- Email: [email protected]
Pulse duration variation and electrode size as factors in pacemaker longevity
Pulse duration variation and electrode size as factors in pacemaker longevity For the past 4 years, we have used highly efficient Medtronic electrodes (Nos. 6901 and 6907), with small surface areas, for cardiac pacemakers. We have found that chronic thresholds average 1.6 Ma. at a 1 msec, pulse duration, compared to 4.9 Ma. for the conventional Medtronic No. 5816 electrodes. These electrodes have been used in association with Medtronic Models 5961 and 5931, ventricular-inhibited and asynchronous pulse generators in which output current and voltage are fixed and pulse duration is variable. Drain from the generator battery is directly related to impulse duration and is reduced at shorter durations. The strength-duration curve of cardiac stimulation suggests and actual long-term pacing has been achieved at an average of 0.2 to 0.3 msec, rather than the conventional 1.0 msec. Drain from the battery is one fourth that of pacemakers of 1970 and one half that of present day, conventional units. This fact suggests that a realistic average life of the pulse generator is 4 years or more.
Seymour Furman, M.D., Julius Garvey, M.D., and Philip Hurzeler, Ph.D., Bronx, N. Y.
P J- robably the single most significant problem with implantable cardiac pacemakers is the limited functional life provided by the mercury-zinc cells powering the unit. Presently, most pulse generators are replaced at intervals of two years or less.1 Although alternative energy sources such as biogalvanic energy,- fuel cell,3 nuclear,4 and lithium5 or sodium halide cells have been investigated and used for units in man, the most common power source for cardiac pacemakers is the mercury-zinc cell.0 The limited capacity of this cell is the cause for frequent reoperation. The major approaches taken for the prolongation of pulse generator longevity are as follows: (1) development of new energy sources; (2) careful follow-up so that pulse generators are removed for causes such as From the Cardiothoracic Service, Division of Surgery, Montefiore Hospital and Medical Center, Bronx, N. Y. Supported in Part by U. S. Public Health Service Grant HE 04666-13. Received for publication Aug. 26, 1974. Address for reprints: Dr. Seymour Furman, 111 East 210th Street, Bronx, N. Y. 10467.
382
battery depletion or electronic malfunction, rather than routinely at an elective period 7 - s ; and (3) reduction of current consumption."' 10 The first approach is beyond the scope of this paper and the second is one in which we have been actively engaged with significant results. The present report is concerned with reduction of current consumption and a projected increase in pacer longevity. Theoretical considerations The current and energy requirements for consistent pacing are based on the material and size of the electrode used,11 the impulse duration,12' 13 and several myocardial factors beyond control of the surgeon. The material and electrode size and impulse duration are controllable and do affect the current and energy requirements for pacing. 1. Although the threshold of stimulation can be measured in a variety of variables, the single most useful parameter to determine pacer longevity is measurement of charge in microcoulombs; that is, the threshold as a function of the product of current
Volume 69 Number 3
Factors in pacemaker longevity
383
March, 1975
(in milliamperes) and duration of current flow (in milliseconds). Charge is the major criterion, because it is the measure of the total flow of current from a cell (in this instance, the Mallory* RM-1 mercuric-oxide-zinc cell). For this reason, the cell capacity is rated in milliampere hours (charge), in this case 1,000 per cell. The output voltage of a cell is an indication that charge remains, but it does not define how much charge remains. One characteristic of the mercury-zinc cell is that whether it is fully charged or partially or substantially discharged, the battery voltage remains almost constant, declining only when the cell is on the verge of exhaustion.11 Consequently, reducing either the duration of current flow or the actual current flow itself decreases the expenditure of charge per impulse. Charge consumed with each stimulus decreases, as a direct function of pulse duration, even though threshold current rises somewhat with the decrease of pulse duration. The decreases in pulse duration are far greater, at threshold, than the increase in current required, once pulse durations below 0.5 msec, are used. Consequently, charge (the product of pulse duration and current) is reduced as pulse duration of the pacemaker is decreased, even if current amplitude is kept constant (Fig. 1). 2. There is a close and direct relationship between the threshold of cardiac contraction measured in current or charge and the electrode surface area.11 As electrode size increases, current and charge thresholds increase linearly, for both are a function of the current density at the electrode endocardia! surface. Electrodes of similar configuration and the same metal but of different surface areas require the same current density (milliamperes per square centimeter) to produce a cardiac contraction. Therefore, smaller electrodes require less total current and consequently less charge than larger electrodes. The total energy at threshold is also decreased with reduction in elec*T. R. Mallory Co., Tarrytown, N. Y.
