Pacemaker Syndrome: an Iatrogenic Condition

Br Heart J 1992;68:163-6 163

VIEWPOINT

Pacemaker syndrome: an iatrogenic condition

Christopher M Travill, Richard Sutton

Pacemaker syndrome was first described in inhibited (VVI) pacing was used. In the re- 1969 by Mitsui when it was referred to as the maining 20 patients with no history of pacemaking syndrome.' The name pacemaker pacemaker syndrome or ventriculoatrial con- syndrome was first coined by Erbel (using the duction Doppler derived cardiac output German Schrittmacher syndrom) in 1979.2 It improved by 14% when pacing was changed can present with symptoms as severe as to DDD from VVI.8 syncope, presyncope, oedema, dyspnoea, and chest pain, or more moderately and subtly as lethargy, palpitation, or an awareness of Haemodynamics of pacemaker venous pulsation all of which may occur when syndrome there is atrial systole during ventricular sys- Contraction of the atria against closed atrio- tole. This is most frequent when there is ventricular valves during ventricular systole ventriculoatrial conduction from the paced leads to raised atrial pressures, loss of atrial ventricle to the atrium usually via the contribution to ventricular filling, and a con- atrioventricular node.34 The symptoms can be sequent fall in cardiac output which causes a identical to those prompting implant.5 fall in arterial pressure if the baroreceptor Occasionally pacemaker syndrome can occur mediated rise in systemic vascular resistance is in the absence of ventriculoatrial conduction insufficient.9 In a study on 20 open-chest dogs when ventricular pacing is in competition with experimental complete heart block left with sinus rhythm.6 atrial angiography showed retrograde blood flow into the pulmonary venous system at Incidence of pacemaker syndrome atrioventricular intervals of -50 and The incidence of pacemaker syndrome varies - 100 ms. Therefore in addition to the loss of with the vigour with which it is sought and it atrial contribution to ventricular filling there is probably affects 7% of all ventricularly paced a "negative atrial kick" further compromising patients in its severe form in which it is haemodynamic function.'0 In three patients essential to revise the pacemaker. If mild to studied by Alicandri et al" there was an moderate symptoms are considered it affects absent or smaller rise in peripheral resistance 20% of the ventricularly paced. This group in response to a fall in cardiac output. The too can benefit clinically from pacemaker mechanism was thought to be due to vaso- upgrade.7 All too often the symptoms and dilatation in response to activation of atrial signs of the syndrome are not sought and stretch receptors by atrial cannon waves patients who previously experienced syncope, which dominated over the baroreceptor now relieved, complain little. This leads to a mediated increase in resistance that occurs widespread impression that the syndrome is secondary to the fall in systemic arterial pres- rare. sure. Similar findings were observed in a group of 20 patients studied haemodynam- ically more than 24 hours after coronary Clinical diagnosis of pacemaker artery bypass grafting or aortic valve re- syndrome placement. Hypotension with ventricular pac- The diagnosis is made by reproduction of ing occurred only in those patients with left symptoms during ventricular pacing and atrial cannon waves.'2 It has been suggested depends on the history and the search for that patients with left ventricular disease, hypotension, signs of congestive cardiac especially hypertrophy of any cause, are more Department of failure, and venous cannon waves associated sensitive to the correct timing of atrial systole Cardiology, with ventricular pacing. Usually the diagnosis and are, therefore more liable to be sympto- Westminster Hospital, can be made without recourse to with atrioventricular con- London clinically matic retrograde C M Travill special investigations. However, Doppler duction.7 R Sutton echocardiography can be useful in its A recent study of a large group of patients Correspondence to diagnosis: Doppler ultrasound measurement with intact ventriculoatrial conduction Dr R Sutton, Department of of cardiac output in nine patients with showed similar results, with peripheral resis- Cardiology, Westminster Hospital, Dean ventriculoatrial conduction or symptoms con- tance failing to rise in seven patients requiring Ryle Street, London SWIP sistent with pacemaker syndrome showed a upgrade to dual chamber mode because of 2AP. 30% when fully automatic of whereas it Accepted for publication improvement symptoms pacemaker syndrome 12 February 1992 (DDD) pacing rather than ventricular rose during ventricular pacing in the symptom 164 Travill, Sutton

