Embed Size (px)
Citation preview
Design-Dependent Cross-Talk in a SecondGeneration DDD Pacemaker
BARNEY B. BEAVER, JAMES D. MALONEY, LON W. CASTLE,VICTOR A. MORANT, JAMES M. KEEFE.* and ELIZABETH CHINCFrom the Department of Cardiology. Cleveland Clinic Foundation, Cleveland Ohio,and *Cordis Corporation. Miami. Florida
BEAVER, B.B., ET AL.: Design-dependent cross-talk in a second generation DDD pacemaker. A uniquemechanism of "cross-lalk" has recently been described in the Gemini 4J5A DDD pacemaker in whichcross-inhibition 0/ the ventricuJar output stimulus may occur in (he early Gemini 415A modeJs which weremanufactured with dual anodaJ rings interconnected hy a wire. The Gemini 415A units with dual anodaJrings were implanted under investigahonai protocol in 53 potients from September 1982 to August 1984.To date, cross-talk has been discovered in six patients during clinical/oJIow-up, moni/esting as inappropriateventricular inhibition. Syncope was documented in one patient. Non-Gordis atrial and ventricular leadswere present in all six patients and jive patients had at least one chronic lead. Cross-talk was neitherevident nor provokable in Gemini 415A units which had compatible Gordis lead systems. Reduction oftheatrial output and/or the ventricular sensitivity prevented cross-inhibition in three patients. Gross-talkpersisted in three patients despite programming to the lowest atrial output and/or ventricular sensitivityparameters. Gross-talk, spontaneous or provoked, has not occurred in 36 patients having currently designed4I5A pulse generators manufactured with a single anodal ring in the ventricular connector channel. Weconclude that: (I) this unique form of cross-talk is "design-dependent" and occurs only in Gemini 415Aswith dual interconnected anodal rings and cannot occur in 4'l5As manufactured with a single anodal ring:{2} this form of cross-talk has occurred only in 4I5As with non-Gordis leads and appears (0 be a time-dependent development; (3] reduction of the atrial output or use of pacing modes without ntrial pacing orventricular sensing can prevent cross-talk; (4j patients with Gemini 415As having dual anodal rings, whichcan be identified radiographically. are at risk for this design-dependent cross-talk and merit close obser-vation. fPAGE, Vol. 9, /anuary-February, 1986)
dual chamber pacemaker. DDD pacemakers, oversensing, cross-talk, cross-inhibition, bipolar DDD systems
Introduction
Electromagnetic interference, myopotentials.and intrinsic cardiac depolarizations are familiarcauses of oversensing which in turn, may be re-sponsible for inappropriate pacemaker inhibition.^The term "cross-talk" is used to describe a type ofpacemaker oversensing which is unique to dualchambered pacemakers; "cross-talk" is the occur-rence of pacing stimulus inhibition in one chamberin response to the inappropriate sensing of an out-put stimulus from the other chamber."
Address for reprints: James D. Maloney, M.D.. Cleveland ClinicFoundation. 9500 Euclid Avenue. Cleveland. OH 44106.
Received luly 1, 1985; revision received September 17. 1985;accepted September 18. 1985.
Conventionally, cross-talk manifests as cross-inhibition and less commonly as cross-stimulation.Cross-talk may originate from either biologic orphysical sources.""^ Biologic cross-inhibition canoccur when an intrinsic cardiac depolarizationfrom one chamber is inappropriately sensed in theother chamber causing inhibition of an outputstimulus. Physical cross-inhibition may arise whenthe afterpotential that follows an output stimulusin one chamber is sensed in the opposite chamberresulting in pacemaker inhibition. Alternatively,biologic cross-stimulation has been reported to oc-cur in the setting of anatomically narrow atrialelectrode-ventricular distances, or displacementof the atrial pacing electrode into the ventriclewhereby an atrial pacing stimulus causes anevoked response in both the atrium and the ven-
PACE. Vol. 9 January-February 1986. Part I 65
BEAVER, ET AL.
tricle.** Physical cross-stimulation can occur withthe development of abnormal current paths tomyocardial tissue as a result of insulation fracturesor fluid bridges tbat may develop between tbeatrial and ventricular connector channels.
