Evidence for a tubular defect in the loop of Henle
Impaired response to furosemide in Hyperprostaglandin E syndrome: Evidence for a tubular defect in the loop of Henle
In hyperprostaglandin E syndrome (HPS) renal wasting of electrolytes and water is consistently associated with enhanced synthesis of prostaglandin E2 . In contrast to Bartter or Gitelman syndrome (BS/GS), HPS is characterized by its severe prenatal manifestation, leading to fetal polyuria, development of polyhydramnios, and premature birth. This disorder mimics furosemide treatment with hypokalemic alkalosis, hypochloremia, isosthenuria, and impaired renal conservation of both calcium and magnesium. Therefore the thick ascending limb of the loop of Henle seems to be involved in HPS.
To characterize the tubular defect we investigated the response to furosemide (2 mg/kg) in HPS (n = 8) and BS/GS (n = 3) 1 week after discontinuation of long-term indomethacin treatment. Sensitivity to furosemide was completely maintained in patients with BS/GS. The diuretic, saluretic, and hormonal responses were similar to those of a control group of healthy children (n = 13), indicating an intact function of the thick ascending limb of the loop of Henle in BS/GS. In contrast, patients with HPS had a marked resistance to this loop diuretic. Furosemide treatment increased urine output by 7.5 ? 0.7 ml/kg per hour in healthy control subjects but only by 4.4 ? 1.2 ml/kg per hour ( p <0.5) in children with HPS. In parallel, the latter also had a markedly impaired saluretic response (DeltaClurine 0.14 ? 0.04 mmol/kg per hour vs 0.85 ? 0.09 mmol/kg per hour, p <0.001; DeltaNaurine 0.23 ? 0.06 mmol/kg per hour vs 0.77 ? 0.09 mmol/kg per hour, p <0.001). Furosemide therapy further enhanced prostaglandin E2 excretion in patients with HPS (54 ? 17 to 107 ? 28 ng/hr per 1.73 m2 , p <0.05), whereas no significant effect was observed in healthy children (20 ? 3 to 12 ? 3 ng/hr per 1.73 m2 ). We conclude that a defect of electrolyte reabsorption in the thick ascending limb of the loop of Henle plays a major role in HPS.
Renal tubular disorders associated with hypokalemic metabolic alkalosis are often compiled as “Bartter syndrome” but probably represent a number of variants resulting from different pathophysiologic events.   Clinically postnatal forms including classic Bartter syndrome  and hypomagnesemic Gitelman syndrome  must be distinguished from prenatal hyperprostaglandin E syndrome. The latter is characterized by polyhydramnios leading to premature delivery, a marked hypercalciuria resulting in nephrocalcinosis, and isosthenuria or even hyposthenuria.    To date no consensus has been achieved regarding the sites of tubular dysfunction in these disorders. Results derived from various clearance studies in patients with BS suggest impaired electrolyte reabsorption in the proximal tubule,  the loop of Henle,   the distal tubule,    or the cortical collecting duct.  In a comparison of two clearance methods, Boer et al.  emphasized the problems in interpreting clearance data as a result of the compensatory mechanisms modulating urinary electrolyte excretion in more distal tubular segments. Therefore exposure to diuretics with a well-defined action on tubular transport mechanisms seems to be a more promising approach. Normal increase in solute excretion after furosemide treatment   but impaired response to chlorothiazide   provide strong evidence that the primary defect in the postnatal variants is located in the distal convoluted tubule rather than in the loop of Henle.
In contrast, HPS, the prenatal variant of BS, resembles clinical and biochemical characteristics also found in infants exposed to long-term furosemide treatment such as hypokalemic alkalosis, normotension despite hyperreninemia and hyperaldosteronism, isosthenuria, and hypercalciuria with nephrocalcinosis.  Additionally a reduction of tubular Tamm-Horsfall protein synthesis and excretion has been observed in HPS but not in BS.  This indicates involvement of the thick ascending limb of the loop of Henle only in HPS. Consequently the aim of our study was to investigate the response to the loop diuretic furosemide in children with HPS and thereby to localize the tubular segment affected in this disorder.
