Electrocardiogram with prolonged QT interval in Gitelman disease

Kidney International, Vol. 62 (2002), pp. 580–584

Electrocardiogram with prolonged QT interval in Gitelman disease ALBERTO BETTINELLI, CAMILLO TOSETTO, GIACOMO COLUSSI, GINALUCA TOMMASINI, ALBERTO EDEFONTI, and MARIO G. BIANCHETTI Pediatric Renal Unit, University of Milan Medical School, Clinica De Marchi, Milan; Division of Pediatrics, Leopoldo Mandic Hospital, Merate; Division of Cardiology, HSR Casa di Cura S. Maria, Castellanza; Renal Unit, Niguarda-Ca⬘ Granda Hospital, Milan; and Division of Cardiology, San Gerardo Hospital, Monza, Italy; and Division of Nephrology, University Children’s Hospital, Inselspital, Berne, Switzerland

Electrocardiogram with prolonged QT interval in Gitelman disease. Background. Potassium and magnesium deficiency prolong the QT interval on a standard electrocardiogram and predispose the patient to dangerous cardiac arrhythmias. No information is available on QT interval in patients diagnosed with Gitelman disease. Methods. The QT interval was assessed on lead II in 27 patients with biochemically and genetically defined Gitelman disease, who had discontinued medical treatment for at least four weeks. They included 15 female and 12 male subjects, aged 6.7 to 40 years old, median 20 years old. The corrected QT interval was calculated from the measured QT interval and heart rate using the Bazett formula. Results. The corrected QT interval was normal (between 391 and 433 msec) in 16 and prolonged in the remaining 11 patients (between 444 and 504 msec). Patients with prolonged and patients with normal QT interval did not significantly differ with respect to female to male ratio, plasma potassium, plasma total magnesium, and plasma ionized calcium. Plasma sodium and chloride values were slightly but significantly lower and bicarbonate levels higher in patients with a prolonged than in those with a normal QT interval. Conclusions. The corrected QT interval is often pathologically prolonged in patients with Gitelman disease, suggesting that there is an increased risk for development of dangerous arrhythmias. Further investigations are required in patients with a prolonged QT interval to assess the true hazard of dangerous arrhythmias.

characteristic set of metabolic abnormalities that include hypokalemia and alkalosis [1–6]. The constellation of findings in Bartter syndrome result from a variety of renal tubular defects [1–6]. Gitelman disease (Mendelian Inheritance in Man # 263800) is also known as the mild variant of Bartter syndrome, as this disease sometimes is diagnosed during routine investigation. In addition to hypokalemia and alkalosis, patients with Gitelman disease display magnesium depletion and a relatively low urinary calcium excretion. Inactivating mutations in the thiazide-sensitive sodium-chloride cotransporter gene cause the disease [1–6]. Potassium and magnesium deficiency secondary to loss of gastric or intestinal secretions, to eating disorders or to use of drugs prolong the repolarization of ventricular cardiomyocytes and predispose to dangerous ventricular arrhythmias [7–10]. A prolonged QT interval on standard electrocardiogram is a simple estimate of extended ventricular repolarization that is assumed to impart an increased risk for development of a ventricular arrhythmia culminating in syncope or sudden death [7–10]. To the best of our knowledge, no information is available on the occurrence of unexplained syncope and sudden death in Gitelman disease, as indicated by the fact that none of the recently published reviews dealing with the diagnosis and management of Gitelman disease refer to these risk factors or suggest that an electrocardiogram should be recorded [1–6]. This is the first systematic analysis on QT interval in patients with clinically, biochemically and genetically defined Gitelman disease.

The term normotensive renal hypokalemic alkalosis or, more frequently, the eponym Bartter syndrome is currently used to describe an inherited renal disorder with a Key words: electrocardiogram, hypokalemia, hypomagnesemia, cardiac arrythmia, Bartter syndrome, sudden death, alkalosis, ventricular repolarization.

