Targeting Calcineurin Inhibitor–Induced Arterial Hypertension in Liver Transplanted Children Using Hydrochlorothiazide

Abstract

OBJECTIVE

Arterial hypertension (AH) is the most common toxic effect of calcineurin inhibitor (CNI)–based immunosuppression in children after liver transplantation (LT). Activation of the renal sodium chloride cotransporter (NCC) by CNIs has been described as a major cause of CNI-induced AH. Thiazides, for example, hydrochlorothiazide (HCTZ), can selectively block the NCC and may ameliorate CNI-induced AH after pediatric LT.

METHODS

From 2005 thru 2015 we conducted a retrospective, single-center analysis of blood pressure in 2 pediatric cohorts (each n = 33) with or without HCTZ in their first year after LT. All patients received CNI-based immunosuppression. According to AAP guidelines, AH was defined as stage 1 and stage 2. Cohort 1 received an HCTZ-containing regimen to target the CNI-induced effect on the NCC, leading to AH. Cohort 2 received standard antihypertensive therapy without HCTZ.

RESULTS

In children who have undergone LT and been treated with CNI, AH overall was observed less frequently in cohort 1 vs cohort 2 (31% vs 44%; ns). Moreover, severe AH (stage 2) was significantly lower in cohort 1 vs 2 (1% vs 18%; p < 0.001). Multivariate analysis revealed HCTZ as the only significant factor with a protective effect on occurrence of severe stage 2 AH. While monitoring safety and tolerability, mild asymptomatic hypokalemia was the only adverse effect observed more frequently in cohort 1 vs 2 (27% vs 3%; p = 0.013).

CONCLUSIONS

Targeting NCC by HCTZ significantly improved control of severe CNI-induced AH and was well tolerated in children who underwent LT. This effect may reduce the risk of long-term end-organ damage and improve quality of life.

Introduction

Pediatric liver transplantation (LT) has become standard-of-care therapy, improving prognosis and quality of life for children and adolescents with terminal liver disease. Ten-year patient and graft survival rates exceeded 80% and 70%, respectively, in current studies.^1–4 However, the required chronic calcineurin inhibitor (CNI)–based immunosuppression, now almost exclusively administered by tacrolimus, leads to significant morbidity and impairment of quality of life. One of the most common toxic effects in children after LT is CNI-induced arterial hypertension (AH),⁵ which develops early after the introduction of CNI but persists dose-dependently after LT.⁶ In long-term survivors, i.e. 5 to 10 years after LT, the prevalence of AH was 17.5% to 27.5%.⁵ Arterial hypertension leads to end-organ damage, including cardiac hypertrophy^7,8 and renal insufficiency,⁹ in pediatric LT recipients. Furthermore, a recent study demonstrated elevated blood pressure (BP) at 1 year after LT as an independent risk factor for renal impairment 5 years after LT.¹⁰

Several mechanisms have been shown to contribute to CNI-induced AH,^11,12 including impaired vasodilatation^13,14 and systemic and renal vasoconstriction, possibly through endothelin mediation.¹⁴ Moreover, CNI-induced AH is salt sensitive.¹⁵ Preclinical studies identified the activation of the renal sodium chloride cotransporter (NCC) by CNI to play a major role (Figure 1).^16,17 The NCC activation leads to salt and water retention, resulting in AH. Thiazides can selectively block NCC. One of the thiazide compounds, hydrochlorothiazide (HCTZ), is widely used in children for various renal disease indications.^18,19 Hydrochlorothiazide has a good safety profile, but adverse effects including electrolyte disturbances and hyperuricemia have to be taken into account. Moreover, antihypertensive therapy with thiazides has been shown to prevent the progression of cardiovascular disease in adults.²⁰ This comparative cohort study analyzes whether specifically targeting the tacrolimus-induced NCC with HCT (Figure 1) can improve BP control in the first year after LT. Further, we assessed the safety of HCT and its effect on renal function.

Figure 1.