We recently described the occurence of aunique form of cross-talk, termed "design-depen-dent cross-talk," which may occur in Gemini*415A DDD pacemakers initially designed andmanufactured witb dual anodal connector ringscoupled by a connecting wire/ The goals of thisreport are to review our experience with this formof cross-talk and to explain the design character-istics of the 415A which are responsible for its oc-currence.
Patients and Methods
From September 1982 through November1984, Gemini 415A DDD pacemakers were im-planted under investigational protocol in 57 malesand 32 females. The primary indications for pace-maker implantation included sick sinus syndromeand congenital or acquired heart block associatedwith symptomatic bradycardia. Ventriculoatrialconduction was assessed in all patients prior to,or at tbe time of, pacemaker implantation. Ourfollow-up protocol consisted of an initial period ofcontinuous bedside monitoring following pace-maker implantation and 24-hour Holter monitor-ing before hospital discharge. Pacemaker operationwas analyzed prior to hospital discbarge, then after6 weeks, 3 months. 6 months, and 1 year. Thisanalysis consisted of fluoroscopic examination ofthe pulse generator and leads, threshold deter-minations, and reprogramming to verify propertelemetry function. Cross-talk provocation wasattempted in all patients by progressively increas-ing the atrial pulse duration/current output andtbe ventricular sensitivity. When necessary, theprogrammed rate was increased and tbe AV delaywas shortened to initiate dual chamber stimula-tion, thereby facilitating cross-talk provocation.Additionally, each patient received a telephonetransmitter for pacemaker evaluation as neededbetween clinic visits.
Materials
Pulse Generator
* Cordis Corp., Miami. FL, USA.
The Gemini 415A fulfills ICHD criteria as aDDD. ME pacing device." The constant currentpulse generator incorporates a custom micropro-cessor powered by a lithium battery in which theprogrammable pulse duration is coupled with thecurrent output in a fixed manner. For example, apulse duration of 0.5 ms is coupled witb a currentoutput of 5 mA, etc. Single or dual cbamberedpacing operation is possible with independentlyprogrammable refractory periods, sensitivity, andpulse duration/current output in each channel.Lead polarity is programmable in channel 2 whenthe VVI pacing mode is operating.
Single and dual cbambered triggered modesprovide the ability to troublesboot sensing prob-lems and to document ventriculoatrial conduction.Additionally, programming ibe 415A to a DDTmode (PST,,/PSTvTJ^ permits the implanted pulsegenerator to be used for tachyarrhythmia induc-tion and termination.'"
The basic timing cycle for the415A operatingin a DDD mode is represented in Figure 1. Wbenoperating in this mode, tbe 415A automaticallyshifts from atrial demand to atrial synchronizedto dual chambered demand operation dependingon tbe patient's spontaneous atrial and ventricularrate and AV conduction velocity. Tbe maximumventricular paced rate when synchronized to in-trinsic atrial activity will be determined by theprogrammed atrial refractory period plus the AVdelay.
During the AV delay and the postventricularatrial refractory period (PVARF), the atrial channelis refractory to any sensing input. Since the pro-grammed atrial refractory period or PVARP beginsat the end of the AV delay (or tbe ventricular pacedor sensed event), the "total" ARP will be equal tothe AV delay plus the PVARP whereas sensedventricular events during the AV delay will ini-tiate tbe PVARP and provide a total ARP equal tothe P-R interval plus the PVARP.