METHODS Patients

Received for publication November 26, 2001 and in revised form February 7, 2002 Accepted for publication March 8, 2002

The study population consisted of 27 patients (15 female and 12 male subjects, aged 6.7 to 40 years, median 20 years old) from 22 kindreds with Gitelman disease in

 2002 by the International Society of Nephrology

580

Bettinelli et al: QT interval in Gitelman disease

Italy (N ⫽ 21) and Switzerland (N ⫽ 5). They included five families, each with two patients, and 17 families with one participant patient each. Apart from absent arterial hypertension, the diagnosis of Gitelman disease was based on the following criteria [11]: (1) hypokalemia of renal origin, defined as plasma potassium ⬍3.6 mmol/L and fractional potassium excretion ⬎15.0 ⫻ 10⫺2; (2) plasma bicarbonate ⬎27.0 mmol/L; (3) hypomagnesemia of renal origin, defined as plasma magnesium ⬍0.70 mmol/L and fractional magnesium excretion ⬎4.00 ⫻ 10⫺2; (4) fractional chloride excretion ⬎0.500 ⫻ 10⫺2; (5) urinary calcium/creatinine ratio ⬍0.100 mol/mol. Mutations in the thiazide-sensitive sodium chloride cotransporter gene had been identified in 18 out of the 27 patients by single stranded conformation polymorphism, using polymerase chain reaction (PCR) primers designed from intronic sequences surrounding all exons of the gene, as previously described [12, 13]. None of the 27 Gitelman patients had a history of unexplained loss of consciousness with postural collapse, non-febrile seizures, chest discomfort, palpitations and exertional or nocturnal shortness of breath. In addition, a meticulous clinical examination failed to reveal any cardiovascular abnormality. At the time of the study the patients had not taken any medication for at least one month. They had been instructed to collect urine during 24 hours for determination of creatinine and to attend the outpatient clinic after overnight fasting. Body weight, height and sitting arterial pressure were measured, and a blood sample collected with minimal stasis and without movements of the forearm. The hydrogen ion concentration, carbon dioxide pressure, sodium, potassium, chloride, ionized calcium, total magnesium and protein, and creatinine were assessed in blood. The mentioned laboratory values also were assessed in a control group of 27 healthy subjects matched for age and sex. Finally, all patients underwent a standard 12-lead resting electrocardiogram. Blood hydrogen ion concentration, blood carbon dioxide pressure, plasma sodium, potassium, chloride and ionized calcium were measured using selective electrodes immediately after collection. Plasma bicarbonate was calculated from blood hydrogen concentration and blood carbon dioxide pressure according to the Henderson-Hasselbalch equation. Plasma total protein (biuret method) and plasma and urinary creatinine (kinetic alkaline picrate method) were determined by colorimetry. Plasma total magnesium was assessed by atomic absorption spectrophotometry. The glomerular filtration rate was calculated from creatinine clearance. The electrocardiogram was independently analyzed by two experienced cardiologists who were unaware of the patient’s diagnosis. The QT intervals were assessed at a paper speed of 50 mm/sec and an amplification of 0.1 mV/mm both on leads II and