Activation of NCC by CNI. CNIs increase NCC activity by affecting the kinases WNK and SPAK. Thiazides can selectively block the NCC. (Adapted from Coffman.³⁴)

Patients and Methods

Study Design. This retrospective, single-center, cohort study analyzes the effect of HCT on CNI-induced AH in children after LT. The study was performed at the University Children's Hospital Tübingen, Germany, a national referral center for pediatric LT. In this study, we included children (ages 0–18 years) receiving solitary LT (i.e., excluding combined-organ transplantations) between 2005 and 2015. We excluded children with preexisting AH and chronic renal insufficiency (estimated glomerular filtration rate [eGFR] < 50 mL/min/1.73 m²). We focused on the first year after LT, when children were exposed to high CNI levels with consequently increased risk of AH, potentially contributing to long-term sequelae.

Study Population and Study Treatment. Based on preclinical results on the role of renal NCC in CNI-induced AH,^16,21 we modified our AH therapy to include HCTZ in 2012. Apart from adding HCTZ, our therapy protocol, including CNI-based immunosuppressive treatment, remained unchanged. In this single-center study, we studied the efficacy of HCTZ by conducting a retrospective analysis of BP and clinical parameters in 2 pediatric LT cohorts who underwent solitary LT from 2005 to 2015. Cohort 1 (HCTZ group; n = 33; transplantation 2011–2015) received the modified therapy protocol, including HCTZ at a dose of 1 to 2 mg/kg body weight daily. We initiated treatment upon detection of elevated BP values in the first week after LT. Cohort 2 (n = 33) consisted of patients who underwent transplantation from 2005 to 2011 who received antihypertensive treatment without HCTZ (control group). In both cohorts, upon detection of AH, we began stepwise antihypertensive treatment according to our in-house protocol (see below). All patients received an immunosuppressive regimen consisting of induction therapy with basiliximab and glucocorticosteroids, maintenance therapy with tacrolimus (aiming at a trough plasma concentration of 10–12 ng/mL during the first 6 months and 5–10 ng/mL during the second 6 months after LT) or, rarely, cyclosporine (n = 5), and low-dose glucocorticosteroids in the first year after LT.

BP Measurements and Definition of AH. Blood pressure measurements were performed at least 3 times daily before discharge after transplantation. Treatment was initiated on detection of hypertensive BP values on more than 3 occasions and then adapted accordingly. After discharge, BP values and antihypertensive treatment were reevaluated on outpatient visits 3, 6, and 12 months after LT. Blood pressure was measured in the outpatient clinic using an automatic oscillometric device with a size-appropriate cuff. In general, BP was measured in the sitting position and after a rest of at least 5 minutes. Measurement was performed at the dominant arm at least twice at each visit, and the lowest value was documented. The BP values were assessed based on sex, age, and height, according to the 2017 revised percentiles and guidelines of the American Academy of Pediatrics (AAP).²² Arterial hypertension was defined according to the AAP 2017 guidelines and categorized as stage 1 and stage 2 AH (Table 1). We implemented the AAP 2017 guideline's percentiles and policies to diagnose high BP in children and adolescents because they provided the highest level of sensitivity.²³

Table 1.

Definition of Hypertensive BP According to the 2017 AAP Guidelines²²

Age Range, yr*	Parameter
1–13	Normal BP: both systolic BP (SBP) and diastolic BP (DBP) <90th percentile.
	Elevated BP: SBP and/or DBP ≥90th percentile but <95th percentile, or 120/80 mm Hg to <95th percentile (whichever is lower). Elevated BP is predictive of arterial hypertension.
	Stage 1 arterial hypertension: SBP and/or DBP ≥95th percentile to <95th percentile + 12 mm Hg, or 130/80 to 139/89 mm Hg (whichever is lower).
	Stage 2 arterial hypertension: SBP and/or DBP ≥95th percentile + 12 mm Hg, or ≥140/90 mm Hg (whichever is lower).
≥13	Normal BP: BP <120/80 mm Hg
	Elevated BP: SBP between 120 and 129 with a DBP <80 mm Hg.
	Stage 1 arterial hypertension: BP between 130/80 and 139/89 mm Hg.
	Stage 2 arterial hypertension: BP ≥140/90 mm Hg.