<■> 0.1
0.02 0.05 0.1 Q2
0.5
1
Pulse Duration, Milliseconds
Fig. 1. The value of charge (microcoulombs) plotted against pulse duration indicates that, at threshold, charge is minimal and rises as pulse duration increases. This occurs because pulse duration decreases much more rapidly than the current threshold increases. The chronic threshold curve is identical to the acute curve, although at a higher level (log scales, clinical data).
trode surface area. Energy is proportional to current- x resistance x time (pulse duration) : E = PRT. With reduction in electrode size, electrode electrical resistance is increased. In constant-voltage pulse generators (Medtronic Models 5931 and 5961), an increase in electrode resistance is accompanied by a decrease in current flow and, consequently, a reduction in energy expended by the pulse generator. The increase in resistance is more than overcome by the square of the reduction of current (see formula). The coupling of a low-threshold electrode (with small surface area) and a variable-pulse duration pulse generator allows a major over-all reduction in current drain from the battery. Materials and methods Since March, 1972, seventeen Medtronic Model 5931 asynchronous units (earlier numerical designation was 1254) and sixteen Medtronic Model 5961 ventricular inhibited pulse generators with constant current and voltage have been implanted. The output voltage is 5.2 and current is 10.4 Ma. into a 500 ohm load. Rate declines and pulse width increases with battery exhaustion.1'' In the standard unit, pulse duration is
The Journal of
3 84
Furman,
Garvey,
Hurzeler
Thoracic and Cardiovascular Surgery
0.2msec/div L.
I
i
Fig. 2. Oscilloscopic display of the pulse generator artifact indicates the wide variability of the pulse duration (adjusted noninvasively) and the consequent generator output. A, Pulse duration 0.7 msec. B, Pulse duration 0.5 msec. C, Pulse duration 0.2 msec. D, Pulse duration 0.05 msec. Incidentally, the final duration (0.05 msec.) was the threshold of stimulation in the patient. The amplitude of the impulse remains constant. (Retouched for emphasis.)
variable from a minimum of 0.15 to a maximum of 1.7 msec. About half of the units that we implanted were standard; the other half were specially designed to have the shortest pulse duration—0.05 msec. The units are unipolar, enclosed in titanium (a nonmagnetic metal), and capable of having the pulse duration changed noninvasively. A magnetically activated gear train connected to a variable potentiometer is incorporated within the unit. The pulse duration is varied by rotation of two external bar magnets (contained in a gearoperated controller) over the implanted pulse generator. Their rotation by operation of the controller's hand crank magnetically couples with and rotates the magnet-bearing potentiometer. The pulse duration increases or decreases correspondingly (Fig. 2). Because the orientation of the potentiometer determines whether rotation in a single direction will lengthen or shorten the pulse duration, it is important to place each unit with the lettering (manufacturer's data)
facing outward. Reversal of the implant will require reversal of the direction of the controller rotation. The Model 5961 unit has been designed so that the fluctuating magnetic field does not inhibit the generator"1 (Fig. 3). The indication for implantation of the Model 5931 unit was complete heart block, either at primary implantation or on replacement of an already existing electrode. In 6 cases, the pulse generators of earlier asynchronous units were replaced. The average pulse duration required for longterm pacing was 0.4 msec. In 11 other instances, Medtronic unipolar No. 6907 electrodes were used for primary implantation. One of these electrodes was injected 8 months after implantation, and both the generator, though functioning well, and the electrode were replaced at another institution. Two pacemakers have exhibited rate instability. One, a replacement unit, continuously increased in rate and was removed after 9 months. The other pacer, a carefully
Volume 69 Number 3 March, 1975
Factors in pacemaker longevity
385
Fig. 3. With the magnet in position, pacemaker sensing is lost (first and second QRS complexes). The rotating magnetic field superimposes a sine wave on the electrocardiogram but does not, in Model 5961, inhibit pacer output. The pacer stimuli persist at the automatic rate.
