The Electrocardiogram 100 Years Later
2002; Lippincott Williams & Wilkins; Volume: 106; Issue: 17 Linguagem: Inglês
10.1161/01.cir.0000036741.27574.ce
ISSN1524-4539
Autores Tópico(s)Receptor Mechanisms and Signaling
ResumoHomeCirculationVol. 106, No. 17The Electrocardiogram 100 Years Later Free AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessReview ArticlePDF/EPUBThe Electrocardiogram 100 Years LaterElectrical Insights Into Molecular Messages Michael R. Rosen, MD Michael R. RosenMichael R. Rosen From the Department of Pharmacology, College of Physicians and Surgeons of Columbia University, New York, NY. Originally published22 Oct 2002https://doi.org/10.1161/01.CIR.0000036741.27574.CECirculation. 2002;106:2173–2179In 1921, Willem Einthoven invited Thomas Lewis to lecture in Leiden. Lewis responded, "The Leiden visit is one to which I am looking forward with great pleasure. It will be most stimulating to come to the Mecca of electrocardiography …"1 It is with much the same feeling that I approach this lectureship, honoring the memory of the man who, in developing the ECG, in effect ushered in the modern era in electrophysiology. In being here, I also pay homage to the Dutch tradition of electrophysiology, which has flourished since Einthoven's time.In 1930, several years after Einthoven's passing, the Russian electrophysiologist Samojloff was invited by Paul Dudley White (later the first Einthoven Lecturer) to reminisce about Einthoven's contributions. Samojloff stated, "… Einthoven worked almost exclusively in the field of electrophysiology. This branch of physiology stood for a long time completely isolated from life, medicine, and even from the general path of development of physiological knowledge; shut off in this way, electrophysiology could not progress and it seemed that it would be difficult to alter this sad situation of the study of animal electricity. [Einthoven's mind] … worked like an instrument of precision. He worked only on what could be measured and his measurements reached the limit of precision possible under the circumstances."2Einthoven's string galvanometer was remarkable in the fidelity and accuracy of its recordings and in its utility for interpreting the rhythm and state of health of the heart. Moreover, in applying the instrument to diagnosis, Einthoven foresaw events that wouldn't occur again for 50 years or more. Not only were patients' ECGs recorded, but the signals were wired the 1-mile distance from hospitalized patient to laboratory for registration and analysis. The off-site diagnostician was often able to identify a patient's affliction faster than the on-site physician.3Einthoven assessed the value of his invention rigorously; in 1922, he wrote to Lewis, "An instrument takes its true value not so much from the work it possibly might do but from the work it really does."4 A very practical approach, and in 1922 the ECG did more than enough to justify its use. Yet, we can only wonder whether a visionary like Einthoven or, indeed, anyone at that time, could foresee how much more the instrument held in store. Certainly, Lewis was not a visionary in this instance. In 1925, he diverted his attention from electrocardiography because he was "weary of being tied to an instrument … the cream was off … there were no longer new things to be discovered in electrocardiography."5Yet, during this period, electrocardiography was not only advancing, but exploration of its possibilities ushered in vectorcardiography,6 phonocardiography,7 and application to other organ systems that emit electrical signals.8 Moreover, electrocardiography has remained a centerpiece of clinical and investigative cardiology since its invention. For example, although nothing might be further from the interests of molecular biologists and geneticists than to register and interpret ECGs, those using mouse models to genetically manipulate cardiac proteins9 find the ECG a prime tool in their armamentarium. Also, as shall be described, their clinical counterparts have found diagnostic clues within the ECG's simple waveforms that reflect highly specific protein alterations in the heart.The ECG as Molecular Messenger: Early InsightsIt is difficult to date the origins of our understanding that electrocardiographic signals may reflect molecular change. That the ECG sums electrical activity generated by billions of cardiac myocytes was recognized from the time investigators began to record action potentials from cardiac cells.10 Adaptation of voltage clamp and single channel recording techniques to the heart heightened our appreciation of the roles of the cell membrane's integral proteins in generating action potentials. Studies of such inherited conditions as congenital long QT syndrome (LQTS)11–13 demonstrated that some genetic anomalies are expressed phenotypically on the ECG.Exemplifying the burgeoning awareness of relationships between molecular change and repolarization was Peter Schwartz's statement in 