(Circulation. 1996;93:91-98.)
© 1996 American Heart Association, Inc.
Articles |
Undetected Ventricular Fibrillation in Transvenous Implantable Cardioverter-Defibrillators
Prospective Comparison of Different Lead System–Device Combinations
From the VA Medical Center/Duke University, Durham, NC; the Sinai Samaritan Medical Center/St Luke's Hospital, University of Wisconsin, Milwaukee (Wis) Clinical Campus; and Methodist Hospital, Houston, Tex.
Correspondence to Andrea Natale, MD, VA Medical Center/Duke University, 508 Fulton St, Box 111A, Durham, NC 27705.
 |
Abstract |
|---|
 Top Abstract IntroductionMethods Results Discussion Addendum References |
|---|
Background The purpose of this study was to prospectively analyze redetection problems after unsuccessful shock with different lead systems and devices.
Methods and Results We prospectively analyzed
detection and redetection characteristics among transvenous implantable
cardioverter-defibrillators (ICDs) using standard bipolar and
integrated bipolar sensing. Monophasic and biphasic ICDs were included.
Subthreshold shocks were intentionally delivered, and redetection of
ventricular fibrillation (VF) was assessed before discharge
and at 1, 3, 6, and 12 months later. Sensing of VF resulting from
antitachycardia pacing and low-energy cardioversion (
2 J)
also was analyzed. Before inclusion in the study, each patient
underwent subthreshold shock testing at three different time intervals.
Among the 160 ICDs with standard bipolar sensing, 530 VF inductions
were analyzed. After the failed shocks, undersensing was more
frequent (3% versus 20%, P<.01) but did not remarkably
prolong redetection (3.1±0.8 versus 3.3±1.1 seconds). Among the
201
ICDs with integrated bipolar sensing, 80 were connected to a CPI device
(60 Ventak 1600–Endotak 60 series and 20 PRx II 1715–Endotak 70
series) and 121 to the Ventritex defibrillator (91 Endotak 60 series,
14 TVL systems, and 16 Endotak 70 series). After 252 failed shocks,
redetection was prolonged with the CPI system (3.1±1.4 versus
4.6±3.6
seconds, P<.05) but did not change after 396 failed shocks
with the Ventritex ICD (5.4±1.9 versus 4.9±2.2 seconds). This
may
reflect different nominal settings for detection and redetection. In 9
of 121 patients with Ventritex and 1 of 80 with the CPI ICDs, the
devices failed to redetect VF. However, redetection malfunction was
never observed in patients with integrated bipolar systems with >6-mm
electrode separation. After antitachycardia pacing in 1 patient
and a 2-J shock in 1 patient, ventricular
tachycardia turned into VF, which was undetected. Both patients
used the Endotak 60 series–Cadence combination. None of the patients
showing VF undersensing had sudden death at follow-up. Only 3 of
the 12 patients with sensing malfunction were on
antiarrhythmia drugs at the time of testing.
Analysis of endocardial electrograms showed that failure to
redetect VF is not associated with a uniform reduction but with a rapid
and repetitive change of electrogram amplitude.
Conclusions Standard bipolar sensing redetects VF more effectively than integrated bipolar sensing. Endocardial electrogram analysis provides insights into the understanding of the mechanism of undersensing, and certain lead-device combinations result in a higher occurrence of VF undersensing. The clinical relevance of this phenomenon remains unknown.
Key Words: defibrillation • fibrillation • death, sudden
|
Introduction |
|---|
|
TopAbstractIntroductionMethods Results Discussion Addendum References |
|---|
The primary purpose of implantable cardioverter-defibrillators (ICDs) is to appropriately detect and treat ventricular fibrillation (VF) and ventricular tachycardia. Prompt and accurate recognition of these arrhythmias depends on the quality and stability of the endocardial electrograms recorded through the transvenous lead used for implantation. It has been suggested that shocks might alter the amplitude of the intracardiac electrograms and delay or prevent redetection and treatment of ventricular arrhythmia attacks.1 2 3 4 5 6 7 8 9 However, the potential for sensing problems might be increased by the use of a particular lead system or sensing algorithm. The purpose of this study was to prospectively analyze redetection problems after unsuccessful shocks with different lead systems and devices.