Fig. 4. The threshold of cardiac stimulation is a linear function of the size of the cathode in contact with the endocardium. The Medtronic No. 5816 electrode (top) has a cathode tip surface area of 87 sq. mm. The No. 6901 electrode (bottom) has an area of 12 sq. mm. The ratio is 7:1, and the threshold relationship is about 4:1.
observed primary implant, has increased slowly in rate and continues to function after 22 months. The remaining 14 units are functioning well, 2 to 24 months after implantation. Sixteen patients underwent initial implantation with Model 5961 (ventricular-inhibited) pulse generators and No. 6907 electrodes. One unit was removed because sensing of ventricular signals was inadequate. However, the unit was functioning normally. One patient died of metastatic carcinoma of the lung 1 month following implantation. Another patient died 8 months after implantation, although normal pacemaker function had been demonstrated on the day preceding death. Electrodes
Until energy sources superior to the mercury-zinc cell are available, it will be necessary to limit current drain from the
battery to achieve increased generator longevity. The single major element common to all of the techniques of reducing battery drain is the use of an efficient electrode. The electrodes described have a small surface area, require a small amount of current at threshold, and consequently, since the current output of the pulse generator is fixed, have very low thresholds as a function of time (pulse duration). Pulse generator output should be matched to the requirements of the electrode system used. In constant-voltage pulse generators, such as those described, a smaller surface area with a higher electrode resistance reduces pacer output, whereas a lower electrode resistance increases it. The Medtronic electrode No. 5816, available since 1965, has a surface area of 87 sq. mm. The acute threshold at implantation is 2.0 Ma. at 1 msec. The average threshold for cardiac stimulation of a
The Journal of
386
Furman, Garvey, Hurzeler
Thoracic and Cardiovascular Surgery
I
I
I 5961 J
(5816 Tip In 2 0 % . / Ringer SoluttonL?'
8. 3 0 " S
.02
.05
.1
.2
.5
1 2
Pulse Duration, Millisec.
Fig. 5. The limiting factor in stimulation with generators that have variable pulse durations is the increase in threshold voltage with decreasing pulse duration. The output voltage of these pacemakers is fixed at 5.5 v., corresponding to a pulse duration threshold of about 0.03 msec, acutely and 0.08 msec, chronically. The chronic curve is identical to the acute curve, although at a higher level (log scales, clinical data). chronically implanted electrode is 4.9 Ma. at a pulse duration of 1 msec.17 Medtronic No. 6901 (bipolar) and No. 6907 (unipolar) electrodes implanted over the last 4 years have been evaluated both acutely and chronically at pulse generator replacement. The average threshold at implantation was 0.5 Ma. at a pulse duration of 1 msec, and the average long-term threshold at the same pulse duration was 1.6 Ma. (Fig. 4 ) . With reduction to one seventh of the electrode size, that is, from the area of a Medtronic No. 5816 (87 sq. mm.) to a Medtronic No. 6901 or 6907 (12 sq. mm.) electrode, the electrode resistance rises from 300 to 1,000 ohms. Threshold is reduced to one quarter and pulse generator output to one half. The pairing of these electrodes with variable-pulse duration generators produces the combination of a small surface area, low current threshold, high resistance electrode with a low-output pulse generator capable of pulse duration variation (Fig. 5). (Pacer output at 5.5 v. through 1,000 ohms is 5 to 6 Ma., whereas at 500 ohms it is 10 to 11 Ma.)
6907
E
zo-
% £
22
c
/Electrode ' 6907
18-
C
16-
3 °
12
14
i / '5931 500 .a.