^* Patients are categorized into 2 age groups. Of note, the recently revised definitions for adolescents align with adult guidelines for the detection of chronically elevated BP.

Treatment of AH. According to our in-house protocol, antihypertensive treatment was initiated for both cohorts in the first week after pediatric LT. Antihypertensive treatment consisted of nifedipine (dosage of 1–3 mg/kg per day), with or without enalapril (dosage of 0.2–0.4 mg/kg per day), and with or without metoprolol (dosage of 2 mg/kg per day). In the HCTZ group, we used HCTZ monotherapy or combination therapy of HCTZ, with or without nifedipine and with or without enalapril.

Analysis. Statistical analyses were performed using SPSS Statistics (IBM, version 25). Continuous variables were reported with medians and interquartile ranges and compared the groups using the Mann-Whitney U test for continuous, non–normally distributed variables. We used Fisher exact test to compare groups for categoric variables. Multivariate analysis with binary logistic regression was employed to determine which covariates were associated with stage 2 AH. All statistical tests were 2-sided, and p < 0.05 was considered statistically significant.

Results

Patient Characteristics. In total, 89 children received a solitary liver transplant at our center between 2005 and 2015. We excluded children with preexisting renal failure (n = 4), preexisting AH (n = 15), and early conversion to mTOR inhibitor (n = 1). Three patients were followed up at other centers after LT. Ultimately, 66 liver transplant recipients, ages 1 month to 16 years (median, 1.5 years), were included in our study. There was no significant difference in terms of patients' sex, age at the time of LT, immunosuppressive therapy regimen, underlying liver disease, and type of graft (dead full-size, dead split, or living donor liver transplant) between the 2 cohorts (Table 2). The concomitant anti-hypertensive treatment was comparable in both cohorts and consisted of calcium channel blockers (63.6% in cohort 1 vs 57.6% in cohort 2; ns); ACEI (36.4% in both cohorts), and in rare cases beta-receptor antagonists (0% in cohort 1 vs 12.1% in cohort 2; ns).

Table 2.

Patient Characteristics ^*

	Cohort 1 (HCT), n = 33	Cohort 2 (Control), n = 33	p value
Median recipient age at LT (IQR), mo	29 (6.0–120.0)	22 (8.0–136.0)	0.67
Median recipient age at LT (IQR), yr	2.0 (0.0–9.5)	1.0 (0.0–11.0)	0.86
Female sex, n (%)	15 (45.5)	13 (39.4)	0.80
Cholestatic liver disease, n (%)	15 (45.5)	20 (62.2)	0.32
Type of graft, n (%)	12 (36.4)	10 (30.3)	0.09
Postmortem full-size graft	7 (21.2)	15 (45.5)
Postmortem split liver	14 (42.4)	8 (24.2)
Living donor liver transplant
CNI	31	30
TAC	2	3
CSA
Median tacrolimus trough level 3 mo after LT (IQR), ng/mL	8.3 (6.10–10.10)	7.7 (6.25–11.20)	0.62
Median tacrolimus trough level 6 mo after LT (IQR), ng/mL	6.2 (4.4–7.25)	6.5 (5.05–10.15)	0.27
Median tacrolimus trough level 12 mo after LT (IQR), ng/mL	5.7 (4.15–6.68)	5.0 (3.80–6.40)	0.27
Low-dose steroids, <0.1 mg/kg body weight, ≥12 mo, n (%)	17 (51.5)	17 (51.5)	1.00

CNI, calcineurin inhibitor; CSA, cyclosporine A; HCT, hydrochlorothiazide; LT, liver transplantation; TAC, tacrolimus

^* Data reported as n (%) have p values from Fisher exact tests; median (IQR) have p values from Mann-Whitney U tests.

