause of CNI-related hypertension is the impairment of the endothelium control of vascular tone. In fact, CNIs reduce the endothelium release of vasodilator factors, such as prostacyclin and nitric oxide, and enhance the release of vasoconstrictor factors, such as endothelin and thromboxane. The net effect of these changes is an imbalance that promotes widespread arterial vasoconstriction.5
A CNI-related activation of sympathetic neural pathways and the renin-angiotensin system and a CNI-related decrease of renal sodium excretion are currently considered to be less important in the development of de novo arterial hypertension after OLT.
Up to now, standard care for the treatment of de novo arterial hypertension after liver transplantation has not been established. Because of its pathogenesis, a vasodilator agent may represent the first-choice drug. Calcium channel blocking agents are the preferred class of antihypertensive drugs because of their efficacy at smooth muscle vasodilation. Among them, the dihydropyridine class is preferred because other compounds such as verapamil and diltiazem interfere with the removal of CNIs and may lead to accumulation of these drugs with potential toxic effects.6
Other antihypertensive agents have been used in the treatment of de novo arterial hypertension after OLT. In particular, some studies have suggested a potential role for β-blockers in this clinical scenario. Carvedilol is an α1-β-blocker with vasodilator properties. Recent studies have suggested its ability to reduce the oxidative stress induced by CNIs at the endothelium vessel wall in renal transplant recipients.7, 8
Therefore, the aim of this study was to compare nifedipine, a calcium channel blocker, and carvedilol in the treatment of de novo arterial hypertension after liver transplantation.
This was a prospective, randomized, open trial comparing nifedipine and carvedilol in the treatment of de novo arterial hypertension after liver transplantation. From January 2001 to January 2005, 73 consecutive outpatients with a diagnosis of de novo arterial hypertension were included in the study. The diagnosis was based on the finding of abnormally high values of arterial pressure in recipients of liver grafts who had never suffered from arterial hypertension before transplantation. For the purpose of the study, arterial hypertension was defined as ambulatory blood pressure values > 140/90 mm Hg in nondiabetic patients9 and > 130/80 mm Hg in diabetic patients. Ambulatory blood pressures were measured with patients in the sitting position after 10 minutes of rest at the brachial artery with a mercury sphygmomanometer. The mean of 3 consecutive measurements was considered the value of the blood pressure. In all patients, the diagnosis of arterial hypertension was confirmed by 24-hour blood pressure monitoring.10, 11 Exclusion criteria were contraindications to the use of the drugs that were provided in the study or to the use of drugs that may affect arterial pressure and/or renal function, with the exception of immunosuppressive drugs such as steroids and CNIs, the presence of severe renal failure (creatinine ≥ 2.5 mg/dL) at the time of randomization, the presence of graft dysfunction (total serum bilirubin > 3 mg/dL and international normalized ratio > 1.5), and the development of de novo arterial hypertension during the period of post-OLT intensive care, that is, when antihypertensive therapy was requested.
Patients included in the study were randomly assigned to receive either nifedipine (group A) or carvedilol (group B). Patients in group A received a long-acting gastrointestinal transport system formulation of nifedipine at the initial dose of 30 mg once a day, whereas patients in group B received carvedilol at the initial dose of 12.5 mg twice a day. When a patient developed severe adverse effects to the randomized drug, the treatment was stopped, and the patient received the other drug (crossing-over). When the value of the arterial blood pressure was not normalized after 4 weeks, the doses of nifedipine and carvedilol were doubled to 60 mg once daily and 25 mg twice daily, respectively. If the arterial blood pressure was not normalized with the highest dose of the assigned drug, patients received a combined therapy by the addition of ramipril, an angiotensin converting enzyme (ACE) inhibitor, to the first drug. Ramipril was added at the initial dose of 2.5 mg once daily, which was increased to 5 mg once daily when required. The response to treatment was defined as complete when the arterial pressure was reduced to values ≤ 140/90 mm Hg in nondiabetic patients and ≤ 130/80 mm Hg in diabetic patients. The response to treatment was defined as incomplete when the arterial pressure was reduced compared to the baseline but was still higher than 140/90 mm Hg in nondiabetic patients and higher than 130/80 mm Hg in diabetic patients.
