Introduction

Arterial hypertension is a major contributor to accelerated atherosclerosis in adult graft recipients in whom myocardial hypertrophy represents a particularly strong risk factor for cardiac death. It is conceivable that this is also the case for pediatric graft recipients, as cardiovascular events are among the most frequent causes for long-term morbidity and mortality in this population [1, 2, 3].

However, there are only limited reports on arterial hypertension and altered myocardial architecture in children. These studies are mostly cross-sectional and their results are contradictory [4, 5, 6].

The aim of this retrospective longitudinal study was to analyze effects of post-transplant changes in arterial hypertension, as assessed by repeated 24-h ambulatory blood pressure measurement (ABPM), on myocardial architecture, as assessed by repeated echocardiography, in a cohort of children after renal transplantation.

Patients and methods

Population

In a retrospective chart review analysis, 39 children, adolescents, and young adults were identified during follow-up after renal transplantation, in whom 24-h ABPM and echocardiography had been assessed at least once within a 3-month interval. Immunosuppression consisted of standard triple drug therapy: cyclosporin A ( n =35) or tacrolimus ( n =4) in combination with prednisone and azathioprine or mycophenolate mofetil. Acute rejection episodes were diagnosed clinically and treated with 3–5 days of high-dose steroids (300–500 mg/m2). Patient data are given in Tables 1 and 2. Twenty repeated pairs of measurements were available to analyze the effects of post-transplant changes of arterial hypertension on myocardial architecture. Patient data from this group with repeat measurement were not different at baseline from the total group.

Table 1 Patient data (BSA body surface area, RTX renal transplantation, ESRD end-stage renal disease, GFR glomerular filtration rate)
Full size table
Table 2 Ambulatory blood pressure measurement (ABPM) and left ventricular mass index (LVMI) data (BP blood pressure, LVH left ventricular hypertrophy)
Full size table

The aim of the antihypertensive treatment was to maintain the blood pressure below the 95th percentile for age. The standard protocol started with a calcium antagonist (nifedipine or nitrendipine). If blood pressure was not adequately controlled, a beta-blocker was added (metoprolol). In isolated cases, other drugs such as angiotensin-converting enzyme inhibitors, hydralazine, or furosemide were used.

Methods

All children and/or parents performed routine home blood pressure measurements with various apparatus. These measurements were compared with the data obtained with an automatic sphygmomanometer (Dinamap, model 1846 SX) during the clinic visit. ABPM was performed on an outpatient basis with a Spacelabs 90207 monitor (Spacelabs Medical, USA); echocardiography was performed using a Vingmed 700, later Vingmed 5 (General Electrics, USA). APBM and cardiac ultrasonography were “randomly” performed according to physician’s choice and independently performed by two different teams (ABPM pediatric nephrology, ultrasonography pediatric cardiology). Only those data pairs that were assessed within a 3-month interval were included in the analysis. Investigators evaluating blood pressure were blinded to results of cardiac ultrasonography and vice versa.

Statistics

For each APBM, data were indexed by dividing the mean patient value for an entire 24-h period, for awake and sleep intervals by the 95th percentiles based on the reference values for mid-European children [7]. Children with an index >1 (mean blood pressure greater than 95th percentile) were considered hypertensive. Dipping was defined as a reduction in night-time blood pressure values of more than 10% compared with daytime values. Left ventricular hypertrophy (LVH) was determined using sex-specific percentiles derived from a study of normal children (the 95th percentile 103 g/m2 for males, 84.2 g/m2 for females) [8]. Our data were analyzed using commercial statistical software (SPSS 10.0). Differences between groups were tested with Mann-Whitney U test; intra-individual differences were assessed using non-parametric Wilcoxon signed rank test. Results were considered significant when P <0.05 or lower when statistically appropriate.

Results

ABPM during follow-up of renal transplantation

At the time of the first measurement, 23 of 39 patients were hypertensive. At the time of the repeat measurement 8 of 20 patients were hypertensive. Of these 8 patients, 6 had already been hypertensive at the initial assessment. Of the 12 patients with normal blood pressure at repeat assessment, 6 patients had been hypertensive at initial assessment. Overall, there were no significant changes in ABPM during a follow-up period of 1.9±1.1 years (range 0.5–3.78 years) (Table 2).

Echocardiographic measurements during follow-up of renal transplantation

At the time of the first measurement, 19 of 39 patients had LVH; at the time of the repeat measurement, 14 of 20 patients had LVH. Of these 14 patients, 12 already had LVH at the initial assessment; 6 patients maintained their normal left ventricular mass index (LVMI). Again, there was no change of LVMI or incidence of LVH during follow-up after renal transplantation (Table 2).

Correlation between ABPM and LVH

To study the influence of blood pressure control on myocardial architecture, we correlated ABPM data with LVMI. In the cross-sectional assessments, there were no significant correlations with any of the ABPM variables at the initial observation. However, at repeat assessment 1.9±1.1 years later, a significant correlation between mean systolic and diastolic blood pressure and LVMI was found (Table 3). Furthermore, when we correlated the longitudinally assessed changes in ABPM with alterations in LVMI, we also found significant effects of blood pressure control on LVH remodeling (Table 4).