Arterial Hypertension. Overall, 37.8% of all patients showed hypertensive BP values within the first year after LT. The prevalence of AH tended to be lower in cohort 1 (HCTZ) vs cohort 2 during the first year of follow-up (31.6% in cohort 1 vs 44.4% in cohort 2; ns; Figure 2). Moreover, the prevalence of severe hypertensive BP values, defined as stage 2, was significantly lower in cohort 1 (HCTZ) vs cohort 2: 1.1% vs 17.8% of BP measurements (p < 0.001; Figure 3). Applying the former 2004 AAP guidelines, used for reference until 2017,²⁴ differences between the cohorts were more distinct: the prevalence of AH was significantly lower in cohort 1 (HCTZ) vs cohort 2: 27.4% vs 42.1% (p = 0.044). Severe hypertensive BP values were not detected in cohort 1 but in 12.2% of measurements in cohort 2 (p < 0.001).

Figure 2.

Prevalence of AH on each follow-up visit at 3, 6, and 12 months after LT. The prevalence of severe AH stage 2 was lower in cohort 1 (HCTZ) than in cohort 2 (control): at 3 months after LT, 0% vs 16% (p = 0.05); at 6 months, 3% vs 19% (p = 0.08); and at 12 months, 0% vs 19% (p = 0.02).

Figure 3.

Comparative cohort analysis of AH during follow-up. During the first year after LT, the overall prevalence of severe AH, defined as stage 2, was 1.1% in cohort 1 (HCTZ) vs 17.8% in cohort 2 (control) (p < 0.001).

In addition, we performed a multiple logistic regression with the presence of stage 2 AH as a dependent and cohort as an independent factor adjusted for prednisolone and HCTZ as factors and tacrolimus trough plasma concentration and GFR as continuous variables. No significant effect of the cohort can be observed (p = 0.513). Hydrochlorothiazide, however, shows a significant protective effect with respect to HTN stage 2 (OR, 0.10 [95% CI, 0.01–0.79]). Other confounders did not reach the level of significance (Table 3).

Table 3.

Logistic Regression Analysis of 66 Children for HTN stage 2 at 3, 6, and 12 Months After Liver Transplantation (LT) ^*

Predictor	β	Wald χ	p value	OR	95% CI
Constant	4.23	5.68	0.017	0.01
6 mo after LT vs 3 mo after LP	0.76	1.18	0.277	2.14	0.54–8.40
12 mo after LT vs 3 mo after LP	0.70	0.84	0.359	2.00	0.45–8.93
Prednisolone	1.34	1.48	0.223	3.83	0.44–33.25
Hydrochlorothiazide	−2.33	4.76	0.029†	0.10	0.01–0.79
Glomerular filtration rate mL/min/1.73 m2	0.00	0.00	0.987	1.00	0.99–1.01
Tacrolimus trough level	0.12	2.24	0.135	1.13	0.96–1.32

^* HCT shows a significant negative effect with respect to severe arterial hypertension (stage 2). Other confounders did not reach the level of significance.

^† Significant at p < 0.05.

Renal Function. Renal function was assessed by the eGFR²⁵ at screening before LT and 3, 6, and 12 months after LT. The eGFR was calculated by serum creatinine and height.²⁵ Overall, the renal function was unimpaired throughout the study period and did not differ between cohort 1 and cohort 2 (median eGFR, 113.3 vs 112.9 mL/min/1.76 m²; ns; Table 4).

Table 4.