free group.'3 The difference between the re- identified during 800 head up tilt than with the sponse of the cardiac output between the two patient supine. In those patients in whom the groups was negligible and the authors argued interatrial conduction delay exceeded 150 ms that the behaviour of the blood pressure was (three of 16 patients) a programmed atrioven- crucial. They advised that a cuff recording of tricular delay of 150 ms resulted in left atrial blood pressure should be made at the onset of activation after the ventricular spike, yielding ventricular pacing during pacemaker implant left atrial contraction during left ventricular and that a fall in systolic pressure of greater systole. The temporal difference between sens- than 25 mm Hg should be regarded as predic- ing of spontaneous right atrial activation and tive for the possible development of right atrial pacing is such that at a given pacemaker syndrome and they recommended atrioventricular interval, the sequence between dual chamber pacing.'3 We suggest that if atrial and ventricular contraction is longer in clinicians allow their choice of pacing mode to the atrial synchronous ventricular inhibited be guided by haemodynamic variables during (VDD) than in the atrioventricular sequential temporary ventricular pacing before implant, mode (DVI). Some pacemakers now take this they must assess the patient in the upright delay into account as a programmable feature. posture (600 head up tilt with appropriate The DDI mode has been recommended for safety restraint) before considering it appro- patients with carotid sinus syndrome, malig- priate to implant a VVI unit.'4 nant vasovagal syndrome, and sick sinus syndrome who do not require atrial tracking (ventricular pacing as a result of atrial sensing) and Pacemaker syndrome in different pacing who often have ventriculoatrial conduction.2224 modes In this mode it is impossible for pacemaker In a small group of patients, paced in VVI mediated tachycardia to occur but pacemaker mode and presenting with symptoms of syndrome can occur with non-conducted pacemaker syndrome, changing the mode from premature atrial beats or with retrograde con- VVI to atrial inhibited (AAI) relieved symp- duction from premature ventricular beats.25 toms and was associated with a fall in both the Either of these events will inhibit atrial output mean right atrial and pulmonary capillary if they occur outside the post ventricular atrial wedge pressures as well as a rise in cardiac refractory period. They are then followed by a output."' In view of the high prevalence of ventricular stimulus at the programmed ven- ventriculoatrial conduction in patients with tricular rate which may be conducted sinus node disease, AAI pacing with the pos- retrogradely if the atrium has had sufficient sible addition of a sensor-driven facility for time to recover. Ventricular pacing will only those patients who do not show an increase in then be inhibited if the sinus rate recovers and sinus rate on exertion (chronotropic incom- is normally conducted. In patients with petence) is advocated by some for all patients in documented ventriculoatrial conduction the whom atrioventricular conduction is sound.'6 atrial refractory period in DDI should be Two large randomised controlled trials to programmed long enough to include the compare ventricular and dual chamber pacing retrograde conduction interval. Thus sensing in patients with sinus rhythm identified of a retrograde P wave and atrial output inhibi- improved well-being even in "asymptomatic" tion is avoided. It is suggested that the atrial patients when the dual chamber mode was refractory period be set to 325 ms at program- used.'718 However, ventricular pacing is the med rates of 55-85 pulses per minute (ppm), preferred mode in patients with symptomatic shorter for faster rates, and longer for slower bradycardia in the presence ofatrial fibrillation. rates. Although the early dual chamber pacing mode (atrial synchronous ventricular inhibited (VDD)) was an improvement over the atrio- Pacemaker syndrome in rate responsive ventricular sequential (DVI) mode, it was still pacing associated with pacemaker syndrome in those Sensor driven ventricular pacing (VVIR) patients whose sinus rate dropped below the would be expected to result in pacemaker programmed lower rate at which point the syndrome at resting heart rates in the same way pacing mode effectively became VVI."9With the as non-sensor driven ventricular pacing. In advent of fully automatic dual chamber pacing patients with sinus node chronotropic incom- systems (DDD) it became apparent that petence, a group of patients who are often pacemaker syndrome can still occur in properly considered suitable for the VVIR mode, ven- functioning dual chamber pacing systems in tricular pacing is often programmed so as to be the DDD mode if the atrioventricular delay is inhibited at rest but triggered by the sensor programmed either too