Summary
Background
Electrophysiological alterations in atrial fibrillation (AF) may be genetically based and may lead to changes in ventricular repolarization. Short QT syndrome is a rare channelopathy with abbreviated ventricular repolarization and a propensity for AF.
Aims
To determine if minor unrecognized forms of short QT syndrome can explain some cases of lone AF.
Methods
We prospectively compared QT intervals in 66 patients with idiopathic lone AF and 132 age- and sex-matched controls. QT intervals were measured during sinus rhythm in each of the 12 surface electrocardiogram leads and corrected using Bazett’s formula (QTc). QT intervals were also corrected using other formulae. Uncorrected QT and heart rate regression lines were compared between AF patients and controls.
Results
AF patients presented with a slower resting heart rate (64 ± 10 beats per minute [bpm] vs 69 ± 9 bpm; P = 0.0006). QTc intervals were shorter in AF patients in 11/12 electrocardiogram leads (significant in 7/12, borderline in 2/12; mean QTc 381 ± 21 ms vs 388 ± 22 ms; P = 0.02). QTc intervals were also shorter in AF patients, significantly or not, using other correction formulae. For similar heart rates, uncorrected QT intervals were shorter in patients when heart rates were greater than 70 bpm and longer when heart rates were less than 60 bpm. AF patients displayed steeper QT/heart rate regression line slopes than controls ( P = 0.009).
Conclusion
Heart rate is significantly slower and the rate dependence of ventricular repolarization is significantly altered in patients with lone AF compared with controls. Further study is warranted to determine if AF induces subsequent ventricular repolarization changes or if these modifications are caused by an underlying primary electrical disease.
Résumé
Hypothèse
Les modifications électrophysiologiques dans la fibrillation atriale sont parfois d’origine génétique et pourraient donc conduire à des modifications de la repolarisation ventriculaire. Le syndrome du QT court est une canalopathie rare caractérisée par une repolarisation ventriculaire raccourcie et une propension pour les arythmies ventriculaires et la fibrillation atriale. Les deux pourraient être liés.
Objectifs
Rechercher une anomalie de la repolarisation chez les patients avec fibrillation atriale idiopathique à la recherche d’un syndrome du QT court fruste sous-jacent.
Méthodes
Les intervalles QT de 66 patients avec fibrillation atriale idiopathique ont été comparés de manière prospective à ceux de 132 témoins appariés en âge et en sexe. Les intervalles QT étaient mesurés en rythme sinusal dans chacune des 12 dérivations de surface et corrigés avec la formule de Bazett (QTc). Les intervalles QT étaient aussi corrigés selon d’autres formules et les pentes des droites de régression entre QT et fréquence étaient comparées entre patients et témoins.
Résultats
Les patients avec fibrillation atriale avaient une fréquence cardiaque plus basse (64 ± 10 bpm vs 69 ± 9 bpm ; p = 0,0006). Les QTc étaient plus courts chez les patients avec fibrillation atriale dans 11/12 dérivations (significatif dans 7/12, limite dans 2/12 ; QTc moyen 381 ± 21 ms vs 388 ± 22 ms ; p = 0,02). Les QTc étaient aussi plus courts chez les patients avec fibrillation atriale, de manière significative ou non, quand d’autres formules de correction étaient utilisées. À fréquence comparable, les QT non corrigés étaient plus courts chez les patients avec fibrillation atriale pour les fréquences supérieures à 70 bpm et plus longs pour les fréquences inférieures à 60 bpm. La pente de la droite de regression QT–fréquence était plus grande chez les patients que chez les témoins ( p = 0,009).
Conclusion
La fréquence cardiaque est significativement plus lente et la fréquence-dépendence de la repolarisation ventriculaire est significativement altérée chez les patients avec fibrillation atriale idiopathique comparées aux témoins. Il reste à déterminer si la fibrillation atriale induit secondairement des changements dans la repolarisation ventriculaire ou si ces modifications sont le témoin d’une atteinte primitive électrique sous-jacente.
Background
Atrial fibrillation (AF) is the most common cardiac arrhythmia, reaching a prevalence of 1% in the general population . In 2 to 30% of cases, AF occurs without any cardiac alteration or facilitating extracardiac abnormality and is called lone AF . Most of the time, AF is a consequence of a multifactorial process involving various acquired structural alterations, while primitive electrical disease can be suspected in lone AF, where no structural heart disease is present. A genetic background is suspected in familial cases, which account for 15% of lone AF cases . Several mutations or variants have been discovered during the past decade , mainly on genes encoding various membrane ion channels involved in atrial but also ventricular action potentials.
Short QT syndrome (SQTS) is a recently recognized channelopathy associating abbreviated atrial and ventricular repolarizations and a propensity for atrial and ventricular arrhythmias . In particular, AF is frequently observed in SQTS patients . To date, discovered mutations in SQTS mainly involve potassium channels, leading to a gain of function , similar to that observed in some familial cases of AF , although QT intervals are apparently normal in these patients.
The aim of this study was to compare QT interval durations in patients with lone AF and matched controls, based on the hypothesis that minor unrecognized forms of SQTS may explain some cases of lone AF.
Methods
We performed a case-control study at two centres, comparing QT interval durations during sinus rhythm in patients with lone AF and matched controls.