|
Methods |
|---|
|
TopAbstractIntroductionMethods Results Discussion Addendum References |
|---|
Patient Population
Three hundred sixty-one patients who received transvenous ICDs were included in the study. This population comprised 60 patients who previously received a CPI 1600 system with the Endotak C lead (60 series) and 90 patients who received the Transvene system with the PCD 7217B device. The remaining 211 consecutive patients underwent prospective evaluation. Fifty-eight patients were implanted with the Transvene system associated with the Medtronic device model 7219D. Ninety-one patients underwent implantation with the Cadence device V100 and the Endotak lead 60 series. Twelve patients received the Medtronic device model 7219C with the Transvene lead, 20 received the Endotak lead 70 series and the CPI device PRx II, 14 received a Cadence device V100/V110 and the TVL lead system, and 16 underwent implantation with the Cadence device V110 and the Endotak 70 series lead system. There were 315 men and 46 women; their mean age was 62.5±9.8 years (range, 36 to 82 years). The left ventricular ejection fraction ranged from 15% to 65% (mean, 29.6±11%). Two hundred eight-five patients (79%) had coronary artery disease, 58 (16%) had idiophatic dilated cardiomyopathy, and the remaining 18 (5%) had no structural heart disease or congenital long QT syndrome. In each patient and with each lead system, subthreshold shock and redetection were evaluated at implantation; before discharge; and 1, 3, 6, and 12 months later in follow-up testing. Analysis was performed in only those patients who had at least three different follow-up ICD testings during which redetection after subthreshold shock was performed.
Lead Systems
In 160 patients, the Transvene system
(Medtronic, Inc),
consisting of a 10.5F right ventricular catheter model 6966
and a 7F superior vena cava coil model 6963, was used. The right
ventricular catheter used a true bipolar sensing system
from the distal helix to the proximal ring electrode with an
interelectrode distance of 10 mm. With the Transvene system, the two
5-cm coils in the right ventricle (surface area, 426 mm2)
and superior vena cava (surface area, 90 mm2) are used only
for energy delivery. A total of 14 patients underwent implantation and
testing with the Cadence TVL system (Ventritex, Inc), which includes a
right ventricular lead model RV-1101 (surface area, 470
mm2) and a superior vena cava lead model SV-1101 (surface
area, 550 mm2). The right ventricular lead used
passive fixation with tines and provided rate sensing between the 5-cm
right ventricular defibrillation coil and the distal pacing
tip. The interelectrode distance was 11 mm. In the remaining patients,
the Endotak C 60 series (151 patients) and the Endotak C 70 series (36
patients; CPI, Inc) were used. Both consisted of a tripolar passive
fixation lead that used integrated bipolar sensing between the distal
electrode tip and the distal right ventricular
defibrillation coil. The distal end of the coil is separated from the
electrode tip by 6 mm in the Endotak 60 series and by 12 mm in the 70
series. In the 60 series, the defibrillation system consisted of a
3.8-cm distal coil (surface area, 295 mm2) and a
6.8-cm-long proximal coil (surface area, 617 mm2). The
70 series lead had a 4.7-cm distal coil (surface area, 379
mm2) and a 6.8-cm proximal coil (surface area, 617
mm2).
Devices for Testing and Permanent Implantation
Several
different devices were used, including the PCD 7217B
transvenous (Medtronic), the CPI monophasic device model number 1600,
and three biphasic devices: the PCD Jewel model 7219D/C (Medtronic),
the Cadence (model V100, V110 Ventritex), and the CPI model PRx II
(CPI). With each device, data were collected at implantation; before
discharge; and at 1-, 3-, 6-, and 12-month follow-up testing.
During each testing, both detection and redetection after subthreshold
shocks were analyzed to identify any sensing errors. Before
analysis of the data and inclusion of the patient in the study,
at least three redetection trials were performed at three different
times. Although separate detection and redetection algorithms were
configured as appropriate for individual patients, for the purpose of
this study and when allowed by the defibrillator, both detection and
redetection were programmed by use of the same number of intervals. The
detection algorithm used by each device is mentioned briefly.
