Global Impairment of Cardiac Autonomic Nervous Activity Late After Repair of Tetralogy of Fallot
Background Sustained ventricular tachycardia (VT) and sudden cardiac death (SCD) remain devastating late complications after repair of Tetralogy of Fallot (ToF). Although heart rate variability (HRV) and baroreflex sensitivity (BRS) are recognized as independent markers of autonomic activity and strong predictors of SCD in major cardiovascular disease, little is known about their role in patients with tertralogy.
Methods and Results We measured HRV and BRS in 45 ToF patients (27 male, age 35±12 years, 26±7 years after repair) and 45 matched healthy controls. Subjects underwent 20 minute of resting measurements of heart rate (ECG) and noninvasive beat-to-beat blood pressure recording (Finapres), with 5 minutes of 0.1Hz controlled breathing followed by cardiac MRI. BRS was computed by spectral analysis and the sequence and controlled breathing methods. All HRV time and frequency domain variables were measured. All BRS and HRV variables were significantly reduced in patients compared with controls (P<0.001 in all). HRV tended to increase with years from repair. BRS decreased with previous palliation and increasing patient age. Both HRV and BRS decreased with pulmonary regurgitation, elevated right ventricular end systolic volumes and reduced right and left ventricular ejection fraction. Finally, there was an inverse relation between QRS duration (predictor of sustained VT and SCD) and indices of HRV but no relation with indices of BRS.
Conclusion There is global impairment of autonomic nervous system regulation late after repair of tetralogy with marked reduction of BRS and HRV. This seems to relate to previous surgical intervention/s, their timing and current right and left-sided hemodynamics. Reduced HRV also related to markers of sustained VT and SCD, suggesting possible common pathogenic mechanisms. Further studies are required to examine the prognostic significance of impaired BRS and HRV in these patients.
Repair of tetralogy of Fallot (ToF) has an excellent long-term prognosis and a 25-year survival of more than 94%.1 However, these patients face considerable late morbidity and mortality. The main cause of late mortality is sudden cardiac death, accounting for more than 1 third of late deaths.1–3 Sudden death is usually because of sustained ventricular tachycardia (VT). Pulmonary regurgitation with resultant right ventricular dilatation are the common underlying hemodynamic substrates.3 Maximum QRS duration on the surface ECG and its rate of increase relate to right ventricular dilatation and are markers of sustained VT and SCD.3,4
Deranged cardiac autonomic nervous activity (CANA) is associated with cardiovascular mortality (including SCD), and worse prognosis in coronary artery disease and heart failure.5,6 Heart rate variability (HRV) and baroreflex sensitivity (BRS) are independent quantitative markers of CANA.6 Reduced HRV is a strong independent predictor of SCD after myocardial infarction (MI) and may be used to estimate the risk of SCD.7,8 BRS is more powerful than HRV in predicting sustained VT and cardiac death post MI9,10 and is a significant predictor of mortality in chronic heart failure.11
Limited data exist on CANA in congenital heart disease after surgical intervention and even less so in ToF late after repair, mostly focusing on HRV.12,13 Our prospective study aimed to assess global CANA late after ToF repair and examine its relation with surrogate markers for SCD, namely underlying hemodynamics and ECG parameters.
Forty-five patients (27 male) with tetralogy repair aged 15 or older, were recruited from the Royal Brompton Adult Congenital Heart Program, London, UK. Exclusion criteria included diabetes mellitus, atrial fibrillation or>2 ventricular ectopic beats per min during data acquisition, permanent pacing, clinical instability within the preceding 3 months, and antiarrhythmic drugs affecting HRV or BRS. Surgical details were obtained from operative notes. The patients’ symptomatic status was defined according to NYHA functional classification. Forty-five age- and sex-matched healthy controls were also studied. The study received Ethics Committee approval and all subjects gave informed consent.