581

V2 [14, 15]. The QT interval was measured manually from the earliest onset of the QRS complex to the latest point at which the T wave crosses the baseline (U waves were not measured as part of the QT interval). The QT interval and preceding RR intervals were measured for five consecutive cycles and averaged. The QT interval corrected for heart rate was calculated from the measured QT interval and from heart rate (RR interval) using the Bazett formula [15]. The normal value for the corrected QT interval is ⬍440 msec [15]. The two-tailed Kruskal-Wallis test (nonparametric analysis of variance for independent samples) with the Bonferroni adjustment, followed by a posteriori multivariate analysis, and Bland-Altman residual analysis were used for analysis. Test for bimodality was performed using the Lilliefors test, a recognized modification of the Kolmogorov-Smirnov normality test. Significance was assumed when P ⬍ 0.05. RESULTS Hypokalemia (plasma potassium ranging from 2.61 to 3.50 mmol/L), metabolic alkalosis (plasma bicarbonate from 27.3 to 39.5 mmol/L) and hypomagnesemia (plasma magnesium from 0.44 to 0.69 mmol/L) were noted in all 27 patients with Gitelman disease (Table 1). Patients with Gitelman disease and control subjects did not significantly differ with respect to age, female-to-male ratio, body weight, arterial pressure, glomerular filtration rate and plasma sodium. Plasma chloride and ionized calcium were significantly lower and plasma total protein higher in the group of patients with Gitelman disease than in the control group (Table 1). Depression of the ST segment ⬎0.5 mm, flat or low amplitude T waves, and U waves amplitudes ⬎1 mm were not observed in the 27 Gitelman patients. The QT interval measured on lead II varied from 542 to 647, median 580 msec, and that measured on lead V2 varied from 541 to 650, median 578 msec. The comparison of the QT interval measured on lead II and that measured on lead V2 by means of the Bland-Altman residual analysis given in Figure 1 indicates that the difference between leads is not clinically important and that the two methods of measurement can be used interchangeably. As a consequence only the numbers measured on lead II, the most commonly used measurement in clinical routine, will be used in this article. The lead II QT interval corrected for heart rate ranged between 391 and 504 msec in the 27 Gitelman patients. It was normal (between 391 and 433 msec) in 16 and prolonged in the remaining 11 patients. (The QT interval was assessed again three to ten months later in seven of the patients with a prolonged QT interval. Again, the corrected QT interval on lead II was ⬎440 msec.) The corrected QT interval ranged between 444 and 491 msec in

582

Bettinelli et al: QT interval in Gitelman disease Table 1. Clinical and laboratory parameters in 27 patients with Gitelman disease and in 27 control subjects Patients with Gitelman disease All patients

N Age years Gender, female:male Body weight kg Height m Arterial pressure mm Hg Heart rate per min Plasma potassiuma mmol/L Plasma bicarbonatea mmol/L Total plasma magnesiuma mmol/L Glomerular filtration rate mL/min ⫻ 1.73 m2 Plasma sodium mmol/L Plasma chloride mmol/L Plasma ionized calcium mmol/L Plasma total protein g/L Measured QT interval (lead II)a msec Corrected QT interval (lead II)a msec

Corrected QT-interval normal

Corrected QT-interval prolonged

Control subjects

27 20 (14–25) 15:12 52.0 (44.1–56.6) 1.59 (1.45–1.71) 113 (105–117)/76 (53–78) 64.4 [61.9–70.1]

16 23 (16–26) 9:7 55.9 (51.6–58.8) 1.62 (1.53–1.74) 117 (105–127)/77 (69–80) 64.2 [61.3–69.2]

11 17 (13–19)c 6:5 44.5 (42.2–50.9)c 1.45 (1.37–1.52)c 114 (111–120)/74 (62–75) 64.4 [62.4–70.1]

27 20 (14–26) 15:12 56.7 (47.5–61.1) 1.66 (1.59–1.74) 120 (113–130)/76 (68–83) 63.0 [59.5–71.4]

3.02 (2.90–3.39)

2.96 (2.82–3.35)

3.01 (2.89–3.38)

4.52 (4.40–4.74)

33.4 (30.3–34.1)

31.6 (28.9–33.5)

34.2 (33.0–38.0)c

25.2 (24.3–26.7)

0.59 (0.55–0.64)

0.59 (0.54–0.61)

0.61 (0.57–0.67)

0.79 (0.74–0.83)

106 (101–115) 140.0 (138.2–141.5) 101 (98–103)a

105 (99–116) 140.6 (139.8–142.0) 102 (101–104)

1.30 (1.23–1.35)b 77.0 (75.2–79.3)b

1.30 (1.22–1.33) 76.9 (75.0–78.9)

1.31 (1.26–1.36) 77.0 (75.6–79.8)

585 (565–604)

568 (552–575)

616 (600–637)

—

429 (421–466)

423 (409–482)

471 (463–482)