Renal Function, the Occurrence of Electrolyte Disorders, and Hyperuricemia, Measured at 3, 6, and 12 Months After Liver Transplantation ^*

	Cohort 1 (HCT), n = 33	Cohort 2 (Control), n = 33	p value
Median eGFR (IQR), mL/min/1.73 m2	113.3 (90.0–152.8)	112.5 (92.2–157.2)	0.64
Hyponatremia, n (%)
Mild: Na < 135; ≥130 mmol/L	1 (3.0)	2 (6.1)	1.00
Moderate: Na < 130; ≥120 mmol/L	0 (0.0)	0 (0.0)	NA
Severe: Na < 120 mmol/L	0 (0.0)	0 (0.0)	NA
Hypokalemia, n (%)
Mild: K < 3.5; ≥3.0 mmol/L	9 (27.3)	1 (3.0)	0.013†
Moderate: K < 3.0; ≥2.5 mmol/L	1 (3.0)	0 (0.0)	1.00
Severe: K < 2.5 mmol/L	0 (0.0)	0 (0.0)	NA
Hyperkalemia: K > 5.5 mmol/L, n (%)	1 (3.0)	1 (3.0)	1.00
Hyperuricemia: >7.0 mg/dL, n (%)	3 (17.6); n = 17	4 (14.8%); n = 27	1.00

eGFR, estimated glomerular filtration rate; K, potassium; Na, sodium; NA, not applicable

^* Data reported as n (%) have p values from Fisher exact tests; median (IQR) have p values from Mann-Whitney U tests.

^† Significant at p < 0.05.

Adverse Effects of HCTZ. We observed no differences in the occurrence of hyponatremia and blood uric acid concentrations between the 2 groups (Table 4). In cohort 1 (HCTZ), we found a significantly higher prevalence of mild hypokalemia (potassium [K] <3.5 and ≥3.0 mmol/L; Table 4: 27.3% vs 3.0%; p = 0.013). However, the prevalence of moderate (K < 3.0 mmol/L) or even severe (K < 2.5 mmol/L) hypokalemia was not elevated. No cases of symptomatic hypokalemia were observed. After reduction of HCTZ dosage or initiation of potassium supplementation, potassium levels normalized. Finally, there was no graft loss or death in both groups during the whole study period.

Discussion

Activation of the NCC is a major pathomechanism of CNI-induced AH. Thiazides can selectively block the NCC and thereby reduce the risk of AH development.¹⁶ In children who have undergone LT, CNI medication mainly accounts for dose-dependent development of de novo AH.⁶ Therefore, the introduction of thiazides, for example, HCTZ (1–2 mg/kg/day), may specifically counteract the CNI-inducing effect on the NCC, leading to AH. This study is the first to evaluate thiazides blocking the CNI-inducing effect on the NCC, analyzing the effect on BP control in children after LT. Here we demonstrate that HCTZ leads to overall improved BP control and that severe CNI-induced AH is less frequently observed using HCTZ treatment alone or combined with other antihypertensives in the vulnerable group of children who have received a LT.

Arterial hypertension is one of the major complications accounting for extrahepatic morbidity after pediatric LT. McLin et al⁵ demonstrated that up to 27.5% of long-term survivors of LT had elevated BP measurements. In our study, the overall prevalence of hypertensive BP values was 37.8% in the first year after LT. The higher prevalence may be attributable to higher tacrolimus trough concentrations and higher mean glucocorticosteroid dosage in the first year after LT. High CNI trough concentrations⁶ and glucocorticosteroids⁵ are the main risk factors for AH after pediatric LT.

Morath et al¹⁰ studied the impact of hypertensive BP values 1 year after LT on long-term renal function. Their study showed that systolic BP values >140 mm Hg (corresponding to stage 2 AH) at 1 year after LT were an independent risk factor for renal impairment 5 years after LT. Our study demonstrated that stage 2 AH was significantly less frequently observed in patients treated with the HCTZ-containing regimen than in controls. Thus, this regimen may ameliorate the risk of renal injury after CNI-induced AH in pediatric LT recipients. However, long-term data on renal function after HCTZ treatment are needed.

Recently, studies on antihypertensive treatment after adult kidney transplantation showed thiazides to be safe and effective.^26,27 In those studies, not all kidney transplant recipients reached BP aims despite multiple combinations of antihypertensives.²⁷ However, after kidney transplantation, AH is multifactorial, and other factors may contribute, such as transplant nephropathy or the presence of non-functional native kidneys. In our cohort, patients had no preexisting AH or kidney disease. Therefore, CNIs and glucocorticosteroids can be regarded as the predominant cause of AH after pediatric LT,⁶ and targeted therapy of CNI-induced AH can be most beneficial.