For the purpose of the study, patients were defined as intolerant to nifedipine or carvedilol when they developed adverse effects that were intolerable enough for patients to represent a problem for their compliance with therapy. Patients were classified as full or partial responders when a complete or incomplete response to nifedipine or carvedilol was detected, respectively.
Partial responders to nifedipine or carvedilol received the combined-therapy treatment. The study was approved by the local ethics committee, and all the patients gave their written informed consent.
Patients were monitored monthly until 1 year after the inclusion. Each visit included the measurement of the ambulatory blood pressure and heart rate (HR), as previously reported. In addition, plasma samples were collected for the detection of routine laboratory parameters such as total bilirubin, alanine aminotransferase, aspartate aminotransferase, gamma glutamyl transpeptidase, cholesterol, triglycerides, uric acid, blood glucose level, and trough levels of cyclosporine and tacrolimus. After 6 and 12 months from the inclusion, the 24-hour blood pressure monitoring was also repeated.
All patients received a CNI as basal immunosuppressive therapy. During the study, the blood trough level of tacrolimus was maintained within 10–15 μg/L during the first month after OLT and within 5–10 μg/L over the long term. The blood concentration of cyclosporine 2 hours after oral ingestion was maintained within 800–1000 μg/L during the first month after OLT and within 300–600 μg/L over the long term. Corticosteroids were withdrawn within 6 months after OLT in all patients. Diuretics, azathioprine, and mycophenolate mofetil were not used.
Evaluation of Renal Function
Before randomization, samples of plasma were collected to estimate levels of urea, creatinine, electrolytes, and plasma renin activity (PRA). The glomerular filtration rate (GFR) and effective renal plasmatic flow (RPF) were determined by the clearance of inulin and para-aminohippurate, respectively. Plasma urea, creatinine, sodium, and potassium were re-evaluated monthly during each visit. GFR, inulin, and PRA were re-evaluated 6 and 12 months after inclusion.
A power calculation could not be performed because this is the first controlled clinical trial on this issue in liver transplant recipients. The primary endpoint was the response to antihypertensive monotherapy (nifedipine or carvedilol), which was classified as complete, partial, or intolerant. Patients were analyzed either as randomized (group A and group B) or after the crossing-over (group C and group D). Differences in the response rates between the 2 groups were evaluated by means of the cumulative logit analysis. A P value of less than 0.05 was considered to be indicative of statistical significance. Categorical data were presented as number and percentage and compared with Fisher's exact test or the χ2 test, whichever was more applicable. Continuous variables were expressed as means ± standard deviation, unless specified otherwise, and analyzed with the Student t test or Mann-Whitney test. All statistical analyses were performed with SAS software (version 8.2).
Fifty of 73 patients that were considered for the study were randomized. Twenty-three patients were excluded: 10 patients for severe renal failure, 6 patients for the development of arterial hypertension during the intensive care period after OLT, 5 patients because they were hypertensive before the development of end-stage liver disease and so before OLT, and 2 for graft dysfunction. Twenty-five patients were randomized to receive nifedipine, and 25 patients were randomized to receive carvedilol. There was a case of dropout in group B, so that the final number of patients who received carvedilol was 24. The reason for the dropout in group B was the withdrawal of the consensus. During the follow-up, 12 patients in group A (48.0%) and 3 patients in group B (12.5%) developed serious adverse effects (P < 0.01). In group A, 7 patients developed serious adverse effects within the first month of treatment, and 5 developed serious adverse effects between the first and sixth months. In group B, 2 patients developed serious adverse effects between the first and sixth months of treatment, and the other patient developed serious adverse effects after the sixth month. As a result, they stopped the randomized drug and were classified as intolerant patients. Subsequently, they received the other drug. There was another dropout in the group of patients that developed adverse effects to nifedipine before the crossing-over (Fig. 1); as a result, the total number of patients who experienced either nifedipine or carvedilol was 28 (group C) and 35 (group D), respectively. The adverse effects observed during the study are summarized in Table 1.