Blanking on tbe ventricular channel (channel2) occurs following any atrial output pulse (chan-nel 1) and has a blanking duration of 39-47 ms.This mecbanism prevents tbe ventricular cbanneifrom detecting tbe atrial output pulse or its after-potential which could result in inappropriate
66 January-February 1986, Part 1 PACE, Vol. 9
DESIGN-DEPENDENT CROSS-TALK
415A Gemini Basic Timing Cycle
out
Channel 1
Channel 2
Total Atrial Refractory
Atrial Escape interval
Blanking .Ventricular,Refractory
Blanking
Refractory Alert
Figure 1. Gemini 415A basic timing cycie. Channel t—atrial channel; Channel 2—ventricularchannel; A out = alrial outpul; V oul = venlricular oulpul.
cross-inhibition of the ventricular output stimulus.During this blanking period, any input to thechannel 2 sense amplifier will not be recognized.
Lead System
Atrial and ventricular leads from severalmanufacturers were used (Table I). The stimula-tion thresholds and sensing characteristics ob-tained at the initial implantation are summarizedin Table II. Lead selection was dependent uponthe clinical situation and physician preference al-though active fixation leads were uniformly im-planted in patients with truncated atrial append-ages. Sleeve adaptors were utilized as recom-mended with non-Cordis lead systems to obtainan adequate seal between the connector pin andthe connector channel.
Background
The Model 415A pacemaker was initially de-signed with programmable unipolar or bipolar dual
cbambered pacing configurations. The pacer neckassembly (Fig. 2) incorporated anodal rings inchannel 1 and channel 2 for tbe use of bipolarleads with linear terminals. When a linear bipolarlead was inserted into the connector cbannei ofthe 415A. the anodal ring of the bipolar lead con-
Abial
Endocardial401 r327-752t6991 U"
Endocardial ActiveFixation
6957 J "6957-'
TABLE 1.
Acute Leads Utilized
No.
2931
125
Ventricular
Endocardial4010'327-2621327-263t
No.
22315
• CPI, St. Paul. MNt Cordis, Miami, FL•' Medtronic, Minneapolis, MN
PACE. Vol. 9 January-February 1986. Part I 67
BEAVER. ET AL.
LeadLocation
AtriumVentricie
TABLE l[.
Lead Electrical Characteristics (Acute)
Current Threshold(@ 0.5 ms Pulse
Width)
Mean Range
1.5 mA 0.5-3.7 mA1.3 mA 0.1-11 mA
Sensing Threshold
Mean
2.8 mV10.3 mV
Range
1.2-6.0 mV4.6-23 mV
• All measurements were obtained using a Medtronic 5311 PSA.
tacted the anodal ring in the connector channel.It is important to note that an interconnecting wirewas used to couple the channel 1 and channel 2anodal rings and when a bipolar pacing mode was
selected, the two anodal rings were held at thesame ground potential.
Evaluation of the bipolar DDD pacing modewas performed in dog models during pre-clinicaltesting.'' "Worst case" programming situationswere examined in these dog models in which thegreatest stimulation and sensing parameters wereselected for bipolar DDD pacing. Intermittentcross-inhibition of ventricular pacing was dem-onstrated during "worst case" testing which couldhe resolved with atrial pulse width/current outputreduction. Of particular interest was the recog-nition of cross-stimulation of the ventricle as a re-sult of an atrial output stimulus. Analysis of thiscross-stimulation occurrence confirmed that theanodal rings in channel 1 and channel 2 wereelectrically common as a result of the wire that
Pacer Neck Assembly and Leads
Key
Current PathwayFluid in Neck AssemblyNeck AssemblyLead Body and Terminal Pin
Figure 2. Pacer neck assembly and lead connector bodies within the connector channels. Anodolrings with interconnecting wire are highlighted in fhe lower right hand corner.
68 lanuary-February 1986. Part I PACE. Vol. 9
DESIGN-DEPENDENT CROSS-TALK
joined them. Moreover, since the anodal rings inthe connector channels were common, the anodalrings on the distal portion of each lead were alsocommon. Thus, any output pulse originating inone channel had its current divided between theanodal rings in each connector channel as well asthe anodal rings of each lead. Consequently, if theoutput pulse of one chamher was of sufficientmagnitude, the opposite chamber of the heartcould be stimulated via the anodal ring.