—

107 (105–114) 138.3 (137.0–140.1)c 98 (96–102)c

110 (95–115) 40.0 (138.1–142.0) 107 (106–108) 1.36 (1.32–1.39) 73.3 (70.1–77.2)

Results are given either as absolute values or as as median and interquartile range (which extends from the value at centile 25 to the value at centile 75 and includes half of the data points). a Patients and control subjects, respectively patients with a normal and patients with a prolonged QT interval were not compared; a significant difference was predictable as the value was by definition abnormal in one of the groups b P ⬍ 0.01 vs. control subjects c P ⬍ 0.05 vs. patients with normal corrected QT interval

Fig. 1. Comparison of the QT intervals measured on lead II and the QT measured on anteroseptal lead V2 by means of the Bland Altman diagram of residual analysis in 27 patients with Gitelman disease. The measured QT interval on the average was lower by 1 msec on lead II than on lead V2. Furthermore, the clinically irrelevant limits of agreement, namely the differences within mean ⫾ 2 SD of 9 msec indicate that lead II and lead V2 can be used interchangeably. Please note that the diagram presents the measured QT interval and not the corrected QT interval calculated from the measured QT interval and heart rate using the Bazett formula.

10 patients and was ⬎500 msec (504 msec) in one patient. The frequency distribution of the QT interval corrected for heart rate indicates the possible existence of a bimodal distribution (P ⬍ 0.05; Fig. 2). The analysis of

Fig. 2. Frequency distribution of the QT interval corrected for heart rate on lead II in 27 patients with Gitelman disease. The diagram indicates the possible existence of a bimodal distribution (P ⬍ 0.05). The vertical dotted line separates normal from increased QT intervals corrected for heart rate. Symbols are: ( ) females; (䊐) males.

the five families with two patients in each disclosed the following: in three families the QT interval corrected for heart rate was normal in both patients, in one family the interval was prolonged in both patients, and in one family it was normal in the first and prolonged in the second

Bettinelli et al: QT interval in Gitelman disease

patient. Patients with a normal and patients with a prolonged QT interval did not significantly differ with respect to the female-to-male ratio, arterial pressure, glomerular filtration rate, plasma potassium, plasma total magnesium, and plasma ionized calcium. Total plasma protein was significantly higher in patients. Patients with a prolonged corrected QT interval were slightly younger, smaller and lighter than patients without it. In addition, plasma sodium and chloride values were slightly but significantly lower and bicarbonate higher in patients with a prolonged than in those with a normal QT interval. In the 27 patients with Gitelman disease no significant correlation was noted between circulating potassium, magnesium, calcium or bicarbonate values, taken as independent values, and the corrected QT interval, taken as dependent value. To estimate the independent contribution of the measured variables to the variance of the corrected QT interval, a multiple regression analysis was performed with gender, age, and circulating sodium, chloride and bicarbonate as independent variables. Among these variables, sodium and chloride contributed significantly (P ⬍ 0.05) to the variance of QT, whereas gender, age and chloride did not. DISCUSSION The results of the present study demonstrate a prolongation of the QT interval, which is likely stable over time, in more than 40% of the patients with Gitelman disease. Similar findings have been previously reported in a very heterogeneous group of patients with normotensive renal hypokalemic alkalosis [16]. Gitelman disease is a rare autosomal recessive disorder that is often not diagnosed until late childhood or even adulthood. The disease is considered a benign variant of Bartter syndrome that is sometimes asymptomatic. However, the disease is usually symptomatic and associated with clinical manifestations. Cramps are observed in almost all patients. Affected patients may also present with tetany, particularly in association with vomiting and diarrhea. Fatigue also may be observed and a lower than average blood pressure may be seen, consistent with the tendency to salt wasting. Polyuria and nocturia are found in a minority of these patients. As in other diseases with recessive mode of inheritance, in Gitelman disease there is no direct correlation between the clinical phenotype and the underlying genotypic abnormality [1–6, 17]. In some heart diseases estimates of QT interval may vary in different leads [14, 15]. In our Gitelman patients the QT interval that had been measured manually by two experienced cardiologists was very similar in lead II and in lead V2. A prolonged QT interval is present when the corrected QT interval is more than 440 msec [14, 15]. It represents a phenotypic description predisposing to dangerous ven-

