Walter Bonfig*, and Hans Peter Schwarz

Transcription

1 Clinical Endocrinology (2014) 81, doi: /cen ORIGINAL ARTICLE Blood pressure, fludrocortisone dose and plasma renin activity in children with classic congenital adrenal hyperplasia due to 21-hydroxylase deficiency followed from birth to 4 years of age Walter Bonfig*, and Hans Peter Schwarz *Division of Pediatric Endocrinology & Diabetology, Department of Pediatrics, Technische Universit at M unchen and Division of Pediatric Endocrinology & Diabetology, Department of Pediatrics, Ludwig Maximilian Universit at M unchen, Munich, Germany Summary Introduction Infants with congenital adrenal hyperplasia (CAH) require higher doses of fludrocortisone (FC) due to physiological mineralocorticoid resistance. The adequacy of mineralocorticoid replacement should be closely monitored to avoid hypertension. Objective To evaluate blood pressure (BP) in infants with CAH due to 21-hydroxylase deficiency. Patients and Methods Thirty-three patients (18f/15 m) diagnosed by newborn screening were followed until the age of 4 years. Mean start of HC and FC treatment was day postnatally. Mean daily HC dose ranged from 86 to 123 mg/ m 2 /day. Results During the first year of life prevalence of systolic hypertension was up to 455%. At 12 and at 18 months, BP was highest. Prevalence of systolic hypertension was up to 576% at 18 months of age. After 24 months BP levels were lower and at 48 months prevalence of hypertension decreased to 152%. Systolic and diastolic BP correlated significantly with the administered fludrocortisone dose (r = 03, P = 0005), but not with body mass index. Hypertensive children received significantly higher FC doses and had significantly lower plasma renin activity during the study period. Conclusion High prevalence of transient, most likely FC induced hypertension was found in young children with classic CAH diagnosed by newborn screening. The changing mineralocorticoid sensitivity in infants is a risk factor for the development of hypertension in patients with CAH, who are treated with FC. Therefore suppressed plasma renin activity should be avoided to prevent arterial hypertension. (Received 30 December 2013; returned for revision 18 January 2014; finally revised 1 May 2014; accepted 2 May 2014) Correspondence: Walter Bonfig, Department of Pediatrics, Division of Pediatric Endocrinology, Technische Universit at M unchen, Parzivalstr. 16, D Munich, Germany. Tel.: ; Fax: ; walter.bonfig@lrz.tum.de Introduction Congenital adrenal hyperplasia (CAH) refers to a group of autosomal-recessive conditions characterized by disordered cortisol synthesis. In 90 95% of cases, it is caused by 21-hydroxylase deficiency due to CYP21A2 mutations. 1 Mineralocorticoid deficiency occurs in approximately 75% of cases of 21-hydroxylase deficiency. 2 Treatment of CAH due to 21-hydroxylase deficiency consists of glucocorticoid and mineralocorticoid replacement. In infancy a state of relative aldosterone resistance exists, that requires higher 9a-fludrocortisone replacement doses per kg body weight or per body surface area during infancy than later in childhood and in adults. 3 Due to changing mineralocorticoid sensitivity, it is important to monitor blood pressure, plasma electrolytes and measure renin or plasma renin activity to avoid overdosage. 4 Newborn screening allows early diagnosis and treatment of CAH due to 21-hydroxylase deficiency and is available in many countries nowadays. According to current consensus guidelines, initial treatment consists of both glucocorticoid and mineralocorticoid treatment once 21-hydroxylase deficiency has been diagnosed in a newborn. Overdosage of 9a-fludrocortisone will lead to low potassium and hypertension. 5 The challenge in infants and toddlers with CAH persists in finding the appropriate FC dose to avoid life-threatening salt-wasting crises often triggered by minor illness and, at the same time, avoid FC overdosage to prevent arterial hypertension with possibly long-term consequences. So far, studies on blood pressure in patients with CAH have been performed mostly in older children and adults In some of these studies, a tendency towards hypertension - already present in paediatric patients was found. So far, only one Dutch study has focused on BP during the first year of life in 24 CAH children, in whom mineralocorticoid deficiency was treated with fludrocortisone and salt supplements, and did not find elevated BP at that age. 14 The aim of our study was to evaluate blood pressure, fludrocortisone doses and plasma renin activity (PRA) in children with CAH due to 21-hydroxylase deficiency treated with hydrocortisone and fludrocortisone only during the first 4 years of life. In contrast to the Dutch study, CAH children were not supplemented with 871