In order to prevent the clinical occurrence ofcross-stimulation, dual chambered bipolar pacingwas subsequently "locked out" of the software asa programmable option. Although the early in-vestigational models of the 415A continued to bemanufactured with an anodal ring in channel 1,it could not be activated through the programmer.The capability for single channel bipolar pacingthrough channel 2 could still be accessed. In sub-sequent clinical models of the 415A. the anodalring in channel 1 and the connecting wire wereomitted during the manufacturing process leavingonly the anodal ring in channel 2.
Results
Of the 89 415A pacemakers implanted at thisinstitution, 53 units were models manufacturedwith dual anodal rings and 36 units were modelsmanufactured with a single anodal ring. The an-odal ring configuration is easily identified radio-graphically as demonstrated in Figures 3A and B.During pacemaker follow-up, we identified six pa-tients with cross-talk that manifested as cross-in-hibition of the ventricular channel. Each of thesepatients had a 415A witb dual interconnected an-odal rings. Our initial case was identified followingan emergency room evaluation for unexplainedsyncope. The remaining patients were asymptom-atic and intermittent cross-talk was recognizedincidentally during scheduled pacemaker follow-up, which included provocative testing with highatrial output/ventricular sensitivity parameters.Although none of the patients were totally pace-maker-dependent, three had slow idioventricularescape rhythms.
In each case, the initial recognition of cross-talk was at least 6 weeks following pacemaker im-plantation. Common to all six patients was thepresence of non-Cordis atrial and/or ventricular
leads (Table III) and five patients had at least onechronic lead present at the initial pacemaker im-plantation. Acute stimulation thresholds in theatrium ranged from 0.7 to 3.7 mA in this group ofpatients and at tbe time of initial cross-talk rec-ognition, the programmed atrial pulse widthsranged from 0.5 to 0.8 ms coupled with currentoutputs of 5 to 8 mA, respectively. Programmedventricular sensitivity parameters were either 1.3or 2.5 mV and the most sensitive programmablevalue was 0.5 mV.
An ECG recorded in a representative patientduring a clinic evaluation is presented in Figure4A. A 415A was implanted in this patient forsymptomatic complete heart block associated witha slow idioventricular escape rhythm. At evalu-ation, telemetry interrogation revealed a DDDmode with an AV interval (AV) of 250 ms and anatriai pulse width (APW) of 0.8 ms coupled withan output of 8 mA. The ECC strip demonstratesfour atrial stimulus artifacts with atrial capture inat least three paced events. No ventricular pacingartifacts follow the atrial outputs at any AV inter-val. Both atrial and ventricular sensing appear tobe appropriate in this tracing.
This initial ECC led us to suspect that cross-inhibition was occurring from the atrial to theventricular channels in which the afterpotentialof an atrial ontput stimulus was being sensed bythe ventricular sense amplifier and thereby inhib-iting the ventricular output. To prove this, weprogrammed the pacemaker to a DDT (PSTg/PSTyTJ mode (Fig. 4B) in wbich an atrial pacingstimulus is triggered after any sensed event in theatrium and a ventricular pacing stimulus is trig-gered following any sensed event in the ventriclewhich occurs beyond the ventricular blanking pe-riod. While programmed to the DDT mode, eachatrial pacing artifact is followed in approximately50 ms by a triggered ventricular pacing stimuluswhich captures the ventricle. Because there is nonative ventricular activity to trigger an output inthe ventricle and the triggered ventricular pacingstimulus occurred just beyond the blanking period,we assumed that the ventricular channel wassensing an afterpotential from the atrial channelwhich triggered a ventricular output. To documentthis assumption with greater certainty, we pro-grammed the pacemaker to a VDD (S/PSIvT,) mode(Fig. 4C) in which atrial pacing is absent. This
PACE, Vol. 9 January-February 1986, Part I 69
BEAVER. ET AL.