Patients were studied between 1:00 pm and 5:00 pm under standardized conditions, in a quiet room. All were fasted for at least 2 hours and were not allowed to smoke or drink alcohol- or caffeine-containing beverages for 24 hours. Subjects rested supine for 15 minutes and then underwent 2 investigations: 1 successive 20 minute recording at rest and 1 5 minute recording during controlled breathing at 0.1 Hz by following a sinusoidal visual and auditory signal.14 During each investigation, RR interval, BP and respiration were recorded. Procedures were performed in the same order in all subjects. The resting recording was performed first in case a period of controlled breathing caused excessive conscious awareness of respiration, altering the natural breathing pattern. BP was measured by a Finapres device (model 2300; Ohmeda, CA), with the cuffed finger resting comfortably at the level of the heart. The Finapres cuff was wrapped around the index finger of the left hand (right hand for patients with a left Blalock-Taussing shunt). Subjects underwent several minutes of accustomisation to the Finapres and the servo-adjust mechanism was turned off before recording. ECG was acquired from the limb lead with the largest R wave (typically lead II). Respiratory rate was measured by impedance plethysmography. All data were sampled at 1000 Hz on a computer using an analogue-to-digital converter (National Instruments). Readings were saved into floppy disk and analyzed off-line with custom designed software. The program measured RR intervals and beat-to-beat systolic pressure while ectopic beats were corrected by linear interpolation.14
Heart Rate Variability
For BRS assessment, we measured average amplitudes of oscillations in RR interval and in systolic blood pressure (SBP). BRS was assessed non-invasively by spectral analysis, the sequence method, and the controlled breathing method (Table 1).14
ECG and Cardiac Magnetic Resonance (CMR)
QRS duration was measured manually from 12 lead ECGs as previously described.4 CMR was performed in 36 patients on the same day using a 1.5 tesla Siemens Sonata system. Volume measurements, indexed for body surface area (m2), and pulmonary regurgitant fraction (PRF) were made as previously described.16
Data were expressed as frequency for the nominal variables and as mean ± standard deviation (SD) for continuous variables. Indices with skewed distribution were log transformed. Data were analyzed with Statview 5.0 (Abacus Concepts, SAS Institute, Cary, NC). Differences between patients and controls were assessed with unpaired Student’s t test, or the Mann-Whitney rank-sum for data failing tests of normality. Univariate and multivariate linear regression was used for determination of independent predictors of the assessed variables. All tests were 2 sided. P<0.05 was considered significant.
Clinical, surgical, ECG and hemodynamic data are shown in Table 2. Two patients were on furosemide, 1 on bendroflurazide and 1 on phenytoin and clobazam. None of the patients had previous history of sustained VT.
Heart Rate Variability
All HRV indices in patients were significantly reduced, compared with controls (Table 3). Within the ToF group, years after repair affected positively SDNN (b=1.2, 95% CI 0.4 to 2, P=0.005), RMSSD (b=1.7, 95% CI 0.5 to 2.9, P=0.007) and triangular index (b=2.3, 95% CI 0.024 to 4.5, P=0.048) (Table 4). Palliation, type of palliation, years remained palliated or years after palliation had no effect on time or frequency domain indices of HRV. PRF had a significant negative effect on SDNN (b=−0.42, 95% CI −0.74 to −0.1, P=0.011) and on RMSSD (b=−0.56, 95% CI −1.07 to −0.05, P=0.034) whereas right ventricular end-systolic volume index (RVESVi) had a significant negative effect on the triangular index (b=−0.86, 95% CI −1.4 to −0.3, P=0.003) and log LF (b=−0.016, 95% CI −0.029 to −0.003, P=0.02). LVEF had a positive effect on SDNN (b=0.92, 95% CI 0.24 to 1.6, P=0.009), triangular index (b=2.99, 95% CI 1.36 to 4.63, P=0.001) and log LF (b=0.057, 95% CI 0.02 to 0.09, P=0.004). QRS duration tended to prolong with decreasing SDNN (b=−0.48, 95% CI −0.95 to −0.003, P=0.049).