583

tricular arrhythmias [7–10]. A QT prolongation can be inherited or acquired as an adverse response to medication or metabolic abnormalities [7–19]. The inherited causes of prolonged QT interval result from an intrinsic abnormality in the mechanisms responsible for cardiac repolarization. The defect underlying Gitelman disease is an abnormality in the gene coding for the thiazidesensitive sodium chloride cotransporter in the distal tubule [1–6, 13, 17]. This gene is not expressed in heart tissue [12]. Therefore, it is assumed that the QT prolongation noted in our patients with Gitelman disease does not directly result from an altered intracardial expression of the thiazide-sensitive sodium chloride cotransporter. The analysis of the families each with two patients included in our study further supports this hypothesis. The results of the present study are not surprising. It is well known that electrolyte disturbances such as those noted in Gitelman disease, namely moderate to severe hypokalemia and hypomagnesemia, and a mild tendency toward hypocalcemia, may result in QT prolongation [7–10]. We failed to observe any significant correlation between extracellular levels of the mentioned electrolytes and the corrected QT interval. Apart from the rather small size of patients with this rare disease included in the study, a possible explanation for this result is that circulating electrolytes may not be representative of the corresponding myocardial levels. A complementary explanation is that it is the altered potassium gradient across the cardiomyocytic membrane that induces changes on the electrocardiogram. In our patients, however, the intramyocardial potassium concentration, and thus the corresponding gradient across cell membrane, is unknown. A further explanation is that potassium-, magnesium- and calcium-deficiency act in concert to prolong the repolarization of ventricular cardiomyocytes [7–9]. In our patients circulating sodium, chloride and bicarbonate weakly predicted the duration of the corrected QT interval. These correlations, which also have been noted in anorexia nervosa [18], support the traditional assumption that in Bartter syndrome and its variants, chloride depletion plays a pivotal role in the development of hypokalemia and alkalosis [19]. A prolongation of the QT interval imparts a relevant risk for development of a ventricular arrhythmia if there is history of syncope or if the corrected QT interval is longer than 500 msec [10, 20]. Consequently, no more than one of our 27 patients is considered to have a high probability of developing a dangerous arrhythmia. Further studies will be needed to show whether the mild QT prolongation noted in a large subset of our patients with Gitelman disease is relevant to their prognosis. We believe that the results of the study have at least four possible implications: (1) It may be prudent to evaluate all Gitelman patients for the possible occurrence of a syncope and to repeatedly record an electrocardio-

584

Bettinelli et al: QT interval in Gitelman disease

gram. A history of a syncope or a corrected QT interval of more than 500 msec should be taken as documentation that the risk of malignant ventricular arrhythmia is increased. In children a common presentation of syncope are seizures that may go for years without appropriate diagnosis. (2) In Gitelman disease acute diarrhea or vomiting further exacerbate hypokalemia and hypomagnesemia. We hypothesize that the risk for development of arrhythmias is especially high in these particular circumstances. (3) Several drugs prolong the QT interval and have been associated with cardiac arrhythmias. In addition to some anti-arrhythmic agents, QT-prolonging drugs include among others some antihistamines, some macrolides, some antifungals, some psychotropics, ␤2-agonists and cisapride [10, 20]. The potential of the mentioned drugs to induce arrhythmias is favored by concurrent imbalance in potassium or magnesium levels. In view of that a case can be made for avoiding the use of these drugs in Gitelman disease. (4) In patients with QT prolongation sudden death is often precipitated by physical activity [8]. We currently recommend that, like patients with inherited causes of QT prolongation, Gitelman patients with this finding be excluded from competitive sports [21]. In conclusion, the QT interval is often pathologically prolonged in patients with Gitelman disease, suggesting that they are at an increased risk for the development of ventricular arrhythmias at least in some circumstances [10, 20, 21]. Further investigations are required in Gitelman patients with a prolonged QT interval to assess the true hazard of dangerous arrhythmias. ACKNOWLEDGMENT We acknowledge the support of the Associazione Bambino Nefropatico. Reprint requests to Mario G. Bianchetti, M.D., Division of Pediatric Nephrology, Inselspital, CH-3010 Berne, Switzerland. E-mail: [email protected].