Figure 3A. Gemini 4I5A ivilh dual anodaJ rings indicaled by Ihe arrows.
tracing demonstrates "P" waves which are sensedappropriately and following the programmed AVinterval a ventricular pacing stimulus emittedwith capture. In the absence of an atrial pacingstimulus, inhibition of ventricular pacing or cross-talk did not occur.
In practice, this type of cross-talk can usuallybe managed with a reduction in the atrial outputand/or the ventricular sensitivity. Although wewere able to prevent the occurrence of cross-talkin three patients with these programming maneu-vers, the remaining three patients continued tohave cross-talk despite programming of the atrialontput and ventricular sensitivity to their lowestparameters. This development prompted us to in-
vestigate alternative mechanisms which mightexplain the occurrence of this unusual form ofcross-talk.
Based on the cross-inhibition and cross-stim-ulation problems identified in the pre-clinicalstudies, it was known that cross-talk could occurbetween the atria! and ventricular channels, viathe interconnected anodal rings, when the 415Awas programmed to a DDD bipolar pacing mode.Although this mode was subsequently "lockedout" of the programmer, it was hypothesized thatin the presence of non-Cordis atrial and ventric-ular leads, body fluid could penetrate into eachconnector channel and provide a conductive pathwhereby an atrial output stimulus in channel 1
70 January-February 1986. Part I PACE, Vol. 9
DESrGN-DEPENDENT CROSS-TALK
Figure 3B. Gemini 415A wHh a single anodal ring in channel 2 indicated with an arrow.
PT
1234
56
TABLE III.
Incidence of Cross-Talk
Atrial Lead
AcutetChronic"Acute*Acute'Chronic**Acute*
Ventricular Lead
AcutetChronic"*ChronictChronicttChronic'*Chronict
Onset
6 wks6 wks6 wks6 wks6 wks6 wks
*CPit Medtronic" Intermedjcst t CordisAdaptors utilized with non-Cordis toads.
could be conducted to the channel 2 sense ampli-fier via the wire joining the anodal rings. Thisconcept is illustrated in Figure 4.
In an attempt to simulate this clinically ob-served cross-talk, a bench test model was con-structed with a 415A using non-Cordis atrial andventricular leads [Fig. 5). Materials for testing in-cluded a 415A breadboard, hybrids, oscilloscope,saline tank, programmer, and smooth tipped pac-ing electrodes. The suspicion that the connectingwire between the channel 1 and channel 2 anodalrings might be responsible for the observed cross-talk led to the insertion of a simulated "leakagepathway" between cbannels 1 and 2 using a vari-able resistor (12 k-28 kilohms).
PACE, Vol. 9 January-Eebruary 1986, Part I 71
BEAVER, ET AI..
GEMINI 415A
CC6
200mi»c
I I I I
DDDPSIalvPSIvTa 1 V I 1V 1
VV=750msecAV=250msecAPW=O.8msec
550
ECG
DDT PSTa
PSTvTa
VV-750msecAV="250msecAPW-0.8mtec
\i \i \i
300 300
ECG
VDD K KPSIvT*
VV-750fn»ecAV"250niiecAPW="OFF" 3pp]
400j ftsa
300
400 1
300|
Figures 4A, B, C. Baseiine ECG and programming ECGs from a patient with design-dependentcross-talk in a 415A. Symbolic language for pacemaker/heart interaction modified from Brownlee.R.R.. e( al^ A = atrial pacing stimulus.