All BRS indices in patients were significantly reduced, compared with controls (Table 3). Within the ToF group, age affected negatively BRScbr (b=−0.16, 95% CI −0.28 to −0.03, P=0.018) (Table 4). Type of repair did not affect BRS. However, previous palliation had a negative effect on BRScbr (b=−2.1, 95% CI −3.65 to −0.55, P=0.009). RVESVi affected negatively BRScbr (b=−0.13, 95% CI −0.2 to −0.05, P=0.003). LVEF had a positive effect on BRSseq (b=0.2, 95% CI 0.06 to 0.35, P=0.007) and RVEF on BRScbr (b=0.26, 95% CI 0.08 to 0.44, P=0.006). RVESVi had a negative effect on log aLF (b=−0.009, 95% CI −0.016 to −0.002, P=0.019). RVEF correlated with LVEF (r=0.71, P<0.001) and PRF had a positive effect on RVESVi (b=0.85, 95%CI 0.42 to 1.27, P<0.001). FIGURE FIGURE
We have shown global impairment of CANA late after ToF repair as evidenced by markedly reduced HRV and BRS. This impairment relates to underlying heamodynamics and may be contributory to arrhythmogenesis in these patients.
HRV is an index of multiple components swaying the heart rate around its mean value. Nonspectral or time domain parameters involve computing indexes that are not directly related to specific cycle lengths. This method offers a simple means of defining patients with decreased variability in the mean and standard deviation of R-R intervals. Spectral analysis provides an assessment of the vagal modulation of the R-R interval. Parasympathetic tone is primarily reflected in the high-frequency (HF) component of spectral analysis. The low-frequency (LF) component is influenced by both the sympathetic and parasympathetic nervous systems.15 BRS is a more specific measure of the vagal ability to modulate heart rate in a manner that buffers spontaneous changes in blood pressure and can be assessed by injection of vasoactive agents or by non-invasive methods. We used a well validated noninvasive method, avoiding possible adverse effects on patients and controls.14 We also employed a controlled breathing protocol to correct for varying respiratory rates and depths that may affect BP and RR interval.17 Steady breathing rate (0.1 Hz), enhances the LF oscillations in both RR and blood pressure, thus optimizing the reproducibility of measurements of BRS.18 The CANA plays a key role in myocardial stability and has been implicated in triggering SCD.5,6 Although many risk factors for SCD have been investigated in patients with repaired ToF, a satisfactory method to risk stratify these patients remains elusive.19
We used a standard 20 minute resting ECG recording for both patients and controls under the same external conditions for HRV assessment. Although this is not a 24 hour recording, it allows ample time for the frequencies of interest to manifest themselves and eliminates the variation in the degree of physical activity between patients that inevitably occurs during 24 hours of unrestricted and unobserved behavior. We found that all time- and frequency-domain HRV indices were reduced in ToF patients compared with healthy controls. This clearly shows that not only the variation in heart rate differs between patients and controls (reduction in SDNN and Triangular index) but a significant withdrawal both in vagal and sympathetic activity and a sympathovagal balance deregulation exists.
McLeod et al reported a reduction in HRV in young ToF patients>10 years after repair.12 The authors showed a reduction in NN50 using 24 hour Holter ECG monitoring, which was weakly associated with increasing age, right ventricular dilatation and right ventricular hypertension. However, age was not adjusted for years after repair. In our study, HRV was reduced with age in both patients and healthy volunteers. However, controlling for age, years after repair proved to be the only independent predictor of HRV. Ohuchi et al in a more recent study13 suggested a surgery-related insult causing a cardiac sympathetic denervation after right ventricular outflow tract (RVOT) reconstruction surgery. This was based on the correlation of decreased [123I] metaiodobenzylguanidine (MIBG) myocardial uptake with the number of surgical procedures for right ventricular outflow tract (RVOT) reconstruction, and on the fact that the number of surgical procedures correlated inversely with HRV and BRS indices. Only 2 patients in our study had undergone a second RVOT operation and we did not use MIBG scintigraphy. Therefore, we cannot conclude on the effect of multiple surgeries on the CANA. However, we cannot reject the hypothesis that an initial damage on the CANA occurred during repair. The positive effect of years after repair on indices of HRV (SDNN, RMSSD and triangular index) supports a degree of cardiac reinnervation- involving at least the CANA of the sinus node- following repair. Similar cardiac autonomic reinnervation has been shown in other surgical groups, such as infants after arterial switch operations20 and heart transplant recipients.21,22 Apart from surgical damage of the cardiac autonomic nerves, transient ischemic injury during bypass is another plausible explanation. Indeed the CANA is very sensitive to ischemia.23 This would lead to improvement of HRV latter after repair. However, one should not underestimate the negative effect of impaired hemodynamics on HRV, which may interact with any possible improvement mediated by any of the above mentioned mechanisms. Longitudinal HRV data are required to confirm our observation.