Medtronic PCD Models 7219D/C and 7217B
With both the
monophasic and biphasic devices, detection and
redetection testing were performed at a sensitivity setting of 0.3 mV.
With both devices, the number of detection and redetection intervals is
programmable, and detection and redetection criteria are met when 75%
of the programmed number of intervals go below the fibrillation
detection interval. For the purpose of this study, 16 intervals were
used, so 12 intervals had to be sensed before detection or redetection
criteria were met.
CPI Model 1600
With this device, both detection and
redetection are the same
and are not programmable. During testing and after the sensing function
is reactivated, the device "counter increases" by 1
every time it senses an electrogram at a shorter cycle length than the
cutoff rate for VF detection. When the counter reaches a value of 8
intervals, the detection criteria are met, and a delay interval
programmable for the first shock and nonprogrammable (equal to 2.5
seconds) for the second and following shocks begins. After an
unsuccessful shock, the same criteria are used to redetect VF. In the
evaluation of the data from this device, the "delay time" was not
included in the detection and redetection analysis.
Ventritex Cadence (V100, V110)
The Cadence devices can be
programmed with up to three
tachyarrhythmia zones. VF detection required that a
minimum number of intervals (nominal, 12) must be classified as less
than the programmed VF detection interval and that the average of the
last interval with the previous 3 intervals must be less than the VF
detection rate.
When both the last and average intervals are in the VF zone, the detection counter is increased by 1. When the last and average intervals are in different tachycardia zones, the last interval is classified in the zone corresponding to the shorter cycle length. If only the last or the average interval is longer than the slowest tachyarrhythmia rate, the last interval is discarded. However, when both the last and the average intervals are longer than the slowest tachyarrhythmia rate, the last interval is classified as a "sinus beat," and the sinus redetection counter is increased by 1. The sinus rhythm counter is reset to zero whenever an interval is classified as VF or ventricular tachycardia. When the sinus counter reaches a programmable number of intervals (nominal, 5; slow, 7; or fast, 3), all counters are reset to zero. Redetection follows the same criteria, but only 6 events need to be counted for redetection. In the Cadence devices, a separate postshock VF detection interval can be programmed. In our study, the number of intervals for the detection of VF always was programmed to nominal, and sinus redetection was nominal (5 intervals) in 20 patients and slow (7 intervals) in the remaining 101 patients.
CPI PRx II Model 1715
In the PRx II device, each zone has a
detection window that
consists of the 10 most recent RR intervals measured by the pulse
generator. When 8 of 10 intervals in a detection window have been
classified as above the rate threshold (fast), the window is considered
satisfied. The detection window then will remain satisfied as long as 6
of 10 intervals remain classified as fast. After the detection criteria
are satisfied, a "duration time" will start that can be
programmed in the VF zone from 1 to 15 seconds. For the purpose of this
study, testing was performed with a 1-second duration time. After an
unsuccessful shock, redetection uses the same detection window process
and programmed tachycardia rate threshold as the initial
detection to confirm or deny the presence of a
tachyarrhythmia (8 of 10 fast intervals). After
redetection is met, a postshock duration time begins. For redetection,
this time interval is not programmable and is fixed at 1 second. The
duration time was not counted in the analysis of detection and
redetection times.
Data Collection
Subthreshold shock testing with the
measurement of detection and
redetection times was performed intraoperatively; at predischarge; and
at 1-, 3-, 6-, and 12-month follow-up testing. Patients were
included in the study and analysis was performed only after
each patient underwent subthreshold shock testing during three
different defibrillator checks. As mentioned, VF was induced, depending
on the device, with 60-Hz AC, rapid burst pacing, or delivery of a
T-wave shock. Defibrillation threshold was determined by use of
2.5-J increments or decrements starting from an energy setting 5 J
higher than the last determined defibrillation threshold.