During continuous oscilloscope monitoring,the bench test model was initially evaluated in aunipolar DDD mode without the leakage pathway(i.e.. variable resistor removed). As demonstratedin the oscilloscope tracing in Figure 6A, the atrial
output stimulus (channel 1) is followed by a ven-tricular output stimulus {channel 2) after the pro-grammed AV interval of 200 ms. No afterpotentialfrom the channel 1 output stimulus is evident onthe channel 2 waveform following the blanking
72 January-February 1986, Part I PACE. Vol. 9
DESIGN-DEPENDENT CROSS-TALK
Cross-Talk Test Configuration
Figure 5. Cross-talk lest con/igurafion.
period. Alternatively, when the channel 1 andchannel 2 anodal rings are connected via the vari-ahle resistor, simulating the leakage pathway, thepolarization waveform on channel 2 changes dra-matically (Fig. 6B). Following the channel 1 de-polarization a small negative polarization occurson the channel 2 waveform following the blankingperiod. This is followed by a return of the wave-form to baseline and absence of ventricular outputstimulus artifact which indicates that the channel2 sense amplifier has sensed the afterpotential fromthe channel 1 depolarization resulting in ventric-ular inhibition. Additionally, the timing cycles arereset at the end of the blanking period indicatingthat the channel 2 sense amplifier has been trig-gered (Fig. 7). Thus it is possible that an afterpo-tential from the channel 1 output stimulus canpersist beyond the blanking period and develop
appropriate frequency and amplitude character-istics to be recognized by the channel 2 sense am-plifier thereby inhibiting a ventricular outputstimulus.
This form of cross-talk can occur with the415A in vivo in an analogous fashion. Using non-Cordis atrial and ventricular leads, it is possiblethat body fluids could more readily penetrate eachconnector channel aperture. Acting as a conductor,the fluid can complete the potential short circuitbetween channel 1 and channel 2 via the wirejoining the anodal rings. An atrial output stimulusand the afterpotential that it develops through thefluid in the connector channel can be conductedto the channel 1 anodal ring, across the connectingwire to the channel 2 anodal ring, and then to thechannel 2 sense amplifier resulting in ventricularinhibition.
PACE. Vol. 9 lanuary-February 1986, Part I 73
Ch. 1
Ch.2
2V 1DiV
2VDIV
. ^ ——1
< ^
Time
BEA
' IC
VEK
)0m£
ET
j e c
AL.
Programmed ValuesRate 70 {857 ms)
_f AV Delav 200Mode DDD
Ch, 1 Ch. 21 Out 1/10 1/10
Sense 0.5 2.5- REF 200 200
DIV
Voltage across Ch. I/case and Ch. 2/case vsTime. Circuit without R |_eakaae
Ch. 1 2VDIV
Ch. 2 200 mVDIV
TimeA100 msec
DIV
++H
Programmed ValuesRate 70 (857 ms)AV Delay 200Mode DDD
Ch.1 Ch.2Out 1/10 1/10Sense 0.5 2.5REF 200 200
Voltage across Ch. I/case and Ch. 2/case vsTime. Circuit without
Figure 6A. Oscilloscope tracings/rom cross-laik lesf model. Channel 2 is unaffected by channel1 oulpul waveform. Channel 1 output is altenualed ivhen no leakage pathway is present. Channel2 output occurs at end of AV delay (200 ms).
Discussion
At the time of this writing, only 11 cases ofcross-talk have been reported in 1.241 patientswith Gemini 415A pacemakers.^^ The unique formof cross-talk that we first described in 1983^ ap-pears to be design-dependent, occurring only in415As with dual anodal rings; it cannot occur in
415As manufactured with a single anodal ring.Importantly, this design-dependent cross-talk doesnot occur uniformly in all patients with 415Ashaving dual anodal rings. This can be supportedby several observations. First, we have only iden-tified cross-talk in 415As in which non-Cordisleads have been used and which are most likelyto allow body fluid to penetrate into the connector
74 January-February 1986, Part I PACE, Vol. 9
DESIGN-DEPENDENT CROSS-TALK
Ch. 1 2VDIV.
Ch. 2 2VDtV.