The degree of pulmonary regurgitation (PR) assessed by CMR had a negative influence on SDNN and RMSSD, suggesting more marked HRV depression with increasing PR. This may be important as severe PR is the most common underlying hemodynamic substrate in repaired ToF patients associated with sustained VT and SCD.3 Pulmonary regurgitant fraction (PRF) influenced RV end systolic volume in our study. The latter had a negative effect on the triangular index and logLF. This may further imply that RV dilatation, associated with PR, may contribute to the deranged CANA of these patients. Indeed there was a negative influence of RV end systolic and diastolic volumes on all time domain and spectral HRV indices, albeit weak. This is probably because of the relatively small number of patients who underwent cardiac MRI. Furthermore, it is likely that multiple factors influence CANA. LVEF showed a significant positive influence on SDNN, Triangular index and log LF. Patients with reduced LVEF had significantly lower Triangular index (91.2±27. versus 8146.5±44, P=0.001) and a tendency for lower SDNN (P=0.057), in keeping with patients with reduced LVEF because of heart failure.11 RVEF did not affect directly HRV indices. It is known, however, that in patients with repaired ToF there is progressive left ventricular dysfunction possibly attributed to long standing PR.24
Results from the different methods used concur that ToF patients late after repair have a significantly reduced BRS compared with healthy controls. RVESVi had a negative effect on log aLF implying an inverse relationship between increased RV volume and baroreceptor sensitivity. BRSseq was positively affected by LVEF. BRScbr was negatively influenced by age, palliation and RVESVi, whereas RVEF had a positive influence. The relationship of BRS indices with the hemodynamic substrate (RVESVi, RVEF and LVEF) implies that biventricular dysfunction may lead to deranged BRS control. Patients with previous palliations had a significantly reduced BRScbr compared with non-palliated patients keeping in accord with the findings of Ohuchi et al13 Longstanding cyanosis in palliated patients and volume overload before repair, may have impacted on the reflex autonomic control of blood pressure. Age at repair was not an independent predictor of BRS in our study, but the inverse relationship with palliation is in favor of early definitive repair. In contrast to HRV indices, BRS had no relationship to years after repair. This implies that parasympathetic cardiac reinervation or recovery after transient ischemic injury as reflected in BRS improvement late after RVOT surgery in the study by Ochuchi et al,13 might be opposed by the impaired hemodynamic substrate of our cohort. Of all BRS indices in our study, BRScbr related with existing surrogate markers of survival in ToF. Its excellent reproducibility and its easy method of assessment commend it as the preferred method for BRS assessment to risk stratify patients with ToF late after repair.
Relation of HRV and BRS Indices With ECG Markers of Arrhythmia
McLeod et al in their study found a relationship between QRS duration and HRV.12 Our results are in agreement with this finding. QRS duration late after repaired ToF is known to be associated with RV dilatation and malignant arrhythmias. SDNN adjusting for RV volumes was the only significant independent predictor of QRS duration with a negative influence on the latter. This may suggest that apart from RV dilatation, a deranged CANA may also affect the depolarization sequence in these patients. We speculate that PR late after ToF repair leads to RV dilatation and stretching and deranges CANA. Both RV dilatation and deranged CANA, in turn, and in conjunction with the original ventriculotomy scar may lead to abnormal and inhomogeneous depolarization with clear arrhythmic implications. There was no relation between QRS duration and BRS indices, suggesting that if depressed BRS plays a role in arrhythmogenesis and SCD, this has to be through a different and perhaps additional pathogenic mechanism.