Defibrillation threshold was considered the lowest energy of the first
shock that successfully ended VF. Once the defibrillation threshold was
obtained, the first shock was reprogrammed with an energy 5 J lower
than the defibrillation threshold, and the second shock was the highest
energy that the device could deliver. At least 3 minutes was allowed
between consecutive inductions of VF. In addition, any
hemodynamic or ECG changes had to recover before
induction of VF. Three surface ECG leads were recorded during
testing. When available, intracardiac electrograms during the VF
episode were retrieved and analyzed. When allowed, the number
of detection and redetection intervals was kept constant within and
among devices. With all Medtronic devices, the number of intervals
chosen for both detection and redetection was 12 of 16. Similarly, the
postdetection delay time (Ventak P 1600) and duration time (PRx II)
were programmed by use of the default setting for redetection, namely
2.5 seconds for delay time and 1 second for duration time. With the
Ventritex and Medtronic devices, detection and redetection times were
defined by subtracting the postinduction and postshock refractory
periods and the capacitor charge time provided on interrogation after
each therapy sequence from the total event time. The availability in
the Medtronic devices of event markers facilitated the identification
of detection and redetection times. Although the programmer provided
detection and redetection times for the CPI devices, these times were
confirmed on paper by subtracting the capacitor charge time and the
delay interval from the total time. If rapid burst pacing or 60-Hz AC
was used, induction time was not included in the analysis.
Analysis of VF episodes induced by unsuccessful
antitachycardia pacing or low-energy cardioversion (
2 J) also was performed. End points were (1) to analyze the
occurrence of inappropriate redetection of sinus rhythm during VF by
use of different lead-device combinations, (2) to analyze
the occurrence of prolonged redetection (>2 SD above the mean
redetection time)3 with different lead-device
combinations, and (3) to document whether VF undersensing resulting
from unsuccessful antitachycardia therapy or low-energy
cardioversion ( 2 J) may occur.
Statistical Analysis
The mean detection and redetection times
from at least three
different subthreshold shock testings were obtained for each patient.
Data are expressed as mean±SD. Comparison between the mean detection
and redetection times within each device-lead combination was
performed with the paired Student's t test. The unpaired
Student's t test was used to compare detection and
redetection times among different devices, and 
2
analysis was performed to compare the number of redetection
problems observed with each lead-device combination. In addition, a
Cox proportional-hazards model was used to determine which factors
were independent predictors of redetection problems. Least-squares
linear regression analysis also was performed to detect any
relation between the subthreshold shock energy and redetection time. A
value of P<.05 was considered statistically
significant.
|
Results |
|---|
|
TopAbstractIntroductionMethods Results Discussion Addendum References |
|---|
There were 1178 episodes of VF that required a second shock for termination. Of these, 530 were with true bipolar sensing and 648 were with integrated bipolar sensing. The mean energy of the unsuccessful shocks was 11.6±7.2 J and did not differ between sensing systems. The mean amplitude of the endocardial electrogram at implant was 10.6±2.9 mV with the integrated systems and 11.4±3.2 mV with the true bipolar electrodes. The mean detection times with the true bipolar and integrated bipolar systems were 3.1±0.8 and 5.0±1.6 seconds, respectively. The mean redetection times were 3.3±1.1 and 4.7±2.1 seconds with the true bipolar and integrated bipolar sensing systems. The energy of the failed shocks did not correlate with the redetection time or the occurrence of VF undersensing (r2=.006, P=NS).
VF Undersensing
Undetection of the initial VF episode was not
observed. Failure to
redetect VF after an unsuccessful first shock was seen in 1 patient
with the CPI 1600–Endotak C 60 series and in 9 patients with the
Cadence–Endotak 60 series combinations. External
transthoracic shocks were given to restore sinus rhythm 24
to 33 seconds after the subthreshold shocks (mean, 28±4 seconds)
because no evidence of appropriate redetection was documented.
Redetection malfunction was observed at the predischarge testing in 2
patients and at one of the follow-up testings in the others. Many
attempts to reproduce the redetection malfunction were made, ranging
from 3 to 16 in the same patient (mean, 5.3±1.2). However, failure to
redetect VF was replicated in only 1 patient during other VF episodes.