4-K+ +++• +4++ H++
Programmed ValuesRate 70 (857 ms)AV Delay 200Mode DDD
Ch.1 Ch.2Out 1/10 1/10Sense 0.5 2.5REF 200 200
Time 100 msecDIV
Voltage across Ch. I/case and Ch. 2/case vsTime. Circuit with R
Ch. 1 2VDIV
Ch. 2 200 mVDIV
Programmed ValuesRate 70 (857 ms)AV Delay 200Mode DDD
Ch. 1 Ch. 2Out 1/10 1/10Sense 0.5 2.5REF 200 200
Time 100 msecDIV
Voltage across Ch. I/case and Ch. 2/case vsTime. Circuit with R
Figure 6B. Oscilloscope (racings from cross-laik test modei. Channel 2 senses channel 1 after-potential when channel 2's blanking ends. The aflerpofenh'al is transmitted to channel 2 via fheconneclor channel j luid and interconnected anodal rings. A sensed even! in channel 2 resets pacertiming. The next atrial paced event occurs at the end of the atriai escape interval.
channel. We have not recognized spontaneous orprovokable design-dependent cross-talk in 415Ashaving Cordis compatible leads. Additionally, therecognition of cross-talk during our follow-up pe-riod was time dependent and probably a functionof the time necessary for fluid to accumulate inthe connector channels. Second, the ability to
simulate cross-talk in the bench model dependedon the magnitude of the pulse duration/currentoutput and sensitivity parameters as well as theunique resistance value developed across theleakage pathway. In vivo this resistance value isdetermined by the fluid in the connector channelsand the resistance across the wire joining the an-
PACE. Vol. 9 lanuary-February 1986, Part I 75
BEAVER, ET AL.
415A Cross Sensing Timing Cycle
out out
Channel 1
Channel 2
-*-Total Atrial Refractory-*
DESIGN-DEPENDENT CROSS-TALK
References
7.
Irnich, VV.: Interference in pacemakers. PACE, 7:1021, 1984.Furman, S., Reicher-Reiss. H.. Escher, D.J.W.:Alrioventricular sequential pacing and pacemak-ers. Chest. 63:783, 1973.Levine. P.A.. Mace. R.C.: Pacing Therapy: A Guideto Gardiac Pacing for Optimum HemodynamicBenefit. Mount Kisco, Futura PubHshing. 1983, p.239.Galfee. R.V.: Dual-chamber committed mode pac-ing. PACE, 6:387. 1983.Tarjan, P.P.: Principles of multimode pacing. PACE.6:357, 1983,Investigators Manual: Gemini 415 A. Gordis Cor-poration, Miami. Florida. 1982,Beaver, B.B.. Ching. E., Castle. L.W., et al.: Cordis
10.
11.
Gemini DDD pacemaker: Initial investigationalexperience with a second generation DDD pace-maker. Circulation. 68:111-378. 1983 (Abstract).Parsonnet. V.. Furman. S.. Smyth. N.P.D.: A revisedcode for pacemaker identification. PACK. 4:400.1981.Brownlee. R.R.. Shimmel-Golden. I.B.. Delmarco,Cl., et al.: A new symbolic language for diagram-ming pacemaker/heart interaction. PACE. 5:700,1982,Fisher, I.D., Furman. S.. Kim, S.G.. et al.: DDD/DDT pacemakers in the treatment of ventriculartachycardia. PACE. 7:173, 1984.Potential crosstalk in early Gemini 415A pacerswith dual anodal rings: Product Safety Alert, Oc-toher 19,1984. Cordis Corporation, Miami. Florida.
Organising Committee:Professor A. J. CammProfessor D, G. JulianDr. A, W. rOathanProfessor J. P. ShillingfordBrigadier M, C. Thursby Pelham
BRFTISH HEART FOUNDATION102 Gloucester Ptace. London W1H 4DH Tel: 01-935 0185
THE INTERNATIONAL CONFERENCEON THE
MANAGEMENT OFCARDIAC ARRHYTHMIAS
September 2-4, 1987London, England
Contact: Professor A.J. Camm
PACE, Vol. 9 January-February 1986, Part I 77