REFERENCES 1. Bettinelli A, Vezzoli G, Colussi G, et al: Genotype-phenotype correlations in normotensive patients with primary renal tubular hypokalemic alkalosis. J Nephrol 11:61–70, 1998 2. Guay-Woodford LM: Bartter syndrome: Unraveling the pathophysiologic enigma. Am J Med 105:151–161, 1998 3. Kurtz I: Molecular pathogenesis of Bartter’s and Gitelman’s syndromes. Kidney Int 54:1396–1410, 1998 4. Rodrı´guez-Soriano J: Bartter and related syndromes: The puzzle is almost solved. Pediatr Nephrol 12:315–327, 1998 5. Schurman SJ, Shoemaker LR: Bartter and Gitelman syndromes. Adv Pediatr 47:223–248, 2000 6. Shaer AJ: Inherited primary renal tubular hypokalemic alkalosis: A review of Gitelman and Bartter syndromes. Am J Med Sci 322: 316–332, 2001 7. Janeira LF: Torsades de pointes and long QT syndromes. Am Fam Physician 52:1447–1453, 1995 8. Weiner ID, Wingo CS: Hypokalemia – consequences, causes, and correction. J Am Soc Nephrol 8:1179–1188, 1997 9. Agus MS, Agus ZS: Cardiovascular actions of magnesium. Crit Care Clin 17:175–186, 2001 10. Towbin JA, Vatta M: Molecular biology and the prolonged QT syndromes. Am J Med 110:385–398, 2001 11. Bettinelli A, Bianchetti MG, Borella P, et al: Genetic heterogeneity in tubular hypomagnesemia-hypokalemia with hypocalciuria (Gitelman syndrome). Kidney Int 47:547–551, 1995 12. Mastroianni N, De Fusco M, Zollo M, et al: Molecular cloning, expression pattern, and chromosomal localization of the human Na-Cl thiazide-sensitive cotransporter (SLC12A3). Genomics 35: 486–493, 1996 13. Mastroianni N, Bettinelli A, Bianchetti M, et al: Novel molecular variants of the Na-Cl cotransporter gene are responsible for Gitelman syndrome. Am J Hum Genet 59:1019–1026, 1996 14. Cowan JC, Yusoff K, Moore M, et al: Importance of lead selection in QT interval measurement. Am J Cardiol 61:83–87, 1988 15. Moss AJ: Measurement of the QT interval and the risk associated with QTc interval prolongation: A review. Am J Cardiol 72:23B– 25B, 1993 16. Blomstro¨m-Lundqvist C, Caidahl K, Olsson SB, Rudin A: Electrocardiographic findings and frequency of arrhythmias in Bartter’s syndrome. Br Heart J 61:274–279, 1989 17. Cruz DN, Shaer AJ, Bia MJ, et al: Gitelman’s syndrome revisited: An evaluation of symptoms and health-related quality of life. Kidney Int 59:710–717, 2001 18. Swenne I, Larsson PT: Heart risk associated with weight loss in anorexia nervosa and eating disorders: Risk factors for QTc interval prolongation and dispersion. Acta Paediatr 88:304–309, 1999 19. Sabatini S: The cellular basis of metabolic alkalosis. Kidney Int 49:906–917, 1996 20. Lipicky RJ: A viewpoint on drugs that prolong the QTc interval. Am J Cardiol 72:53B–54B, 1993 21. Zipes DP, Garson A Jr: 26th Bethesda Conference: Recommendations for determining eligibility for competition in athletes with cardiovascular abnormalities. Task Force 6: Arrhythmias. J Am Coll Cardiol 24:892–899, 1994