1986 that "The LQTS may be due to a primary abnormality in myocellular repolarization as a result of genetic alteration in the voltage dependent protein which regulates outward potassium current during phase 3 of the action potential."14 The possibility that different genetic variants of LQTS might be reflected in different repolarization patterns on the ECG was demonstrated by Moss et al in 1995.15 Although prediction of genotype by electrocardiographic phenotype was imperfect, subsequent investigation has suggested that it is often accurate.Investigators had earlier learned that the T wave reflects transmural and apico-basal gradients for ventricular repolarization16–18 and commenced to design frameworks for exploring the molecular determinism of the T wave19–22 (Figure 1). The results enabled scientists to link the structures of some protein subunits to their biophysical functions and to the ultimate expression of these functions in a cell or in the heart in situ.21–24 It is now only a matter of time before we unravel the needed information about all contributing structures. This might have been a rash prediction a decade ago; now, it is simply inevitable. Download figureDownload PowerPointFigure 1. Representative cardiac action potential and its biophysical and molecular determinants. Left, currents contributing to the action potential; right, genes for the α-subunits of the channels. A rough indicator of the temporal relationships of the currents to the voltage-time course of the action potential is presented in graphic form. Inward currents are downward deflections and outward currents, upward. Adapted from references 19 and 21.My initial recognition of this possibility was fueled by discussions with my mentor, Brian Hoffman, and was further directed by Mauricio Rosenbaum. Mauricio had been intrigued by observations of Chatterjee et al25 in patients manifesting persistent postpacing changes in T-wave morphology. Similar changes had been described in intermittent left bundle branch block, post-tachycardia syndromes, extrasystoles, and ventricular preexcitation.26–34 The fundamental observation that paced or spontaneous ventricular impulses could result in ST-T wave changes that persisted after returning to sinus rhythm was expanded and formalized by Rosenbaum and coworkers.35,36 Given that the T waves of the subsequent sinus beats followed the same vector as the paced or ectopic QRS complex, the provocative term "cardiac memory" was suggested; the T wave "remembered" the ectopic QRS.Mauricio and I would meet regularly on his trips to New York; we sat at the Algonquin's weathered tables, sipped coffee, and discussed life and research. What I cherished most about those afternoons was Mauricio's generosity as teacher, confidante, and friend. With respect to research, I was especially intrigued by his thoughts regarding cardiac memory. Yet whenever I considered designing protocols to follow-up on these, other commitments seemed to intervene. Then, one day Mauricio said, "Mike, you're getting older. You'd better start studying memory while you still have one." I'll now review what we have learned of the association of molecular message with electrical signal, using Einthoven's ECG to take Rosenbaum's advice.The ECG of Cardiac Memory and Its Ionic and Molecular DeterminantsOur studies of cardiac memory have been performed entirely in the dog and rat. When asked why not the human, I think of the remark attributed to Albert Szent-Gyorgi, "Life is a similar process in cabbages and kings; I choose to work on cabbages because they are cheaper and easier to come by."37 We can readily and reproducibly induce memory in canine heart via the use of cardiac pacemakers and, while the rat does not provide a T wave adequate for human application, it does permit us to explore pacing-induced molecular changes in a system presently more accessible to molecular/genetic study than the dog.Figure 2 depicts cardiac memory in a dog paced from the ventricle at a rate slightly faster than sinus for several weeks. On reversion to sinus rhythm, the T wave follows the direction of the paced QRS complex and persists in that pattern for long periods. No pathology, hypertrophy, hemodynamic, or circulatory change accompany this T-wave pattern,38 and it is altered activation, not increased heart rate, that induces cardiac memory.38–40Download figureDownload PowerPointFigure 2. Characteristics of cardiac memory. A, Lead I ECG from a dog chronically instrumented with a ventricular pacemaker and paced slightly faster than sinus. Traces recorded before starting to pace (control), during pacing (VP), and on return to sinus rhythm for 1 hour on days 7,14, and 21 after onset of pacing. The T wave on days 7 to 21 progressively inverts, tracking the paced QRS complex. Calibrations=1 mV, 0.25 sec. B, Frontal plane VCG for the same animal. Left, sinus rhythm; center, ventricular pacing; right, T-wave vectors from control and days 14 and 21 enlarged