short or too long.20 The during exercise. This may result in pacemaker importance of the timing of left atrial depolar- syndrome during exercise with retrograde isation was addressed in a study of 16 patients atrial activation as the rate of the paced rhythm with dual chamber pacemakers who had exceeds that of the intrinsic rhythm.26 Syncope haemodynamic variables measured at different on exercise has been reported even in the atrioventricular delays when they were supine absence of regular ventriculoatrial conduc- and erect.2' A wide range of interatrial conduction." tion delay (70 ms-380 ms using oesophageal recording) between right atrial pacing artefact Sensor driven AAI pacing causing pacemaker and left atrial depolarisation was seen and the syndrome-The onset of pacemaker syndrome optimal atrioventricular delay was more easily was reported during exercise in a patient with a Pacemaker syndrome: an iatrogenic condition 165

dual atrioventricular nodal pathway in whom patients who are mostly in sinus rhythm and conduction through the fast anterograde path- only show rare episodes of conduction defect way was prolonged to such an extent by con- and also for patients who are either severely comitant therapy with disopyramide that disabled and show no retrograde conduction or conduction occurred over the slow pathway.28 who have a poor prognosis because of another This resulted in pacemaker syndrome during disease. moderate exercise when the paced atrial event As the use of dual chamber pacing systems was conducted with a long spike Q interval with the ability to sense P waves and at least one leading to atrial systole just after the preceding other physiological variable (by means of sen- QRS during ventricular systole. sors such as those for respiratory minute volume and QT interval) increases, a far lower incidence of pacemaker syndrome should Treatment of pacemaker syndrome result. Furthermore, development of pace- Because symptoms of pacemaker syndrome maker artificial intelligence to diagnose the overlap with those ofpacemaker malfunction it syndrome and automatic adjustment of the is vital to exclude malfunction as the first step in pacing mode to avoid it is expected within a few investigation. Optimal therapy for sympto- years. matic bradycardia of any cause is use of a pacing system that includes atrial sensing and pacing wherever atrial electrophysiology per- Alternatives to dual chamber pacing mits. In practice this will frequently be a dual therapy of pacemaker syndrome chamber device where the atrial refractory (a) If pacemaker syndrome is encountered, it period should be programmed to at least 25% may be ameliorated by reducing the pacing rate longer than the measured ventriculoatrial con- so that competition between sinus rhythm and duction time to avoid pacemaker mediated or pacing is minimal with the possible addition of endless loop reentry tachycardia. Furman's rate hysteresis in which pacing will only be group evaluated ventriculoatrial conduction in triggered after a pause significantly longer than 432 patients receiving a permanent pace- the pacing interval.30 maker.29 One hundred and sixty two had intact (b) Antiarrhythmic drugs may be used to block ventriculoatrial conduction including 14% of retrograde atrioventricular conduction. patients with complete anterograde atrioven- Flecainide has previously been used for this tricular block. Most patients with intact purpose but in the light of the CAST study we anterograde conduction had ventriculoatrial do not think that it can be recommended.3' It conduction. They evaluated these patients by and other antiarrhythmic drugs can also para- means of incremental ventricular pacing and doxically result in artificial circus movement observed that the ventriculoatrial conduction tachycardias by prolonging the retrograde interval was prolonged in most at faster rates atrio-ventricular conduction time without with complete retrograde ventriculoatrial block complete retrograde block and may result in a at rates exceeding 120 pulses per minute in retrograde P wave occurring outside the atrial 50% of patients with intact atrioventricular refractory period thereby precipitating pace- conduction at rest. They estimated that maker mediated tachycardia.32 programming the postventricular atrial refrac- (c) In those patients who are not pacemaker tory period to 300 ms, an upper rate limit of dependent and in whom the pre-implant symp- 140 ppm and an atrioventricular delay of toms are minor serious consideration should be 125 ms would prevent endless loop tachycardia given to the device being turned off and sub- in 90% of their patients. sequently, if appropriate, explanted. Pacemaker syndrome can be avoided pros- pectively by choice of the most physiologically appropriate pacing mode for the individual New approach to the identification