Among the 9 patients with the Cadence–Endotak 60 series combination,
sinus rhythm redetection at the time of VF undersensing was programmed
to nominal in 1 and to slow in the remaining 8. In addition, in the
patient programmed to nominal, the inappropriate sensing could not be
reproduced even though the sinus redetection setting was left
unchanged. In 9 of 10 patients with failed redetection, an external
transthoracic shock was given to terminate VF. In 1 patient
(Fig 1
), after the device redetected sinus rhythm during
VF, the quality of the electrogram improved. This allowed the device to
sense VF and again deliver the first shock therapy, which, on this
occasion, terminated the arrhythmia. Only 3 of the 10 patients
with undetected VF were on antiarrhythmia drug therapy.
Analysis of the stored electrograms showed that VF undersensing
was associated mostly with rapid and repetitive changes in electrogram
amplitude. Two different signal sequences were observed: one large
electrogram alternating with one small electrogram (Fig 2) and
a series of small-amplitude electrograms
alternating with larger signals (Fig 3). As Fig
1 shows,
the appropriate sensing was not restored unless the above sequence
broke. The cycle length of the large-amplitude signals during
undetected VF episodes ranged from 800 to 395 ms. There were no
episodes of VF undersensing with the true bipolar sensing systems and
the PRx II–Endotak 70 series, Cadence–Endotak 70 series, and
Cadence-TVL combinations.
|
|
) signal. The rapid amplitude variation
prevented adjustment of the gain amplifier to detect the small signals.
Because the rate of the large-amplitude signals followed beyond any
tachycardia detection interval, inappropriate detection of
sinus rhythm occurred (sinus redetection).
|
Prolonged Redetection
Comparison of detection and redetection
times in each
lead-device combination showed that redetection time was not
prolonged in any patient with a true bipolar system (detection,
3.1±0.8 seconds; redetection, 3.3±1.1 seconds;
P=NS).
After unsuccessful shocks, however, the occurrence of single beat
undersensing increased significantly (2% versus 20%,
P<.05). In patients with the PRx II–Endotak 70 series
(3.0±0.7 versus 3.1±0.6 seconds, P=NS), the
Cadence–Endotak 70 series (5.2±0.9 versus 4.1±0.8 seconds,
P<.05), and the Cadence-TVL combinations (5.1±0.8 versus
3.9±0.7 seconds, P<.01), redetection was not prolonged,
and no case of long redetection was observed. Among the 60 patients
with the Ventak 1600–Endotak 60 series, failed shock resulted in
longer redetection times (3.1±1.4 versus 4.6±3.6 seconds,
P<.05). Abnormal prolongation of the redetection time (mean
redetection +2 SD) was seen in 9 of 60 patients. In patients with the
Cadence–Endotak 60 series combination, redetection was not prolonged
(5.4±1.9 versus 4.9±2.2 seconds, P=NS). However,
abnormal
prolongation of the redetection time was observed in 14 of 91
patients.
VF Sensing After Failed Antitachycardia
Therapy
Among devices with antitachycardia pacing, VF sensing
resulting from unsuccessful pacing or low-energy cardioversion (2
J) during ventricular tachycardia was
analyzed.
Sensing failure was not observed among the 41 episodes with true bipolar systems (Transvene lead), 14 with the PRx II–Endotak 70 series, 21 with the Cadence–Endotak 70 series, and 5 with the Cadence-TVL combinations.
However, in 1 of the 32 episodes with the
Cadence–Endotak 60
series, combination VF was not detected after ineffective
antitachycardia pacing (Fig 4). Moreover, in
1 patient (1 of 26 episodes) in the same group, a 2-J shock delivered
during ventricular tachycardia caused
degeneration to VF that was not detected and required a
transthoracic shock for termination. This phenomenon was
not observed during 22 similar episodes in patients receiving a true
bipolar system or 16 episodes in patients using integrated systems with
a larger electrode separation than the Endotak 60 series.
|
Failure to redetect VF and prolonged redetection time appeared to be unique to the Endotak 60 series lead (22% versus 0%, P<.01). In addition, VF undersensing seemed more frequent with the Cadence–Endotak 60 series combination (26% versus 3.3%, P<.05). This may reflect the difference in time that the sense amplifier required to adjust the gain setting or incorporation into the detection and redetection algorithms of the average of the last interval with the preceding three intervals. Among patients with normal and abnormal sensing, there was no difference in mean ejection fraction, defibrillation threshold, age, and mean R-wave amplitude in sinus rhythm.