Table 1. Adverse Effects of Antihypertensive Monotherapy by Randomization Treatment Group
Nifedipine (n = 25)
Carvedilol (n = 24)
Adverse effects [n (%)]
Lower limb edema
Baseline Clinical and Laboratory Features of Patients
The baseline clinical and laboratory features of patients of groups A and B are reported in Table 2. The 2 groups were similar in age, gender, indications for liver transplantation, time of development of de novo arterial hypertension, prevalence of diabetes mellitus, prevalence of patients who received cyclosporine or tacrolimus, and prevalence of patients who also received steroids at inclusion. In addition, baseline levels of the mean arterial pressure (MAP), HR, GFR, RPF, and PRA, trough levels of CNIs, and mean doses of CNIs did not differ between the 2 groups.
Table 2. Patients' Clinical and Laboratory Characteristics at Baseline by Randomization Treatment Group
Response to Antihypertensive Monotherapy at the First Month
Four patients in group A (16%) and 3 patients in group B (12.5%) were classified as full responders [16% versus 12.5%, P = not significant (NS)]. Partial responders numbered 14 in group A (56%) and 21 in group B (87.5%; P < 0.01).
With the crossing-over taken into account, there were 4 (14.28%) full responders in group C and 6 (17.14%) in group D (P = NS). There were 17 (60.7%) partial responders in group C and 29 (82%) in group D (P < 0.01). The values of arterial blood pressure and HR at 1 month in full responders and partial responders to monotherapy are summarized in Tables 3 and 4, respectively.
Table 3. Arterial Blood Pressure and HR in Full Responders to Monotherapy
Response to Antihypertensive Monotherapy at the Sixth Month
Antihypertensive monotherapy at the sixth month was less effective in group A than in group B. In particular, there were 3 (12%) full responders in group A and 6 (25%) in group B (P < 0.01). Ten patients in group A (40%) and 16 patients in group B (66%) were classified as partial responders (P < 0.01).
Likewise, after the crossing-over, monotherapy was less effective in group C than in group D. Full responders turned out to be 5 (17.85%) in group C and 8 (22.85%) in group D (P = NS).
Eleven patients (39.28%) in group C and 25 in group D (71.42%) were classified as partial responders (P < 0.01). The values of arterial blood pressure and HR at 6 months in full responders and partial responders to monotherapy are summarized in Tables 3 and 4, respectively.
Response to Antihypertensive Monotherapy at 1 Year
The results of antihypertensive monotherapy at 1 year in group A and in group B are reported in Fig. 2.
After 1 year, there were 5 (20%) full responders in group A and 8 (33.3%) in group B (P < 0.01). There were 8 (22%) partial responders in group A and 13 (54.1%) in group B (P < 0.01).
After the crossing-over, 7 patients in group C (25.00%) and 11 in group D (31.42%) were classified as full responders (P = NS). There were 9 (32.14%) partial responders in group C and 21 (60%) in group D (P < 0.01).
The values of arterial blood pressure and HR at 1 year in full responders to monotherapy are summarized in Table 3.
Response to the Antihypertensive Combined Therapy (Nifedipine or Carvedilol plus Ramipril)
Most partial responders to nifedipine or carvedilol received ramipril after the sixth month. Considering the crossing-over, we found that 8 of 9 partial responders to nifedipine received the combined antihypertensive therapy with nifedipine plus ramipril. The remaining patient did not receive the combined therapy because he was lost during the follow-up. In 6 of 8 patients (75%) who received the combined therapy, MAP was normalized (Table 5). Twenty of 21 partial responders to carvedilol received carvedilol plus ramipril. The remaining patient was not considered because of a protocol violation. In 6 of 21 patients (30%), MAP was normalized (Table 5). Therefore, nifedipine plus ramipril was more effective than carvedilol plus ramipril (P < 0.01).
Table 5. Values of Arterial Blood Pressure and HR in Responders to Combined Therapy at 1 Year
Responders to combined therapy (nifedipine + ramipril)
n = 6
SBP (mm Hg)
121.50 ± 1.13
DBP (mm Hg)
74.00 ± 2.45
MAP (mm Hg)
89.83 ± 1.16
69.50 ± 4.23
Responders to combined therapy (carvedilol + ramipril)
n = 6
SBP (mm Hg)
119.66 ± 1.34
DBP (mm Hg)
77.66 ± 2.32
MAP (mm Hg)
91.66 ± 1.41
66.16 ± 1.81
In patients who were partial responders to nifedipine, the level of PRA was higher than in partial responders to carvedilol, but the difference was not statistically significant (12.41 ± 4.02 versus 9.74 ± 3.14 ng/mL/hour; P = NS).