Thiazide may lead to electrolyte disturbances, and we observed a significantly higher prevalence of mild hypokalemia in the HCTZ group. However, the prevalence of moderate or severe hypokalemia was not elevated. Moreover, all patients with documented hypokalemia were asymptomatic. However, severe hypokalemia may lead to adverse effects, such as muscular weakness and cardiac arrhythmias.²⁸ Therefore, upon evidence of hypokalemia, HCTZ dosage could be reduced, and/or potassium supplementation added. Although electrolyte dysbalances of thiazides are known to be dose-dependent, we did not find an association between HCTZ dosage and hypokalemia in our cohort.²⁹ We found no association between HCTZ and the prevalence of hyponatremia. A gradual dose escalation of HCTZ may help to prevent severe electrolyte disturbances: we recommend starting with 0.5 to 1.0 mg/kg body weight HCTZ with close monitoring of blood electrolyte concentrations. According to the current guidelines, the target dose should be between 1 and 2 mg/kg/day and not exceed 37.5 mg in total per day.²² Of note, HCTZ and ACEI combination can counteract the development of hypokalemia. Thiazide diuretics may be especially effective in patients who have other renal tubular effects of CNIs, including hyperkalemia and acidosis, in addition to AH. In this process, the polymorphism of enzymes and transporters involved in CNI pharmacokinetics may play a role,³⁰ highlighting the need to individualize AH therapy after solid organ transplantation.

Our study has several limitations. It is a retrospective, non-randomized study with historical controls, and patients were not randomized to the treatment arms. Our cohorts were too small to analyze subgroups (e.g., focusing on factors associated with thiazide efficacy). Our study focuses on the treatment of AH early after LT, when exposure to CNI toxicity is high and specific blockage of NCC is most beneficial.

Because HCTZ has been widely used in Europe, HCTZ was the thiazide of choice at our center. However, recent retrospective studies reported the correlation between long-term HCTZ treatment and the occurrence of non-melanoma skin cancer.³¹ Therefore, we recommend prescribing HCTZ for the limited period of the first year after LT when exposure to CNI toxicity is high and specific blockage of NCC is most beneficial. In general, the implementation of screening for skin cancer is essential for solid-organ transplant recipients.^32,33 Finally, chlorthalidone may be a potent alternative and has proven efficacy in treating CNI-induced AH in adult kidney transplant recipients.^26,27

Our results show that HCTZ can play an essential role in treating AH in children after LT. Thiazides can be applied as monotherapy or combination therapy with an ACEI or calcium-channel blocker. The AAP has recommended ACEIs use for children after solid-organ transplantation.²² Therefore, combination therapy with thiazides and ACEIs may increase antihypertensive treatment efficacy and help to prevent electrolyte disturbances. A combination therapy with HCT may also improve the treatment of CNI-induced AH in adult LT recipients. Future studies in pediatric and adult LT recipients are needed.

In conclusion, using thiazides as a selective blocker for CNI-induced NCC activation may improve BP control in pediatric LT recipients compared with standard therapy without thiazides. Patients with severe AH may benefit in particular. Improved BP control may reduce the risk for chronic kidney injury and cardiovascular target organ damage in pediatric LT recipients, thereby improving long-term quality of life. We recommend prospective clinical trials to further evaluate the efficacy of a thiazide-based antihypertensive treatment regimen for children after LT.

ABBREVIATIONS

AAP

American Academy of Pediatrics

ACEI

angiotensin-converting enzyme inhibitor

AH

arterial hypertension

BP

blood pressure

CNI

calcineurin inhibitor

eGFR

estimated glomerular filtration rate

HCTZ

hydrochlorothiazide

K

potassium

LT

liver transplantation

NCC

sodium chloride cotransporter