and superimposed. Note their shift to follow the paced QRS vector. C, Temporal accumulation of the T-vector change over 35 days (mean±SEM % change in angle) using 21 days as a reference point in 16 dogs. Change accumulates over two weeks, when a plateau is attained. D, Resolution of T-vector (amplitude) change during 30 days of recovery after cessation of pacing in two groups of 3 dogs, paced for 21 to 25 and 42 to 52 days, respectively. Peak T-wave vector amplitude is slightly greater than 0.8 mV at the end of pacing (day 0). R indicates the control before pacing. Both groups sustain the memory attained for several days after cessation of pacing. Memory declines thereafter, such that the group paced for the shorter period reaches values near control at 30 days, while the group paced for the longer period still is markedly different from control. Adapted from reference 38.Most of our work on cardiac memory has been in dogs such as those in Figure 2, paced for weeks and showing stable, persistent repolarization changes.38,41,42 Our initial goal was to test the relationship between the T-wave change and the transmural repolarization gradient; one of whose determinants, the transient outward potassium current, Ito, is large in ventricular epicardial myocytes and small in endocardium.43,44Figure 3 demonstrates that cardiac memory is associated with decreases in Ito density and mRNA for Kv4.3 (which encodes the α-subunit of the channel protein41) and a reduction in the epicardial action potential notch (determined by Ito). Also altered are Ito kinetics; activation voltage shifts positively, and recovery from inactivation prolongs about 20-fold.41 Accompanying these molecular/biophysical changes is an altered transmural repolarization gradient (Figure 3).41Download figureDownload PowerPointFigure 3. Relationship of the action potential (AP) changes of cardiac memory to the transmural gradient for repolarization and the biophysical and molecular determinants of the AP notch. A, Left ventricular epicardial free wall AP (cycle length=650 ms) for a control dog and one with cardiac memory. The control AP has a deeper notch, a lower plateau, and a shorter duration. B, mean AP durations to 50% and 90% repolarization measured from the LV epicardium (Epi) and endocardium (Endo) of a control dog and a dog with cardiac memory. Number of impalements is listed in the figure. Horizontal axis=drive cycle length (ms). In control, epicardial APs are shorter than endocardial APs, and a large gradient exists between them. In memory, both epicardial and endocardial APs are prolonged, the former more than the latter. As a result, the gradient between them decreases. This change would contribute to an altered T wave. C, Conductance of the channel carrying Ito (see Figure 1) in epicardial myocytes from dogs with cardiac memory (black circles, n=19) and controls (white circles, n=26). There is significantly greater conductance (reflecting current) throughout the range of physiologically relevant membrane potentials (−20 to 20 mV) in the controls. D, RNase protection analysis of mRNA for Kv4.3, the gene responsible for the α-subunit of the channel carrying Ito in the dog (see Figure 1). Lanes are LV epicardium from 3 control and 3 memory animals. Bands on the bottom are cyclophylin internal controls. On the right, control values for Kv4.3 are expressed as 100% and the values for the memory animals as a percent of these. There is about a 1/3 reduction in message for Kv4.3. Adapted from references 38 and 41.If Ito is a major factor in induction of cardiac memory, then its absence should prevent memory from occurring. We tested this supposition by administering the Ito blocker 4-aminopyridine to intact dogs39 or isolated tissues,45 after which memory could no longer be induced. The role of Ito was also tested using a developmental approach, deriving from our earlier observation that myocytes from dogs under two months of age do not manifest Ito or a phase 1 notch.46 The magnitude of the memory induced increased in parallel with evolution of the phase 1 notch over the first 120 days of life.47However, Ito is not the only current to change in cardiac memory. ICa,L density is comparable to controls, but its activation and recovery from inactivation are altered in a manner consistent with the elevation and prolongation of the action potential plateau (Figure 3A).48 Memory is also associated with reduction in density of the gap junctional protein connexin43 (Cx43) and with changes in Cx43 distribution, from concentration at the longitudinal poles of myocytes to more uniform patterning across the lateral cell margins.49 These changes in Cx43 are regionally nonuniform; ie, they are greater epicardially than endocardially and near to rather than far from the pacing electrode.49 These factors likely all contribute to the altered transmural repolarization gradient.How Are Pacing-Induced Signals Transduced?Understanding the transduction of