2 872 W. Bonfig and H. Peter Schwarz salt, which might be an additional risk factor for development of hypertension in patients with simple virilizing CAH. Patients and methods Patients born between 1999 and 2008 diagnosed by newborn screening with classic CAH due to 21-hydroxylase deficiency were included in the study. In patients born before 2003, data were analysed retrospectively, and patients born after 2003, data were followed prospectively. Patients were followed for 4 years at a single centre. The study protocol was approved by local ethics committee. Thirty-three patients (18 females, 15 males) were identified, and diagnosis of CAH was reconfirmed by mutation analysis of CYP21A2. In seven patients, the genotype would have suggested simple virilizing CAH, but clinically this was only the case in two patients, in whom fludrocortisone substitution could be tapered during the study period. In the neonatal period, all patients were treated with hydrocortisone (1 mg three times daily) and fludrocortisone (01 mg to 015 mg daily). Mean age at start of treatment was day postnatally. Fludrocortisone dose was adjusted according to plasma renin activity (PRA, treatment goal: PRA <18 ng/ml/h until 6 months of age and <55 ng/ml/h above the age of 6 months) and presence of normal sodium concentration. PRA was determined monthly during the first 3 months of life and every 3 months thereafter. Also electrolytes were routinely monitored with each blood sample. None of the patients received salt supplementation. After 2 years of age, patients were followed every 6 months. Until the age of 2 years, supine length was measured in a paediatric measuring device (infantometer) and after 2 years of age, standing height was measured using a digital telescopic wall-mounted stadiometer (Ulmer Stadiometer). Weight was determined to the nearest of 01 kg using an electronic scale (Seca 753 E). Body mass index (BMI) was calculated as weight (kg)/height (m 2 ), and BMI- SDS was derived from data published by Cole et al. 15 Bone age was first assessed at 2 years of age and yearly thereafter on x-ray of the left hand and wrist. Bone age was determined with the Greulich & Pyle method, and bone age was read by both radiologist and experienced paediatric endocrinologist. Adjustment of glucocorticoid dose was made using auxological data (goal: linear growth as best clinical surrogate of metabolic control) and hormonal data (treatment goal: morning serum 17-hydroxyprogesterone <18 nmol/l and androstenedione not being fully suppressed). Blood pressure was measured at all visits using a Dinamap ProCare 400 vital signs monitor (GE Healthcare, Finland). The blood pressure was measured by an experienced nurse with parents calming or feeding the child. When the child could not be calmed, blood pressure was measured again at the end of the clinical visit, when the infant was calmed or sleeping. Age and size appropriate cuffs were used. Blood pressure was consecutively measured three times, and the mean blood pressure was used for the data analysis. Blood pressure values were compared with reference values for systolic and diastolic blood pressure from Flynn et al. (0 12 months of age) and from the German KiGGS study (1 4 years of age),. 