During a mean follow-up of 12.4±3.1 months, all but 1 of the patients showing sensing problems received defibrillator therapy. In 1, prolonged detection preceding the defibrillation shock provoked syncope. However, a second shock was never required. On the other hand, low-energy shocks were not programmed in these patients, and antitachycardia therapy was not used unless consistently effective. None of the patients included in this study died suddenly.
|
Discussion |
|---|
|
TopAbstractIntroductionMethods Results Discussion Addendum References |
|---|
ICDs have assumed an important role in the management of patients with life-threatening ventricular arrhythmias. Despite several technical improvements that have facilitated implantation of ICDs and enhanced the capability of this device to detect and terminate arrhythmias, fundamental limitations of the tachycardia detection schemes incorporated into ICDs and sensing lead design may continue to affect the reliable function of this technology. Efficient and appropriate function of ICDs depends on recognition of the electrogram signals from the implanted cardiac electrodes. It is clear that the primary challenge for sensing is to avoid missing too many of the low-amplitude fragmented R waves that might occur during VF. At the same time, high sensitivity must not cause oversensing of external source of interference or other components of the intracardiac signals. It is well known that the electrogram amplitude and shape can be affected by the direction of the local propagation wave front with respect to electrode geometry and design. This issue may become even more important after an unsuccessful shock because other factors such as ischemia, metabolic changes, and the direct effect of the unsuccessful shock increase the likelihood of signal distortion. Considerable variety exists in the sensing threshold settings and algorithm strategies used to sense individual electrograms during VF. In addition, lead designs for sensing and pacing also differ among devices. Although some manufacturers have opted for true bipolar electrodes, others have used integrated bipolar systems.
Unreliable redetection of VF has been reported recently with the use of an integrated bipolar sensing system.1 2 3 4 5 6 7 8 9 This was attributed to a reduction in R-wave amplitude immediately after defibrillation shock delivery.1 4 5 6 9 However, prospective evaluation of this lead system with serial and multiple subthreshold shock testing has not been performed in a large patient population. In addition, whether this phenomenon is limited to the integrated bipolar system has not been assessed. The findings of our study appear to suggest that (1) a true bipolar system performed better than an integrated bipolar system in the redetection of VF; (2) certain detection algorithms might increase the likelihood of redetection problems with the integrated bipolar system; (3) VF undersensing is not associated with a uniform reduction but with rapid and repetitive changes of electrogram amplitudes; and (4) although no detection of VF is more likely after unsuccessful shocks, deterioration of electrogram signals leading to failure to detect VF was occasionally observed even after unsuccessful antitachycardia therapy.
Because degeneration of electrogram amplitude and quality has been associated with a polarizing effect of the shock at the electrode-tissue interface4 9 or with electroporation,4 9 10 11 12 which creates microscopic cardiac cell membrane injuries, it is not surprising that redetection undersensing is unique to the integrated bipolar system, which uses the same electrodes to sense and deliver energy. It is possible that a higher voltage gradient is seen by the myocardial tissue confined to the immediate region of the integrated system compared with the true bipolar sensing system. In this regard, preliminary results from Penzotti and coauthors,13 who used a finite-element analysis of postshock sensing performance with different transvenous lead systems, showed that the area surrounding the sensing cathode of the integrated bipolar system is subjected to a potential gradient that is double the values predicted at the distal true bipolar electrode.
Although it is true that failed shocks favor inappropriate redetection of VF, we have seen examples of undetected VF after unsuccessful antitachycardia therapy. This might suggest that other components of the electrodes may be responsible for some of the sensing dysfunction observed. The distance between the distal tip and the defibrillation coil might also play an important role in appropriate sensing of electrogram signals. In fact, lead systems using more than 6-mm interelectrode spacing appear to perform appropriately.