signals requires a road map, commencing with cellular excitation and terminating at end-organ response. We know that pacing the ventricle alters its pathway for activation and stress/strain relationships in the myocardium.50–52 We also know that changing stress/strain relationships on cardiac cell cultures increases release of angiotensin II,53,54 a hypertrophic hormone that modulates cardiac electrical activity and structure.55,56 Experiments on memory of short duration in dogs demonstrated that ACE inhibition or AT-1 receptor blockade prevent its initiation.52 Moreover, incubation of epicardial ventricular myocytes from control dogs with angiotensin II for hours/days results in changes in the action potential notch and Ito magnitude and kinetics identical to those described above for cardiac memory.57 Together, these results suggest a central role for angiotensin II in the genesis of cardiac memory.However, the story is more complex. Despite the effect of angiotensin-converting enzyme (ACE) inhibition or AT-1 receptor blockade to attenuate memory initiated by brief periods of pacing, neither intervention prevents the effects of 3 weeks of pacing to induce memory of long duration (unpublished data). This suggests that while angiotensin II may initiate memory, it does not sustain the process of accumulation. In seeking other processes involved, a significant role for calcium has been uncovered. Specifically, administration of ICa,L blockers suppresses memory induced by brief or protracted periods of pacing (preliminary data). The Ca-associated effect is not altered by β-adrenergic blockade, and may be triggered by another process (perhaps angiotensin II's effect to increase ICa,L).58Another complexity is that Kv4.3 mRNA is unchanged in myocytes exposed to angiotensin II,57 in contrast to pacing-induced memory for which Kv4.3 mRNA is reduced.41 If angiotensin II were truly responsible for pacing-induced memory, then angiotensin II might be expected to reduce Kv4.3 mRNA (which does not occur) or to initiate other processes that reduce either Kv4.3 message or protein. Recent data suggest angiotensin II induces enzymes involved in protein degradation via a negative feedback loop, regulating and limiting hormonal production.59 Such degradation could explain the reduced message for Kv4.3 in our pacing studies but not the failure to see a reduction of Kv4.3 when myocytes are incubated in angiotensin II. Whether additional changes occur at the level of channel protein synthesis and/or phosphorylation needs to be tested, as does the possibility that the changes seen in memory result from accelerated protein degradation.In summary, we initially hypothesized angiotensin II to be the sole initiator of cardiac memory. The temptation was then to see all of memory as evolving from this single initiator. The problem was that some of our data were at variance with our initial idea, prompting much discussion and some confusion within our research team. Einthoven (quoted by Carl Wiggers)60 had advice to give on this subject: "When evidence is at variance … spend more time in attempting to harmonize differences rather than … direct … energies toward procurement of more data for a favored hypothesis. The truth is all that matters; what you or I think is inconsequential."Neurons as Paradigms: the Role of Nuclear FactorsThus far, I have described a partially tested and circumscribed role for angiotensin II and a more diverse but as yet uncharted role for Ca in cardiac memory. In seeking an interface with transcriptional elements in the heart, we adopted memory in the central nervous system as a paradigm, particularly, studies in the sea slug aplysia used by neuroscientists to study long-term facilitation. Facilitation is induced in aplysia by electrical shocks or serotonin administration, both of which initiate phosphorylation and synthesis of the cAMP response element binding protein (CREB), a pivotal transcriptional factor in the synthesis of new protein that sustains memory.61–63 Initiation of memory commences rapidly after stimulation, such that a single electrical shock or pulse of serotonin exerts a measurable effect, the magnitude of change of which increases over minutes.Whereas in the central nervous system (CNS), processes involved in long term facilitation are initiated and up-regulated over seconds to minutes, cardiac memory appears to be more associated with down-regulation than synthesis (eg, Ito,41 Cx4349). Although CREB phosphorylation initially increases on pacing the canine or rat heart,64,65 pacing the dog for 120 minutes results in decreased CREB levels.65 The CNS parallel, then, might be long-term depression instead of facilitation.Decreases in CREB levels might be expected if CREB were an upstream transcriptional factor whose down-regulation were required for expression of cardiac memory or if a factor or factors not yet identified were down-regulating both. That either possibility may prove true is suggested