of patient. While this will inevitably lead to a potential pacemaker syndrome higher incidence of dual chamber pacing and If a policy of DDD pacing for all suitable increased cost, the cost ofupgrading as many as patients is not adopted a reliable test to identify 20% of ventricularly paced patients must be patients at risk of developing pacemaker syn- considered together with the technical difficul- drome would be clinically valuable. ties of passing a second lead and the increased Atrial natriuretic peptide was discovered in risk of infection that is attendant upon second 1981 and is released directly into the circulation and subsequent pacemaker procedures. in response to atrial wall stretch within the Increased cost need not be enormous if AAI heart and it has potent vasodilator properties in pacing is considered as first choice mode for addition to natriuretic and diuretic ones.33 patients with sick sinus syndrome. Further- Peripheral venous plasma atrial natriuretic more, the cost of many excellent dual chamber peptide concentrations were measured in units has now fallen to within the range ofmany patients with DDD and VVI pacemakers and of the most sophisticated single chamber units. they were significantly lower in both the short If improved physiological performance of dual and long term when the DDD mode was in chamber pacemakers with well demonstrated operation and the concentrations were similar improved quality of life'7 18 is ignored and to those observed in controls.34 This difference patients most likely to have retrograde conduc- is maintained during exercise.35 Concentra- tion are targeted the increased cost is moderate tions of atrial natriuretic peptide were lower at and comparable with the need for VVI up- physiological delays than at either long or short grade. VVI pacing can still be considered for atrioventricular delays.36 Atrial natriuretic Sutton 166 Travill, were measured in regurgitation during abnormal atrioventricular sequenc- peptide concentrations ing. Am Heart J 1983;105:295-302. patients with ventriculoatrial conduction both 11 Alicandri C, Fouad FM, Tarazi RC, Castle L, Morant V. with and without symptoms of pacemaker Three cases of hypotension and syncope with ventricular pacing: possible role of atrial reflexes. Am J Cardiol 1978; syndrome. Concentrations were normal during 42:137-42. sinus rhythm or DDD pacing (atrioventricular 12 Erlebacher JA, Danner RL, Stelzer PE. Hypotension with ventricular pacing: an atrial vasodepressor reflex in human delay 150 ms) in patients with pacemaker syn- beings. J Am Coll Cardiol 1984;4:550-5. but rose to 7-8 fold during ventricular 13 Witte J, Bondke H, Muller S. The pacemaker syndrome: a drome haemodynamic complication of ventricular pacing. Cor pacing with intact ventriculoatrial conduc- Vasa 1988;30:393-9. tion.37 These data prompt reconsideration of 14 Fitzpatrick AP, Travill CM, Vardas PE, et al. Recurrent symptoms after ventricular pacing in unexplained syn- the mechanism of hypotension in pacemaker cope. PACE 1990;13:619-24. syndrome. Animal studies38 (which have not 15 El Gamal MIH, Van Gelder LM. Chronic ventricular pacing with ventriculo-atrial conduction versus atrial been reproduced by others39) offer only weak pacing in three patients with symptomatic sinus brady- support to the suggestions of Alicandri et al, " cardia. PACE 1981;4:100-5. 16 Ryden L. Atrial inhibited pacing-an underused mode of Erlebacher et al,'2 and Witte et al,'3 that the cardiac stimulation. PACE 1988;1 1:1375-9. mechanism is an exaggerated or inappropriate 17 Kruse I, Arrman K, Conradson TB, Ryden LA. A compar- ison of the acute and long-term haemodynamic effects of neural reflex. Atrial natriuretic peptide has well ventricular inhibited and atrial synchronous ventricular properties at concentrations inhibited pacing. Circulation 1982;65:846-55. defined vasodilator 18 Perrins EJ, Morley CA, Chan SL, Sutton R. Randomised seen in patients with symptomatic pacemaker controlled trial of physiological and ventricular pacing. Br effects may be additive to the Heart J 1983;50:112-7. syndrome. These 19 Levine PA, Seltzer JP, Pirzada FA. The "pacemaker syn- neural reflex or may be the sole mechanism of drome" in a properly functioning physiological pacing vasodilation. The higher con- system. PACE 1983;6:279-82. inappropriate 20 Torresani J, Ebagosti A, Allard-Latour G. Pacemaker centrations found in patients with symptomatic syndrome with DDD pacing. PACE 1984;7:1148-51. syndrome compared with those who 21 Wish M, Fletcher RD, Gottdiener JS, Cohen AI. Import- pacemaker ance of left atrial timing in the programming of dual had retrograde atrioventricular conduction chamber pacemakers. Am J Cardiol 1987;60:566-71. without supports the involvement of 22 Floro J, Castellanet M, Florio J, Messenger J. DDI: A new symptoms mode for cardiac pacing. Clin Prog Pacing Electrophysiol an endocrine mechanism in this syndrome. 