Although it is well known that high-energy shocks uniformly reduce the electrogram amplitude recorded with all the integrated bipolar systems, redetection problems are unique to the Endotak 60 series and are not reproducible despite several attempts and the lack of changes in the defibrillator parameters. This argues against a uniform reduction of the electrogram amplitude as the critical factor responsible for redetection dysfunction. Our data, consistent with previous observations in single patients,1 2 3 appear to suggest that rapid and significant variations in electrogram amplitude and not a uniform reduction predispose to inappropriate redetection of VF. It is unclear whether electrogram quality alteration is exclusive to the myocardium in the vicinity of the sensing system or whether subthreshold shocks modify the characteristic of the electrograms throughout the entire myocardium. The fact that a similar phenomenon was never observed in devices using a true bipolar system lead us to suppose that electrogram deterioration is confined to the immediate region of the high-energy shock field. On the other hand, the occurrence of similar sensing problems after antitachycardia pacing and low-energy cardioversion shocks, which are unlikely to achieve high potential gradient, implies that electrode separation may be more important than lead configuration (true bipolar versus integrated bipolar).
It has been suggested that drug therapy with antiarrhythmic medication might play a role in favoring degeneration of the electrogram quality.2 In our patient population, however, only 3 of the 10 patients who had VF undersensing were treated with antiarrhythmia drugs at the time of the abnormal findings. It is notable that the sensing performance seemed to be affected not only by integrated bipolar systems with shorter interelectrode distance but also by the use of certain devices. It is possible that the detection algorithm and the time constant of the increase in gain or decrease in threshold used by a specific device may also affect the ability to effectively detect rapid changes in electrogram signals. To correct this problem, several expedients have been proposed, including prolongation of the sinus redetection parameters and the postshock fibrillation detection interval and maximizing the energy of the first shock therapy.1 2 3 As our experience showed, however, prolongation of the sinus redetection parameters does not affect appropriate defibrillator function and previously has been falsely interpreted to be effective in improving redetection problems. On the other hand, the phenomenon described is not reproducible, which makes retesting of parameter changes difficult to assess and interpret. In our patient population, VF undersensing was observed even in patients with sinus redetection parameters programmed to slow, whereas additional testing without modification of the nominal value for sinus redetection was still associated with appropriate sensing. On the other hand, as a patient demonstrated, regardless of the sinus redetection intervals, appropriate VF sensing will not occur unless the rapid variation of electrogram amplitude breaks. From our experience, it appears that failed shocks predispose to rapid and repetitive changes in the electrogram signals, which can be prevented only by maximizing the chances of success of the initial therapy. An alternative approach is represented by programming the redetection VF interval to a slower rate, which might account for the lack of sensing of the small-amplitude electrograms. Among our patients, the rate of the large-amplitude electrograms ranged from 800 to 395 ms.
The clinical relevance of this finding remains unclear because all but 1 of the patients showing redetection problems received appropriate shock without any evidence of device malfunction. However, this phenomenon certainly can represent a possible mechanism of defibrillator failure in preventing sudden death in patients with life-threatening ventricular arrhythmias. Although subthreshold shock testing is recommended, the inability to document any redetection malfunction does not guarantee that this problem will not occur in the future, given the random nature of this phenomenon. Although our studies seem to suggest that electrode separation may ultimately overcome the problem described with integrated bipolar sensing systems, it appears safer at this time to maximize the chances of success of the first shock or antitachycardia therapy and to prolong, whenever possible, the postshock fibrillation detection interval to avoid electrogram dropout that will eventually alter redetection of VF. This is of paramount importance with the Cadence/Endotak 60 series combination. With such a system, it may be prudent to program high-energy shocks only.
|
Addendum |
|---|
|
TopAbstractIntroductionMethods Results Discussion Addendum References |
|---|
After this article was accepted, testing was performed in 29 additional patients using the Endotak 70 series and the PRx III device in 9 patients, the Cadence–Endotak 70 series in 11, and the Medtronic 7218C–Transvene system in the remaining 9. In 1 patient with the PRx III–Endotak 70 series conbination, after unseccessful shock, undetection of VF was observed that required transthoracic cardioversion. This indicates that an integrated bipolar system with larger interelectrode space might limit but not eliminate the occurrence of this phenomenon, which is still possible. Ultimately, true bipolar sensing appears to be safer.