16,17 Blood pressure was adjusted for height and age. A systolic and diastolic blood pressure above the 95th centile was considered hypertensive. Plasma renin activity (PRA) was measured with an angiotensin I generation radioimmunoassay (RENCTK, DiaSorin S.p.A., Saluggia, Italy) in our own endocrine laboratory. Within-assay variability was 75 99%, between-assay variability was %. PRA reference values were derived from healthy controls: PRA ng/ml/h until 6 months of age and ng/ml/h above the age of 6 months. When persistent hypertension was present in a patient and the high blood pressure did not respond to a fludrocortisone dose reduction, diagnostic work-up for hypertension (cardiologic and nephrologic examination) was initiated. For statistical analysis, IBM SPSS (Windows version 20 Armonk, NY, USA) was used. For correlation analysis, (correlation between blood pressure and fludrocortisone dose and PRA) nonparametric Spearman 0 s Rho correlation was used. Nonparametric Mann Whitney U-test was used for between-group comparisons (normotensive vs hypertensive children). A P-value <005 was considered significant (2-sided). For multivariate regression analysis, SAS (version 9.2 Cary, NC, USA) was used, and fludrocortison and hydrocortisone dose and PRA were investigated as independent variables after adjustment for multiple testing (Bonferroni method). Results Mean systolic and diastolic blood pressure (BP) was in the higher normal range at three and at 6 months of age (Table 2), with 10 of 33 (303%) children being hypertensive at 3 months and 13 (394%) children being hypertensive at 6 months of age. At twelve, 18 and 24 months of age mean systolic and diastolic BP was highest and statistically significantly elevated (P < 001). The highest rate of hypertension was found at 18 months of age with 576% of children being hypertensive. At 36 and 48 months of age, BP declined and was only mildly increased. Accordingly the percentage of hypertensive patients decreased until the end of the study period to 91%. Fig. 1 shows the course of systolic and diastolic BP. In three children only, hypertension persisted until the age of 4 years despite appropriate FC dose reduction. One patient also required antihypertensive treatment with calcium channel blockers. No vascular, cardiac or renal cause of hypertension was found in these three children, so that the diagnosis of essential hypertension remains. Family history of hypertension was negative in these patients. Also no cardiologic complications such as cardiac hypertrophy have been found in the CAH children that underwent cardiologic work-up. From three to 18 months of age mean daily FC dose was mg and mean HC dose was around 10 mg/m 2 /day (Table 2). Plasma renin activity declined from relatively high levels ( ng/ml/h) at 3 months of age to ng/ ml/h at 18 months (Table 1). At 24 months of age, the lowest

3 Blood pressure in children with CAH 873 BP [mmhg] Fig. 1 Course of systolic and diastolic BP (RR: Riva Rocci BP). Boxplots showing median, lower and upper quartile, whiskers 3rd and 97th centile, circles: outliers. Dotted line: 95th centile for systolic and diastolic BP for 50th height centile. found between blood pressure and hydrocortisone dose (P = 0175). BMI-SDS was elevated at 3 months of age (mean BMI-SDS ) and was below average between 6 and 24 months of age (mean BMI-SDS at 6 months 03 11, at 12 months 04 13, at 18 months and at 24 months 02 09). At 36 months of age, BMI-SDS was around average (01 09 SDS) and increased to SDS at 48 months. Comparing hypertensive and normotensive patients (Table 2), FC dose per body surface area was significantly higher in patients with hypertension at all time points (except 48 months of age, most likely due to the small patient number of hypertensive patients at 48 months), and PRA was significantly lower in hypertensive CAH patients throughout the study period. Patients with risk for hypertension had either PRA concentrations at the very low end of the normal range or even suppressed PRA. As shown by the height-sds data in Table 1, the CAH children in this study exhibited a notable growth pattern with catch-down growth during the first 18 months of life despite extremely low doses of hydrocortisone during that period. Bone age (BA) was not significantly advanced in these children with mean BA years at 2 years of chronological age, mean BA years at 3 years and mean BA years at the chronological age of 4 years. mean plasma renin activity (18 27 ng/ml/h) was measured and therefore FC dose was decreased to a mean daily dose of 005 mg (Table 2). Looking at the FC dose per body surface area per day, a decrease from lg/m 2 /day at 3 months to lg/m 2 /day at 48 months can be observed, corresponding to the increase in mineralocorticoid sensitivity over time. Both systolic and diastolic BP correlated significantly with the administered FC dose (r = 03, P = 0005), and a negative correlation between BP and PRA was found (r = 05, P = 0003). No statistically significant correlation between BP and BMI-SDS was found. Also no correlation between glucocorticoid dose and BP was found. Also in a multivariate regression model, blood pressure correlated positively with fludrocortisone dose (P = 0015, estimate 007, standard error of the mean 001, 95% confidence interval 0065 to 0075) and negatively with PRA (P = 002, estimate 015, standard error of the mean 002, 95% confidence interval 018 to 012). No correlation was Discussion Between 12 and 24 months of age, we found the highest prevalence of hypertension in young CAH children diagnosed by newborn screening and treated early with glucocorticoids and mineralocorticoids. High fludrocortisone dose and low plasma renin activity were associated with hypertension as one would expect by its mechanism of physiological action. Therefore, careful fludrocortisone dosing, assessment of renin or plasma renin activity and accurate blood pressure measurement are indispensable to avoid iatrogenic hypertension in children with CAH. Plasma renin activity should not be completely suppressed to avoid arterial hypertension. Comparing hypertensive vs normotensive patients, mean fludrocortisone dose was significantly higher in hypertensive patients (>250 lg/m 2 /day) and indicate fludrocortisone overtreatment in that group, which explains the high prevalence of transient hypertension. Table 1. Anthropometric data, blood pressure (BP), fludrocortisone (FC) and hydrocortisone (HC) dose and plasma rennin activity (PRA) over the study period (mean SD) Age (months) Height-SDS Height-SDS-Target Height-SDS BMI-SDS RR sys (mmhg) RR dias (mmhg) FC dose (mg/d) FC dose (lg/m 2 /d) HC dose (mg/m 2 /d) PRA (ng/ml/h)

4 874 W. Bonfig and H. Peter Schwarz Table 2. Comparison of FC dose per kg per day and plasma renin activity (mean SD) between hypertensive and normotensive children Hypertensive patients Normotensive patients P value 3 months Fludrocortisone lg/m 2 /day Plasma Renin Activity < months Fludrocortisone lg/m 2 /day Plasma Renin Activity < months Fludrocortisone lg/m 2 /day Plasma Renin Activity < months Fludrocortisone lg/m 2 /day Plasma Renin Activity months Fludrocortisone lg/m 2 /day Plasma Renin Activity months Fludrocortisone lg/m 2 /day Plasma Renin Activity months Fludrocortisone lg/m 2 /day Plasma Renin Activity months Fludrocortisone lg/m 2 /day Plasma Renin Activity So far there is only one study published assessing BP in children with CAH due to 21-hydroxylase deficiency in the first year of life. 14 In this Dutch study, BP was not elevated and no correlations between renin or fludrocortisone dosage were found. However, these 24 children were only followed until 1 year of age, and our data show that the risk for fludrocortisone induced hypertension was highest at 18 months of age in our study. In the Dutch study, children received salt supplementation in addition to fludrocortisone therapy. We did not supplement salt in our patients, as it is difficult to administer in toddlers and may also cause hypertension per se, especially in children with simple virilizing CAH. The changing mineralocorticoid sensitivity during infancy seems to play a key role in the development of transient hypertension, as BP responded to the reduction of fludrocortisone dose in the majority of our patients. In a retrospective chart review, Nebesio and Eugster 18 found essential hypertension in 5 of 91 children (55%) with CAH due to 21-hydroxylase deficiency. Diagnosis of essential hypertension was established between 2 months and 126 years in that cohort and patients needed to be treated with antihypertensive drugs (including hydralazine, calcium channel blockers, ACE inhibitors and hydrochlorothiazide) for a duration of 9 months up to 71 years. In our cohort, only one patient had essential hypertension requiring antihypertensive treatment with calcium channel blockers. In another two patients, hypertension persisted until the age of 4 years despite appropriate FC dose reduction. In older children and adolescents with CAH due to 21-hydroxylase deficiency, conflicting data on the prevalence of hypertension are reported: V olkl et al. 8 found altered 24-h blood pressure profiles with elevated systolic blood pressure in 55 children and adolescents with CAH (age years). BP correlated