Eleven patients with urinary salt loss and hypokalemic alkalosis were enrolled in the study, among them eight children with HPS (median age 10.5 years, range 84 to 11.8 years), two girls with GS (8.7 and 9.3 years), and one boy with classic BS (14.4 years). Classification into these distinct disorders was made according to the history, clinical, and biochemical characteristics as depicted in Table I . Children with HPS were hypercalciuric, and nephrocalcinosis was present in each of them as shown by renal ultrasonography. All patients had been treated with indomethacin (patients with HPS: 2.9 ? 0.8 mg/kg per day; patients with GS/BS: 3, 1.9, and 3.1 mg/kg per day) for several years; two patients also were treated with additional potassium supplementation (one patient with HPS: 1.7 mmol/kg per day; one patient with GS: 3.5 mmol/kg per day). Because of the very low serum levels of potassium this supplementation was continued during the entire study period. No other medication was taken by any of the subjects.
Long-term indomethacin treatment was discontinued 1 week before the children were exposed to furosemide to prevent any drug interaction and to study the tubular dysfunction in a “native” state. Blood samples and 24-hour urine collections were obtained before and 6 days after withdrawal of indomethacin, respectively. During this time patients were allowed to be on a self-demand dietary schedule required by the individual losses of electrolytes and free water. Withdrawal of indomethacin resulted in recurrence of the clinical picture and laboratory characteristics of the disorder within a few days (Tables II and III) .
Thirteen age-matched healthy children (median age 10.2 years, range 6.4 to 15.1 years) volunteered as control subjects during furosemide administration. In 10 of these children separate 24-hour urine collections were obtained under basal conditions. For ethical reasons, however, no blood samples were collected in this group. In addition three unaffected siblings (aged 8.8, 15.3, and 15.5 years) of patients with either HPS or BS were studied under the same protocol as their diseased family members. The protocol was approved by the university ethics committee. Informed parental consent and oral assent of the children was obtained before enrollment in the study.
The study was performed on two succeeding days according to the schedule shown in Fig. 1 . Urine was collected for a 3-hour-period subsequent to voiding at 9:30 am. The first day served as the control period. On the second day a single oral dose of furosemide (2 mg/kg as Lasix Liquidum) was administered. Fluid intake and composition of breakfast were the same during the test periods on both days. Blood samples were taken in the middle of each observation period.
The diuretic, saluretic, and hormonal changes observed between day 1 and day 2 were considered to be induced by furosemide administration. Because loop diuretics act from the luminal side of the tubule, their efficiency completely depends on urinary drug levels. Thus furosemide-induced inhibition of chloride reabsorption (DeltaClfuro ) can be calculated by the formula DeltaClfuro = DeltaClurine /Furourine , where DeltaClurine is the change in chloride excretion observed after furosemide administration and Furourine is the urinary excretion rate of furosemide. By analogy furosemide-induced sodium excretion and diuresis can be expressed as DeltaNafuro = DeltaNa urine /Furourine , and DeltaVfuro = DeltaVurine /Furourine , respectively.
Urinary and serum electrolyte concentrations, osmolality, and creatinine concentrations were measured by conventional laboratory methods. Plasma renin activity and aldosterone were assayed by radioimmunologic methods; upper normal limits in our laboratory are 5 ng/ml per hour and 15 ng/dl, respectively. Serum levels of indomethacin  and urinary levels of furosemide  were determined by high-performance liquid chromatography. To quantify the biosynthesis of prostaglandin E2 , urinary excretion of PGE2 and its major metabolite 7alpha5,11-diketo-tetranorprosta-1,16-dioic acid were measured by gas chromatography-mass spectrometry as described by Schweer et al.  Excretion of PGE-M is regarded to reflect predominantly the extent of systemic PGE2 production, whereas urinary PGE2 represents renal biosynthesis.  
Data are reported as means ? SEM in the control group and in the HPS group. The Student t test for related or unrelated samples was used to determine statistical difference. Values of p <0.05 were considered significant. Data of the three patients with either BS or GS and of the three healthy siblings are given as single values.
Oral administration of 2 mg/kg furosemide in healthy children not related to the patients induced isosthenuria and a sixfold increase in urine output (Table IV , Fig. 2) . The saluretic response was based mainly on the inhibition of chloride and sodium reabsorption (Fig. 3) , whereas potassium excretion was elevated only moderately. As is typical of loop diuretics, furosemide caused marked calciuria and magnesiuria (Table IV) . Urinary excretion of PGE2 and PGE-M decreased slightly but not significantly (Table V) . Normal sensitivity to furosemide also was found in the patients’ siblings. The renal loss of water and electrolytes equaled that observed in unrelated healthy control subjects (Table IV , Figs. 2 and 3) and was accompanied by hyperreninemia and hyperaldosteronism (Table V) . However, PGE-M but not PGE2 release was slightly increased in each of the three siblings (Table V) .
Print This Post