Received June 21, 1995; revision received August 10, 1995; accepted August 16, 1995.
|
References |
|---|
|
TopAbstractIntroductionMethods Results Discussion Addendum References |
|---|
-
Jung W, Manz M, Moosdorf R, Luderitz B.
Failure of an implantable cardioverter-defibrillator to redetect
ventricular fibrillation in patients with a nonthoracotomy
lead system. Circulation. 1992;86:1217-1222.
[Abstract/Free Full Text]
-
Mann DE, Kelly PA, Damle RS, Reiter MJ.
Undersensing during ventricular
tachyarrhythmias in a third-generation implantable
cardioverter defibrillator: diagnosis using stored electrograms and
correction with programming. Pacing Clin
Electrophysiol. 1994;17:1525-1530. [Medline]
[Order article via Infotrieve]
-
Berul CI, Callans DJ, Schwartzman DS, Preminger MW,
Gottlieb CD, Marchlinski FE. Comparison of initial detection and
redetection of ventricular fibrillation in a transvenous
defibrillator system with automatic gain control. J
Am Coll Cardiol. 1995;25:431-436. [Abstract]
-
Yee R, Jones DL, Klein GJ. Pacing threshold
changes after transvenous catheter countershock. Am J
Cardiol. 1984;53:503-507. [Medline]
[Order article via Infotrieve]
-
Isbruch F, Block M, Hammel D, Borggrefe M, Clerc G,
Dees H, Breithardt G. Influence of defibrillation shocks on the
endocardial sensing signal of an integrated sense/pace/defibrillation
lead. Eur Heart J. 1991;12:363. Abstract.
[Abstract/Free Full Text]
-
Isbruch F, Block M, Wietholt D, Bocker D, Borggrefe M,
Clerc G, Dees H, Breithardt G. Reduction of the endocardial
sensing signal of an integrated sense/pace/defibrillation lead after
application of defibrillation shocks. PACE Pacing Clin
Electrophysiol. 1992;15:II-562. Abstract.
-
Herre JM, for the USA Endotak Investigators. Detection
and redetection characteristics of the Endotak C transvenous lead
system. Circulation. 1993;88(suppl I):I-154.
Abstract.
-
Jung W, Manz M, Accorti P, Moosdorf R, Piel M, Bauer
T, Korte T, Luderitz B. Effect of biphasic shocks on electrogram
amplitudes using a new integrated bipolar transvenous dual chamber
defibrillator electrode system. J Am Coll
Cardiol. 1994;23:420A. Abstract.
-
Yee R, Jones DL, Jarvis E, Donner AP, Klein GJ.
Changes in pacing threshold and R wave amplitude after transvenous
catheter countershock. J Am Coll Cardiol. 1984;4:543-549. [Abstract]
-
Jones JL, Lepeschkin E, Jones RE, Rush S.
Response of cultured myocardial cells to countershock-type
electrical field stimulation. Am J Physiol. 1978;235:H214-H222.
-
Jones JL, Jones RE, Balasky G. Microlesion
formation in myocardial cells by high-intensity electric field
stimulation. Am J Physiol. 1987;253:H480-H486.
[Abstract/Free Full Text]
-
Crowley JM. Electrical breakdown of biomolecular
membranes as an electromechanical instability. Biophys
J. 1973;13:711-724.
-
Penzotti JL, Callans DJ, Gottlieb CD, Foster KR.
Finite element analysis of post-shock sensing
performance in transvenous implantable defibrillator lead
systems. J Am Coll Cardiol. 1995;25:316A. Abstract.
This article has been cited by other articles:
 |
|
 |
|
  C. D. Swerdlow Implantation of Cardioverter Defibrillators Without Induction of Ventricular Fibrillation Circulation, May 1, 2001; 103(17): 2159 - 2164. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||