Global Response to Antihypertensive Therapy
Considering the patients as a whole, we found that responders to monotherapy (with nifedipine and carvedilol) were 15.9% after 1 month of treatment, 20.6% at 6 months, and 28.6% at 12 months. At 1 year of combined therapy (nifedipine or carvedilol plus ramipril), blood pressure was normalized in 19% of partial responders to monotherapy. As a result, at 1 year, 47.6% of patients who were included in the study normalized MAP by means of either monotherapy or combined therapy.
Effects of Antihypertensive Therapy on Renal Function
In responders to antihypertensive therapy, either monotherapy or combined therapy, levels of GFR and RPF at 1 year were significantly higher than baseline levels (GFR was 86.43 ± 23.21 versus 69.92 ± 26.02 mL/minute, P < 0.05; RPF was 447.15 ± 65.32 versus 336.12 ± 89.16 mL/minute, P < 0.01). In parallel, in responders, the values of serum urea (7.55 ± 1.78 versus 9.39 ± 2.18 mmol/L) and creatinine (108.50 ± 21.01 versus 113.66 ± 19.41 μmol/L) were reduced at 1 year compared to the baseline values, but only the reduction of serum urea achieved statistical significance (P < 0.01). In contrast, no improvement in RPF, GFR, serum urea, or serum creatinine was observed in nonresponders to antihypertensive therapy (data not shown).
Effects of Antihypertensive Treatment on Immunosuppression
The potential effects of nifedipine and carvedilol were evaluated only on pharmacokinetics of tacrolimus because in full responders to nifedipine, the number of patients that received cyclosporine was very small (0% and 10% before and after the crossing-over, respectively).
In full responders to carvedilol, the daily dose of tacrolimus at 6 months and 1 year should be reduced to almost 50% compared to that at the inclusion time in order to maintain the trough blood levels of tacrolimus in the therapeutic range. However, full responders to nifedipine did not require any adjustment of the daily dose of tacrolimus (Fig. 3).
The prevalence of de novo arterial hypertension is about 50%–85%.1, 12, 13 The main cause of de novo arterial hypertension is thought to be the use of CNIs. In fact, CNIs increase arterial vascular resistances, reducing the production of vasodilator factors and enhancing the production of vasoconstrictor factors. Another mechanism underlying CNI-related hypertension is a reduction of renal sodium excretion,14, 15 whereas the activation of both sympathetic and renin-angiotensin systems is thought to be less important. Because of the pathogenetic mechanism of de novo arterial hypertension, vasodilator agents have been claimed to be the first-choice antihypertensive drugs for its treatment. Nifedipine, a calcium antagonist, has been preferred among the vasodilatatory drugs because it is devoid of interactions with immunosuppressive agents. To date, there are no controlled clinical trials on the treatment of de novo arterial hypertension after OLT. In this study, the results of a controlled clinical trial comparing nifedipine and carvedilol, an α- and β-blocker, are presented.