Patients with lone AF were prospectively recruited over the past 2 years at our two institutions (University Hospital Rangueil, Toulouse, France and Hôpital de la Tour, Meyrin, Switzerland). To be included, patients had to be aged less than 75 years and present with a history of documented paroxysmal AF without any detectable underlying heart disease and without diabetes, hypertension or any precipitating condition, such as a history of renal, hormonal or pulmonary disease . Transthoracic echocardiography was available in each case and had to be compatible with the diagnosis of lone AF (no ventricular hypertrophy or dilatation, normal ventricular systolic function, no valvular or pericardial abnormality, no increased left ventricular diastolic filling pressure). An isolated left atrial dilatation was not an exclusion criterion . Patients with a recent episode of AF occurring during the previous 48 h – either documented or suspected by suggestive symptoms – were excluded, as well as patients under antiarrhythmic drugs, digoxin, beta-blockers, calcium channel blockers or, more generally, under any cardioactive drug or medication with some electrophysiological action or known to alter cardiac repolarization or to perturb ionic homeostasis. Patients aged < 16 years and those with a familial history of AF were excluded, as were professional sportsmen or patients practising intensive physical activity (> 7 h a week).
Once this population was selected, we prospectively recruited age- and sex-matched controls over the following months. Controls were matched with a 2:1 ratio according to sex and age (patients and controls were grouped in sections of 10 years of age). Control patients were recruited during preoperative consultations before non-cardiac surgery or by means of consultations for non-organic cardiovascular symptoms or from the medical or paramedical staff. These recruits were men or women without any personal or familial cardiovascular history, with unremarkable physical examination and without any cardioactive medication. Exclusion criteria for patients with AF also applied for controls. Patients or controls with a resting sinus node heart rate that was too high (> 85 beats per minute [bpm]) or too low (< 45 bpm) were excluded because of the inaccuracy of most QT correction formulae at these rates .
One standard 12-lead surface electrocardiogram (ECG) was used for analysis for each patient or control. The ECG was recorded at rest during the same standard clinical conditions in patients and controls (i.e. common conditions for ECG recording as performed during usual medical consultation, at a paper speed of 25 mm/s with a standard ECG recorder [1 kHz sampling rate, 0.1–250 Hz filters]). For each tracing, sinus node heart rate, PR interval and QRS duration were measured in lead II. The QT interval duration was then measured in each of the 12 ECG leads, from the QRS onset to the end of the T wave, as defined by the intersection of the tangent of the steepest slope of the last part of the T wave and the isoelectric line, according to the technique described by Surawicz . QT was not measured in case of flat T wave (less than 0.1 mV) or when the end of the T wave was difficult to determine. Measurements were averaged over five successive beats in case of sinus node arrhythmia (defined by instantaneous heart rate variations ≥ 10% of the mean heart rate). ECG analysis was independently performed by two cardiologists blinded to the groups in a subset of 20 cases, for evaluation of interobserver variability.
To allow determination of corrected QT (QTc), QT durations were then corrected according to Bazett’s formula (QTc = QT/√RR) . For each ECG, the mean QTc was determined over the 12 leads. QT durations were additionally corrected according to the formulae by Hodges (QTc = QT + 1.75 [heart rate – 60]) , Fridericia (QTc = QT/√3RR) , Rautaharju (QTc = QT + [410 – 656/(1 + heart rate/100)]) and Sagie Framingham (QTc = QT + 154 [1 – RR]) .
Statistical analysis
Statistical analysis was performed using StatView 5 (version 4.57; Abacus Concepts, Inc., Berkeley, CA, USA; 1992–1996) and StatA (version 11.1; StataCorp LP, College Station, TX, USA). Numerical values are presented as means ± standard deviations (ranges) and categorical data as proportions. Normal distribution was expected, given the significant number of patients in each group, so unpaired t tests were used to compare numerical data between groups. Proportions were compared using the Chi 2 test. The intraclass correlation coefficient (ICC) was calculated to determine the reliability of QT measurements; interobserver variability was assessed by comparing the results from observers 1 and 2. Furthermore, due to the imperfection of correction formulae, especially in case of different heart rates between groups, uncorrected QT interval durations were compared in subgroups sharing a similar heart rate. In addition, significant interaction between uncorrected QT/heart rate regression lines for patients and controls was determined. A P value < 0.05 was considered significant for each analysis.
Results
Sixty-six successive patients with paroxysmal lone AF were prospectively included and compared with 132 matched controls. Two controls – but no patients – were excluded before enrolment because of intensive sporting activity. There were no differences in sex (72% men) or age (mean, 49 ± 13 years) between groups due to the matching. All patients or controls were Caucasian. ECGs were always recorded during the daytime (12:00 ± 3 h without difference between groups). Mean heart rate was 67 ± 10 bpm (45–85 bpm). Mean PR interval was 156 ± 26 ms (100–220 ms) and mean QRS duration was 82 ± 4 ms (80–100 ms).
Heart rate was significantly slower in patients with lone AF than in controls (64 ± 10 bpm vs 69 ± 9 bpm; P = 0.0006). Prevalence of sinus node arrhythmia was similar between groups (17% in both groups; P = 0,8). The PR interval was longer in patients with lone AF (160 ± 29 ms) than in controls (154 ± 24 ms), although the difference was not significant ( P = 0.15). No complete bundle branch block or pre-excitation was observed in any ECG. There was no difference in QRS width between groups (82 ± 4 ms for both groups).
Repolarization was found to be unremarkable on every ECG (positive T wave in each lead except V1, V2, VR and sometimes III). There was good agreement between observers for the determination of QT intervals (ICC 0.83). Differences in QTc intervals are shown in Table 1 . In each ECG lead (except V2), Bazett-corrected QT intervals were found to be shorter in patients with lone AF than in controls. Differences were significant for 7/12 ECG leads and borderline in two other leads ( Fig. 1 ): in V5, for example, QTc was 384 ± 22 ms vs 393 ± 24 ms in controls ( P = 0.01). Mean QTc was 381 ± 21 ms in AF patients and 388 ± 22 ms in controls ( P = 0.02) (a difference of 7 ms).