We could not examine the direct influence of surgery on CANA because preoperative and immediate postoperative HRV and BRS data were not available. Trying to keep the study non-invasive and needle-free, we did not assess catecholamine levels or muscle sympathetic nerve activity. Additional predictors of CANA dysfunction may exist, and be identified with a larger patient sample and longer period of observation in future studies. Furthermore, the potential effects of surgery and late intervention on CANA in patients with repaired ToF need prospective assessment.
HRV and BRS are markedly depressed late after repair of ToF. This may be the result of an initial insult during surgical repair, but the relationship of these indices with the hemodynamic substrate of the patients implies that the latter may play an important role in the CANA derangement. HRV indices showed a positive relationship with years after repair, suggesting a possible improvement in CANA because of reinervation. Reduced HRV also related to established markers of sustained VT and SCD, suggesting possible common pathogenetic mechanisms. Further studies are required to examine the prognostic importance and the impact of impaired BRS and HRV on arrhythmia and SCD in patients after ToF repair.
Drs Davos and Davlouros were supported by a Hellenic Society of Cardiology grant.
Gatzoulis MA, Till JA, Somerville J, et al. Mechanoelectrical interaction in tetralogy of Fallot. QRS prolongation relates to right ventricular size and predicts malignant ventricular arrhythmias and sudden death. Circulation. 1995; 92: 231–237.
Zipes DP, Levy MN, Cobb LA, et al. Sudden cardiac death. Neural-cardiac interactions. Circulation. 1987; 76: I202–I207.
Schwartz PJ. The autonomic nervous system and sudden death. Eur Heart J. 1998; 19 (suppl F): F72–F80.
Bigger JT Jr, Fleiss JL, Steinman RC, et al. Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation. 1992; 85: 164–171.
Farrell TG, Paul V, Cripps TR, et al. Baroreflex sensitivity and electrophysiological correlates in patients after acute myocardial infarction. Circulation. 1991; 83: 945–952.
McLeod KA, Hillis WS, Houston AB, et al. Reduced heart rate variability following repair of tetralogy of Fallot. Heart. 1999; 81: 656–660.
Ohuchi H, Suzuki H, Toyohara K, et al. Abnormal cardiac autonomic nervous activity after right ventricular outflow tract reconstruction. Circulation. 2000; 102: 2732–2738.
Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North Am Society of Pacing and Electrophysiology. Circulation. 1996; 93: 1043–1065.
Brown TE, Beightol LA, Koh J, et al. Important influence of respiration on human R-R interval power spectra is largely ignored. J Appl Physiol. 1993; 75: 2310–2317.
Cooke WH, Cox JF, Diedrich AM, et al. Controlled breathing protocols probe human autonomic cardiovascular rhythms. Am J Physiol. 1998; 274: H709–H718.
Bricker JT. Sudden death and tetralogy of Fallot. Risks, markers, and causes. Circulation. 1995; 92: 158–159.
Kondo C, Nakazawa M, Momma K, et al. Sympathetic denervation and reinnervation after arterial switch operation for complete transposition. Circulation. 1998; 97: 2414–2419.
Wilson RF, Laxson DD, Christensen BV, et al. Regional differences in sympathetic reinnervation after human orthotopic cardiac transplantation. Circulation. 1993; 88: 165–171.
Bernardi L, Valenti C, Wdowczyck-Szulc J, et al. Influence of type of surgery on the occurrence of parasympathetic reinnervation after cardiac transplantation. Circulation. 1998; 97: 1368–13674.
Ponikowski P, Chua TP, Piepoli M, et al. Augmented peripheral chemosensitivity as a potential input to baroreflex impairment and autonomic imbalance in chronic heart failure. Circulation. 1997; 96: 2586–2594.