Table 3 Correlations with LVMI (cross-sectional analysis)
Full size table
Table 4 Correlations with Δ LVMI (longitudinal changes) ( BMI body mass index, GFR glomerular filtration rate)
Full size table

Correlation between the LVMI and associated risk factors

To identify factors associated with LVH remodeling after renal transplantation, we also correlated LVMI with patient characteristics, laboratory values, graft function, and medications at initial assessment. Where applicable, alterations in LVMI during follow-up were also correlated with changes of these variables during follow-up. Primary renal disease had no influence on LVMI. At initial assessment only glomerular filtration rate and number of antihypertensive drugs were correlated with LVMI; at follow-up none of these variables reached statistical significance (Table 3). Longitudinal changes in LVMI during follow-up correlated with changes in cyclosporine dosage (Table 4). The occurrence of acute rejection was not associated with graft function, blood pressure, or LVMI.

Discussion

In adult transplant recipients, arterial hypertension and LVH are considered independent cardiovascular risk factors, which significantly increase the chance of sudden death, heart failure, coronary artery disease, and cardiac arrhythmias [9].

In recent years the increasing relevance of these complications has also been recognized in pediatric end-stage renal disease; cardiovascular events also constitute one of the most frequent causes of mortality in this population [1]. In our study, most children after renal transplantation were hypertensive and exhibited LVH. These data confirm other recent studies in children with renal transplants, describing arterial hypertension and LVH in similar percentages of the study populations [5, 6, 10, 11].

The etiology of these complications remains uncertain, but multiple etiological factors, including cyclosporine, steroids, native or recurrent kidney disease, rejection, and renal artery stenosis of the transplanted kidney, may play a role in patients with renal transplants [12]. The relative refractoriness to therapy in our population was evidenced by the use of multiple antihypertensive medications; most of our patients were on two or more antihypertensives at the time of assessment.

The cross-sectional data of our study failed to show an association between ABPM data and LVMI in the total group at initial assessment. Although LVH in chronic renal failure and renal replacement therapy is generally thought to result from left ventricular pressure overload, few studies have assessed the relationship between LVMI and ABPM in pediatric renal transplant recipients, and most of these are contradictory.

Such variability is not unexpected in small sample studies in pediatric renal transplantation, where several confounding factors, such as differential graft function and different doses of immunosuppressive and/or antihypertensive medications, will strongly influence the ability of blood pressure data to predict LVMI at any given time point. More than 10 years ago, Soergel et al. [13] described a significantly higher prevalence of LVH in 6 children with hypertension detected by ABPM than in 17 children who were normotensive. More recently, Matteucci et al. [5] also performed ABPM and echocardiography in 28 children after renal transplant and found LVH in more than 80% of patients. In their study, LVMI was significantly correlated with the mean systolic, but not diastolic ABPM data. However, Johnstone et al. [14] reported that neither systolic nor diastolic blood pressure was correlated with LVMI in 30 children and young adults after renal transplantation. Finally, Morgan et al. [6] also found no significant relationship in 45 children between LVMI and several indices of mean ABPM.

It has therefore been suggested that left ventricular abnormalities might rather be related to the cumulative effects of renal disease, its complications, and treatment than to actual hypertension at the time of assessment. In this respect, it was interesting that allograft rejection was not associated with graft function, blood pressure, or LVMI. At the initial assessment, graft function was the only independent parameter that was significantly correlated with both arterial hypertension and LVH.

When we analyzed the data of the subgroup of children in whom repeat assessment pairs were available for analysis, a significant correlation was found between ABPM and LVMI about 2 years after the initial measurement. Interestingly, in the longitudinal analysis none of the tested variables other than cyclosporine dosage and blood pressure control was correlated with changes of LVMI during the follow-up between the first and repeat measurement. These results are in good agreement with a recent prospective study where regression of LVH in adult transplant recipients with good graft function was dependent on a decrease in blood pressure during follow-up as assessed by ABPM [11]. Our data also extend the only other study evaluating longitudinal effects of blood pressure control (based on casual blood pressure measurements) upon myocardial architecture in children and adolescents after renal transplantation [4]. The results of our study clearly underline the importance of blood pressure control in the preservation of myocardial architecture.

A major limitation of our study is the small patient number, in particular in the group with repeat assessment data. Although the patient variables of this subgroup were not different from the total group at baseline, we cannot exclude selection bias due to the retrospective design. Moreover, these limitations precluded any analysis regarding potential therapeutic interventions, such as alterations of antihypertensive treatment.

In summary, our retrospective study demonstrates that control of blood pressure correlates with changes of LVMI in children with renal allografts. In the cross-sectional assessments, we detected a high prevalence of arterial hypertension by ABPM and of LVH by echocardiography in children after renal transplantation. In the longitudinal assessment, left ventricular remodeling was dependent on interval changes of blood pressure levels. Prospective interventional trials are indicated to further determine effects of active antihypertensive treatment on myocardial architecture.