significantly with BMI but not with fludrocortisone or hydrocortisone dose. Diastolic BP was significantly lower during daytime and normal at night. A physiologic night-time decrease in systolic BP was found, but not in diastolic BP in these 55 CAH patients. Taken together the results of V olkl et al. and the results of our study, one can conclude that during infancy high fludrocortisone dose is the main risk factor for arterial hypertension whereas later in life high BMI is the primary risk factor for high BP. In contrast to V olkl et al., Hoepffner 0 s group 9 found average blood pressure values in the upper normal range in 23 children with CAH aged 6 17 years. In this cohort, however, BP was only measured by 24-h profile in eleven patients. Hoeppfner et al. also compared BP measurements in different settings: the highest BP measurements were observed in inpatient measurements, whereas BP in a calm and relaxed atmosphere in the outpatient clinic was significantly lower. When compared to 24-h ambulatory blood pressure monitoring, no significant difference to the outpatient measurements was found. Hoepffner et al. 9 concluded that BP measured in outpatients in a relaxed and calm atmosphere meets the requirements for monitoring of treatment and that measurement of BP on the ward leads to falsely high results and that 24-h ambulatory blood pressure monitoring might not be necessary. These results strengthen the validity of the outpatient BP measurement as performed in our study. One must also consider that in infants manual or automated BP measurement with the appropriate cuff size under calm conditions is the only option for assessment of hypertension. Roche et al. 12 performed 24-h ambulatory BP measurement in 38 British children with CAH and found a high prevalence of 58% of systolic hypertension. Eighty-four percent of patients had absence of the physiological nocturnal dip in systolic blood pressure. Age ranged between 61 and 182 years. Systolic BP correlated with BMI-SDS, particularly in females. In contrast, in our study, prevalence of hypertension was much lower in children aged 3 4 years. In the most recent UK study on BP in children with CAH, Subbarayan et al. 11 found systolic hypertension in 19 of 91 patients (209%) at an age between 04 and 205 years (mean age 92 years). Only mean systolic BP was significantly higher than in the normal population, whereas diastolic BP was not significantly elevated. As in our study, no correlation between BP and BMI-SDS was detected. Fludrocortisone doses used in the study by Subbarayan et al. were significantly lower, but one has to be careful in comparing

5 Blood pressure in children with CAH 875 these results to our study, because patients between 0 and 99 years were pooled to one age group in the UK study. Not only FC overtreatment leads to arterial hypertension, but also glucocorticoid overtreatment has to be taken into account. By the glucocorticoid sparing effect of FC, we were able to treat our patients with low doses of hydrocortisone and for this reason origination of hypertension from glucocorticoid overtreatment is not very likely in our cohort. Also no statistical correlation between glucocorticoid dose and BP was found in our analysis. According to the CAH consensus guideline, 5 all children with CAH are treated with fludrocortisone from time of diagnosis and many paediatric endocrinologists have developed the strategy to continue treatment with fludrocortisone also in simple virilizing CAH to reduce glucocorticoid dose. This practice goes along with the risk iatrogenic induction of hypertension, which definitely should be avoided in the long run. Currently it is unclear, if temporary hypertension and fludrocortisone over dosage put CAH children at a higher risk for development long-term hypertension and if temporary hypertension is significantly harmful to the kidney and the cardiovascular system. In summary, we found a high prevalence of transient, fludrocortisone induced hypertension between one and 2 years of age, which is basically caused by the changing mineralocorticoid sensitivity in that period of time. The correlations between BP and both fludrocortisone dose and plasma renin activity were statistically significant. Therefore reliable measurement of BP, renin or plasma renin activity and adjustment of fludrocortisone