1984;2:255-60. Pacemaker syndrome particularly in its 23 Barold SS. The DDI mode of cardiac pacing. PACE 1987; 10:480-4. milder form probably has a higher incidence 24 Sutton R. Pacing in patients with carotid sinus and vasovagal and prevalence than is currently realised. It syndromes. PACE 1989;12:1260-3. 25 Cunningham TM. Pacemaker syndrome due to retrograde may be avoided by carefully programmed dual conduction in a DDI pacemaker. Am Heart J 1988;115: chamber pacing for all patients with normal 478-9. 26 Liebert HP, O'Donoghue S, Tullner WF, Platia EV. atrial activity and atrioventricular conduction Pacemaker syndrome in activity-responsive VVI pacing. disease and by atrial pacing in patients with Am J Cardiol 1989;64:124-6. 27 Wish M, Cohen A, Swartz J, Fletcher R. Pacemaker intact atrioventricular conduction. Pacemaker syndrome due to a rate-responsive ventricular pacemaker. upgrade should be avoided on grounds of J Electrophysiol 1988;2:504-7. 28 Den Dulk K, Lindemans FW, Brugada P, Smeets JLRM, patient discomfort (including infection risk), Wellens HJJ. Pacemaker syndrome with AAI rate variable operator difficulty in passing a second lead, and pacing: importance ofatrioventricular conduction properties, medication and pacemaker programmability. PACE cost. Thus a policy of VVI pacing for all is no 1988;11:1226-33. longer tenable. 29 Klementowicz P, Ausubel K, Furman S. The dynamic nature of ventriculo-atrial conduction. PACE 1986;9: 1050-54. 30 Preston TA, Preston AW Jr. The automatic rate adjustment 1 Mitsui T, Hori M, Suma K, Wanibuchi Y, Saigusa M. The pacemaker. The possibilities of rate hysteresis. PACE "pacemaking syndrome" [abstract]. In: Jacobs JE, eds. 1978;1:178-85. Proceedings ofthe eighth annual international conference 31 The CAST investigators, Preliminary report: Effect of on medical and biological engineering. Chicago: Associa- encainide and flecainide on mortality in a randomized trial tion for the Advancement of Medical Instrumentation. of arrhythmia suppression and myocardial infarction. N 1969;29:3. EnglJ Med 1989;321:406-12. 2 Erbel R. Pacemaker syndrome. Am J Cardiol 1979;44:771-2. 32 Bathen J, Gundersen T, Forfang K. Tachycardias related to 3 Werres R, Parsonnet V, Gilbert L, Zucker IR. Symptomatic atrial synchronous ventricular pacing. PACE 1982;5: unilateral cannon "a" waves in a patient with a ventricular 471-5. pacemaker. Chest 1978;73:539-43. 33 De Bold AJ, Borenstein HB, Veress AT, Sonnenberg H. A 4 Johnson AD, Kaiken SL, Engler RL. Hemodynamic com- rapid and potent natriuretic response to intravenous promise associated with ventriculo-atrial conduction fol- injection of atrial myocardial extract in rats. Life Sci lowing transvenous pacemaker placement. Am J Med 1981;28:89-94. 1978;65:75-9. 34 Vardas PE, Travill CM, Williams TDM, Ingram A, 5 Amikam S, Riss E. Untoward haemodynamic consequences Lightman SL, Sutton R. Effect ofdual chamber pacing on of permanent ventricular pacing associated with ven- raised plasma atrial natriuretic peptide concentrations in triculo-atrial conduction. In: Meere C, ed. Proceedings of complete atrioventricular block. BMJ 1988;296:94. VI world symposium on cardiac pacing. PACESYMP, 35 Stangl K, Weil J, Seitz K, Laule M, Gerzer R. Influence of 1979, chapter 15-6. AV synchrony on the plasma levels of atrial natriuretic 6 Nishimura RA, Gersh BJ, Vlietstra RE, Osborn MJ, Ilstrup peptide (ANP) in patients with total AV block. PACE DM, Holmes DR. Haemodynamic and symptomatic con- 1988;11:1 176-81. sequences of ventricular pacing. PACE 1982;5:903-10. 36 Noll B, Krappe J, Goke B, Maisch B. Influence of pacing 7 Ausubel K, Furman S. The pacemaker syndrome. Ann mode and rate on peripheral levels of atrial natriuretic Intern Med 1985;103:420-9. peptide (ANP). PACE 1989;12:1763-68. 8 Stewart WJ, Dicola VC, Harthorne JW, Gillam LD, 37 Travill CM, Williams TDM, Vardas P, et al. Hypotension in Weyman AE. Doppler ultrasound measurement ofcardiac pacemaker syndrome is associated with marked atrial output in patients with physiologic pacemakers. Am J natriuretic peptide (ANP) release [abstract]. PACE Cardiol 1984;54:308-12. 1989;12:93. 9 Nishimura RA, Gersh BJ. Holmes DR Jr, Vlietstra RE, 38 Edis AJ, Donald DE, Shepherd JT. Cardiovascular reflexes Broadbent JC. Outcome of dual-chamber pacing for the from stretch ofpulmonary vein-atrial junctions in the dog. pacemaker syndrome. Mayo Clin Proc 1983;58:452-6. Circ Res 1970;27:1091-100. 10 Naito M, Dreifus LS, David D, Michelson EL, Mardelli TJ, 39 Carswell F, Hainsworth R, LedsomeJR. Effects ofdistension Kmetzo JJ. Reevaluation of the role of atrial systole to of the pulmonary vein-atrial junctions upon peripheral cardiac haemodynamics: Evidence for pulmonary venous vascular resistance. J Physiol 1970;207:1-14.