by the identification of an apparent degradation product of CREB increasing in amount as memory evolves and as CREB levels decrease (preliminary data). In any event, the sequence of CREB changes parallels the temporal course of the ECG changes of cardiac memory.65Clinical Applications of Cardiac MemoryThe final point I will consider is the potential clinical meaning of cardiac memory, summarized as responses to four questions: (1) To the extent that angiotensin II is involved in at least the onset of memory, might memory be an initial sign of hypertrophy to come? This is possible, although at least two observations are in opposition: (1) the failure to demonstrate long-term involvement of angiotensin II in cardiac memory; and (2) the demonstration of feedback loops that ultimately limit angiotensin II availability.59(2) Might cardiac memory impact on antiarrhythmic drug treatment? Ito and IKr-blocking drugs suppress the occurrence of memory, and memory itself (as might be induced by a ventricular arrhythmia) alters the effects of some drugs, eg, quinidine.42 Hence, memory modulation by drugs and the effect of memory to alter drug actions should be considered as determinants of the expression of drug effects.(3) Might pacing to induce cardiac memory be antiarrhythmic? We and others have shown that pacing of the sort that induces memory and/or altered stretch can modulate ventricular and atrial effective refractory periods in ways that might be anti- or proarrhythmic.66–68(4) Is there atrial memory, and if so, might it impact on atrial arrhythmias?69–71Figure 4 demonstrates atrial memory in dogs in complete heart block and chronically instrumented with left or right atrial pacemakers. Atrioventricular (AV) sequential pacing is employed to prolong the PR interval, facilitating recording of the Ta wave. We use modified Frank leads to record P and Ta waves and have adapted the concept of the ventricular gradient72–74 to calculate an atrial gradient (AG)69,75,76 as follows: Download figureDownload figureDownload PowerPoint Figure 4. Measurement of atrial memory via the PTa wave time integral. A, ECG of a dog in chronic AV block, instrumented with an AV-sequential pacemaker. Left, orthogonal X lead showing the signal-averaged, paced P wave and QRS complex. The PR interval is sufficiently long to reveal the entire Ta wave. Right, Enlargement of P and Ta waves, with measured parameters indicated. B, protocols for studying atrial memory. Six animals were instrumented with left and right atrial pacemaker leads. All 3 protocols commenced with right atrial pacing (RAP) for 45 minutes at 111 bpm. To test the effect of altering activation pathway only (protocol 1), during an initial and a second test pacing period, animals were paced from the left atrium (LAP) at 111 bpm. In protocol 2, the effect of altering rate only and leaving activation unchanged was studied by performing RAP during both test-pacing periods at 160 bpm. In protocol 3, effects of altering rate and activation were studied by performing LAP at 160 bpm during both test periods. All postpacing recovery periods incorporated RAP at 111 bpm. C, PTa wave time integrals in protocols 1 to 3. Altering rate alone leaves the integral unchanged (ie, there is no memory). Altering activation alone increases the integral in a way that accumulates further in the second postpacing recovery period. When rate and activation are both altered, the magnitude and accumulation of memory are greatest. Adapted from reference 69. When animals are paced for two hours at rates about 5% faster than sinus (Figure 4), right atrial pacing is unassociated with a changed atrial gradient, whereas left atrial pacing induces a change that accumulates in magnitude over time and as pacing rate increases. We interpret this phenomenon as atrial memory.69Left and, to a lesser extent, right atrial pacing at rates slightly faster than sinus for 3 to 4 weeks induces further changes in atrial gradient.70 Moreover, after 24 hours of right or left atrial pacing, competing atrial tachycardias arise. These persist throughout the pacing period during right atrial pacing and cease when pacing is discontinued. Similar arrhythmias occur during left atrial pacing, but these persist during postpacing recovery and are associated with further change in the atrial gradient.70 Most recently, we have observed that with very rapid atrial pacing (600 to 900 bpm), the atrial gradient continues to evolve until atrial fibrillation occurs.71 At these rapid rates, the atrial gradient is altered regardless of whether pacing is right- or left-sided.71The key to understanding the potential for diagnostic or predictive use of the atrial gradient is that it begins to change in response to pacing before changes in the effective refractory period are noted.70,71 Hence, change in the gradient may represent a noninvasive "early warning" of impending tachyarrhythmia, including fibrillation. However, before waxing too