The main result of this study indicates that an antihypertensive monotherapy is effective in the treatment of de novo arterial hypertension in 20.6% of patients at 6 months and 28.6% at 12 months. Monotherapy with carvedilol is as effective as nifedipine but is far better tolerated. In fact, comparing nifedipine and carvedilol as monotherapy, we found that the percentages of full responders at 6 and 12 were similar, whereas the percentages of adverse effects were significantly higher in patients treated with nifedipine (Fig. 2). The rates of response to monotherapy with nifedipine and carvedilol are lower than those usually observed among patients with essential hypertension treated with the same drugs. In fact, in the treatment of essential hypertension, nifedipine is effective in 50%–85%,16–18 and carvedilol is effective in 65%–85%.19–22 The reason for the restricted efficacy of carvedilol in the treatment of de novo hypertension probably is related to the minor role of activation of the sympathetic pathway in the pathogenesis of CNI-induced arterial hypertension. Meanwhile, the reduced efficacy of nifedipine has mainly been ascribed to the high rates of adverse effects. In agreement with previous observations by Textor et al.,23 42.8% of the patients who received nifedipine in this study had to stop the drug for adverse effects. This rate of adverse effects is almost double that observed in essential hypertensive patients.24, 25 However, as in patients with essential hypertension, the most common adverse effect of nifedipine in our series was the development of lower limb edema.25 In the pathogenesis of nifedipine-induced edema, the decrease of urinary sodium excretion is not very important; the main mechanism underlying this effect seems to be an increase in intracapillary hydrostatic pressure and an increase in capillary permeability.26, 27 The administration of nifedipine induces a decrease of vascular resistance of small arterioles with an increase of intracapillary hydrostatic pressure and induces an increase of transcapillary fluid filtration. The higher prevalence of ankle edema in patients with de novo arterial hypertension who were treated with nifedipine may be related to the persistence after OLT of vascular abnormalities, which were induced by the primitive liver disease.28 In advanced liver diseases, there is vascular remodeling that consists of structural modifications with a diminution of the muscular mass of the arterial wall. The impact of nifedipine on these architectural modifications of the vessel structure can contribute toward explaining the high rate of ankle edema that was observed in our series.
The second result of this study is that only 19% of partial responders to monotherapy (with nifedipine or carvedilol) have good control of blood pressure with combined therapy, in which an ACE inhibitor is added to the first chosen drug. In the treatment of essential hypertension, in fact, the efficacy of the ACE inhibitors is 45%–71%29 when they are used in monotherapy and >70%30 when they are added to other antihypertensive drugs. These data seem to confirm that the activation of the renin-angiotensin system is not an important mechanism underlying CNI-related arterial hypertension.
In our study, the combined therapy carvedilol-ramipril is less effective than nifedipine-ramipril, and this observation can be correlated to the larger antihypertensive effect of nifedipine compared to carvedilol in partial responders and/or to the relative inhibitor effect of carvedilol on the renin-angiotensin system.31, 32 Therefore, the efficacy of the combined therapy nifedipine-ramipril confirms previous data from the literature about the treatment of essential hypertension.33–35
The third result of the study is that the normalization of blood pressure with either monotherapy or combined therapy resulted in an improvement of renal function, as shown by an increase of GFR. Therefore, the decrease of vascular resistances due to nifedipine and carvedilol seems to enable an enhancement of renal perfusion in liver transplant recipients. Taking into account the prevalence and negative prognostic impact of renal dysfunction after OLT, we find that the normalization of blood pressure in liver transplant recipients is a relevant point in the strategy to prevent or improve renal function. The last result of this study deals with the different impacts of nifedipine and carvedilol on CNI pharmacokinetics. In fact, only in the group of patients treated with carvedilol after 6 and 12 months of treatment should the dose of tacrolimus be reduced to almost 50% in order to maintain the trough level in the therapeutic range. To our knowledge, these are the first data on the effect of carvedilol on the bioavailability of tacrolimus in liver transplant recipients. Nevertheless, this observation confirms the results of Kaijser et al.36 in kidney transplant recipients and those of Bader et al.37 in cardiac transplant recipients. In particular, Bader et al. recently supposed that carvedilol influences CNI levels through its effect on a membrane protein that regulates CNI absorption.
In conclusion, a monotherapy antihypertensive schedule is efficient in 20.6% and 28.6% of patients with de novo arterial hypertension after OLT at 6 months and 12 months, respectively. Carvedilol monotherapy is probably preferable as the first-choice drug because it is far better tolerated than nifedipine. Nevertheless, CNIs trough blood levels during treatment with carvedilol should be closely monitored in order to promptly adjust the CNI dose. In partial responders to monotherapy, the introduction of an ACE inhibitor as a second drug normalizes blood pressure in less than 20% of cases. In this context, the combined schedule with nifedipine plus ramipril is more effective than that with carvedilol plus ramipril. The normalization of blood pressure is an important factor for preserving renal function after liver transplantation.