dose are essential in children with CAH. The challenge in infants and toddlers with CAH persists in finding the appropriate fludrocortisone dose to avoid life-threatening salt-wasting and, at the same time, avoid fludrocortisone overdosage. Persistent essential hypertension was infrequent in children with CAH due to 21-hydroxylase deficiency and may be associated more frequently with an increase in BMI and glucocorticoid dose during adolescence and adulthood. One might conclude from our data, that fludrocortisone dose should be carefully decreased between 12 and 18 months of age to prevent hypertension in CAH toddlers and one might also conclude, that renin levels - provided sodium and potassium concentrations are normal - should be kept at the upper normal range or even slightly above the reference range to ensure normal BP. As the patient number is rather small in the current analysis, further studies on BP in young CAH children are need. Disclosure The authors have nothing to disclose. Grant support None. References 1 Trapp, C.M., Speiser, P.W. & Oberfield, S.E. (2011) Congenital adrenal hyperplasia: an update in children. Current opinion in Endocrinology, Diabetes, and Obesity, 18, Merke, D.P. & Bornstein, S.R. (2005) Congenital adrenal hyperplasia. Lancet, 365, Martinerie, L., Pussard, E., Foix-L Helias, L. et al. (2009) Physiological partial aldosterone resistance in human newborns. Pediatric Research, 66, Padidela, R., Hindmarsh, P.C. (2010) Mineralocorticoid deficiency and treatment in congenital adrenal hyperplasia. International journal of Pediatric Endocrinology, doi: /2010/ Speiser, P.W., Azziz, R., Baskin, L.S. et al. (2010) Congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency: an Endocrine Society clinical practice guidline. Journal of Clinical Endocrinology and Metabolism, 95, Harrington, J., Pena, A.S., Gen, R. et al. (2012) Adolescents with congenital adrenal hyperplasia because of 21-hydroxylase deficiency have vascular dysfunction. Clinical Endocrinology, 76, Wasniewska, M., Balsamo, A., Valenzise, M. et al. (2013) Increased large artery intima media thickness in adolescents with either classical or non-classical congenital adrenal hyperplasia. Journal of Endocrinological Investigation, 36, V olkl, T.M.K., Simm, D., D otsch, J. et al. (2006) Altered 24- hour blood pressure profiles in children and adolescents with classical congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Journal of Clinical Endocrinology and Metabolism, 91, Hoepffner, W., Herrmann, A., Willgerodt, H. et al. (2006) Blood pressure in patients with congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Journal of Pediatric Endocrinology and Metabolism, 19, De Silva, K.S., Kanumakala, S., Brown, J.J. et al. (2004) 24-hour ambulatory blood pressure profile in patients with congenital adrenal hyperplasia a preliminary report. Journal of Pediatric Endocrinology and Metabolism, 17, Subbarayan, A., Dattani, M.T., Peters, C.J. et al. (2014) Cardiovascular risk factors in children and adolescents with congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Clinical Endocrinology, 80, doi: /cen Roche, E.F., Charmandari, E., Dattani, M.T. et al. (2003) Blood pressure in children and adolescents with congenital adrenal hyperplasia (21-hydroxylase deficiency): a preliminary report. Clinical Endocrinology, 58, Arlt, W., Willis, D.S., Wild, S.H. et al. (2010) Health status of adults with congenital adrenal hyperplasia: a cohort of 203 patients. Journal of Clinical Endocrinology and Metabolism, 95, Moojij, C.F., Kapusta, L., Otten, B.J. et al. (2010) Blood pressure in the first year of life in children with congenital adrenal hyperplasia due to 21-Hydroxylase deficiency: a pilot study. Hormone Research in Pædiatrics, 74, Cole, T.J., Freeman, J.V. & Preece, M.A. (1995) Body mass index reference curves for the UK, Archives of Disease in Childhood, 73, Flynn, J.T. (2000) Neonatal hypertension: diagnosis and management. Pediatric Nephrology (Berlin, Germany), 14, Neuhauser, H.K., Thamm, M., Ellert, U. et al. (2011) Blood pressure percentiles by age and height from non overweight children and adolescents in Germany. Pediatrics, 127, e978 e Nebisio, T.D. & Eugster, E.A. (2006) Observation of hypertension in children with 21-hydroxylase deficiency: a preliminary report. Endocrine, 30,