enthusiastically over possibilities here, I might well paraphrase Einthoven's admonition quoted earlier;4 the true value of any technique lies not in what it promises but in what it delivers.Closing CommentsIn preparing this talk, I read extensively of Einthoven's life, reread many of his papers, and reviewed some of his correspondence. What impressed me was not only the exemplary caliber of his work but the very human and thoughtful qualities that were commented on by his contemporaries and that come across so clearly in his writings. A characteristic especially meaningful to me has been Einthoven's profound sense of internationalism, which is all the more remarkable in that it persisted through the years leading to, during, and following the Great War. This sentiment is evidenced with exceptional clarity in the closing words of Einthoven's Nobel Prize lecture77: "… a new chapter has been opened in the study of heart diseases, not by the work of a single investigator, but by that of many talented men, who have not been influenced in their work by political boundaries and, distributed over the whole surface of the earth, have devoted their powers to an ideal purpose, the advance of knowledge by which, finally, suffering mankind is helped."As noted by so many in electrophysiology, some of our closest personal and professional bonds and some of our finest moments have been achieved in the company of others born far from us and often living far from us. If ever there were a model demonstrating that disparate peoples can find common ground and kinship despite the distances of culture and personal belief, then our field has shown this, as Einthoven recognized a century ago and as is apparent to this day. With these sentiments in mind, I would like to acknowledge and thank the diverse cast of students, fellows, and faculty who have contributed enthusiastically and selflessly to the work reported here (and whose names may be found in the references), and the larger community of colleagues who have made this career so rewarding.This is the abridged text of the 23rd Einthoven Lecture delivered by the author on receipt of the Einthoven Award, June 10, 2002. The lecture was delivered at a symposium under the auspices of the Einthoven Foundation and the University of Leiden, celebrating the 100th anniversary of Willem Einthoven's invention of the electrocardiogram.My thanks to Charlie Antzelevitch, Martin Schalij, Peter Schwartz, and Albert Waldo, who found background materials for me; to Ira Cohen, Peter Danilo, David McKinnon, Richard Robinson, and Susan Steinberg, my long-term partners in program research; and Eileen Franey for her careful attention to the preparation of the manuscript. Finally, my special thanks to the National Heart, Lung and Blood Institute, whose support over the years has enabled the performance of our research.FootnotesCorrespondence to Michael R. Rosen, MD, Gustavus A. Pfeiffer Professor of Pharmacology, Professor of Pediatrics, Director, Center for Molecular Therapeutics, College of Physicians and Surgeons of Columbia University, 630 West 168 St, PH7West-321, New York, NY 10032. E-mail [email protected] References 1 Lewis T. Quoted by: Snellen HA. Two Pioneers of Electrocardiography. Rotterdam, the Netherlands: Donker Academic Publications; 1983:100.Google Scholar2 Samojloff A. Reminiscences of the late Professor Willem Einthoven. Am Heart J. 1930; 5: 545–548.CrossrefGoogle Scholar3 Einthoven W. Le télécardiogramme. Arch Internat Physiol. 1906; 4: 132–164.Google Scholar4 Einthoven W. Quoted by: Snellen HA. Two Pioneers of Electrocardiography. Rotterdam, the Netherlands: Donker Academic Publications; 1983:48.Google Scholar5 Lewis T, Quoted by: Snellen HA. Two Pioneers of Electrocardiography. Rotterdam, the Netherlands: Donker Academic Publications; 1983:26.Google Scholar6 Williams HB. On the cause of the phase difference frequently observed between homonymous peaks of the electrocardiogram. Am J Physiol. 1914; 35: 292–300.CrossrefGoogle Scholar7 Einthoven W. Die Registrirung der menschlichen herztöne mittels des saitengalvanometers. Arch ges Physiol. 1907; 117: 461–478.CrossrefGoogle Scholar8 Einthoven W, Jolly WA. The form and magnitude of the electrical response of the eye to stimulation by light at various intensities. Quart J Exper Physiol. 1908; 1: 373–416.CrossrefGoogle Scholar9 Rakhit A, Berul CI. Transgenic electrophysiology and small animal models. In: Spooner PM, Rosen MR, eds. Foundations of Cardiac Arrhythmias. New York, NY: Marcel Dekker Inc; 2000:651–664.Google Scholar10 Hoffman B. And then came the microelectrode. Cardiovasc Res. 2002; 53: 1–5.CrossrefMedlineGoogle Scholar11 Romano C, Gemme G, Pongiglione R. Aritmie cardiache rare in eta pediatrica. Clin Pediatr. 1963; 45: 656–683.Google Scholar12 Ward OC. A new familial cardiac syndrome in children. J Irish Med Assoc. 1964; 54: 103–106.MedlineGoogle Scholar13 Jervell A, Lange-Nielsen F. Congenital deaf-mutism, f
Referência(s)