24
"
1
' 5961 500^
'
8 6
O
0.2 0.4 0.6 0.8
1.0
1.2
1.4
1.6 18
20
Pulse Duration (msec) Fig. 6. Battery current drain versus pulse duration. Both pulse duration and electrode cause variation in pulse generator output. The Model 5961 generator at the previously standard pulse duration of 0.85 msec, with the widely used No. 5816 electrode, requires about 28 j«a. The same pulse generator with a No. 6907 electrode, set at 0.3 msec, requires 10 /ia. The Model 5931 unit is asynchronous, without a sensing circuit, and under the same conditions drains 6 /
Volume 69
Factors in pacemaker longevity
Number 3
387
March, 1975
T
S \
35 ua • 23 Months ■ .
\\ C 20
25ua-33 Months
\
Q ♦.
c a>
15
s\
3 io O
\\
\\ J_ 0 2 4 6 810121416
20 24
28 32 36 40 44 48 52 56 60
64 68
72 76 80
Months
Fig. 7. Two generators, one with a drain of 35
At the first patient visit after 1 month, threshold is rechecked. If it has settled to the anticipated chronic level, it is reset between 0.2 and 0.3 msec. All adjustments of pulse duration are performed with a Hewlett-Packard storage oscilloscope and Polaroid photography recording of the impulses. The continuous current drain from the battery of a Model 5961 (ventricularinhibited) generator is 10 ^a* if the unit is pacing continuously at its full output (5 to 6 Ma. per impulse) through a No. 6907 electrode at a pulse duration of 0.2 to 0.3 msec. The drain of a Medtronic Model 5842 (ventricular-inhibited) generator with a pulse duration of 0.9 msec, is 28 /ua through a No. 5816 electrode and 18 fi& through a No. 6901 electrode. With this unit, the saving in charge is 8 to 18 /*a compared to a conventional generator; the amount of charge saved depends on the electrode with which the pacer is used (Fig. 6). Because a constant-voltage pacemaker (10 Ma. output), used with a No. 6907 electrode, will produce normal acute thresholds of 0.05 msec, and because a 0.2 msec. pulse duration is compatible with stable long-term pacing, we have implanted 41 ' A microampere is 1/1,000 of a milliampere.
ventricular-inhibited, unipolar pulse generators with fixed pulse durations of 0.5 msec. Eleven have been in place for over 2 years. These were of two series: Medtronic Model 5843, modified for a reduced pulse duration, and Medtronic Model 5945. No instances of electronic malfunction or battery exhaustion have occurred. Five of the patients have died of causes unrelated to pacer function. One Medtronic Model 5843 unit was replaced by a somewhat higher output pacer (Model 5961) and another by a higher sensitivity Cordis 143K ventricular-inhibited generator. Current drain with continuous pacing is 14 /xa, somewhat higher than the Model 5931 and 5961 units pacing at pulse durations of 0.2 to 0.3 msec. A pulse generator draining about 35 /^a continuously will have an average longevity of about 23 months, whereas a pulse generator operating similarly but with a drain of 25 fxa will have an average longevity of 33 months (Fig. 7). This demonstrated increase of 50 per cent in units now 4 to 5 years old suggests that similar decreases in battery drain might lead to proportional increases in generator longevity. Discussion Threshold current and voltage rise progressively at short pulse durations (i.e., 0.01
The Journal of
388
Fwman,
Garvey, Hurzeler
msec). Total energy (current x voltage x time) consumed at threshold (in microjoules) is highest at ultrashort pulse durations (0.01 to 0.05 msec.) and lowest at pulse durations between 0.2 and 1.0 msec. However, total consumption of electrical charge (microcoulombs) rises rapidly at pulse durations greater than 0.5 msec, but reaches a low, stable plateau at 0.05 to 0.1 msec (Fig. 1). This is true for both acute and chronic pacing, except that the curve of the graph of energy and charge consumption at threshold in chronic stimulation lies above acute levels. A stimulus duration in excess of 1 msec. results in highly inefficient use of electrical charge for both short- and long-term stimulation in man. Considerable savings can be effected by simply reducing impulse duration from 1.0 to 0.5 msec, and maintaining the current output constant. Most pacemakers (including the Medtronic Models 5931 and 5961) have higher current and voltage outputs than required, in order to exceed the expected acute and self-limited rise in threshold after implantation. As voltage and current remain constant, reduction of pulse duration reduces charge, the actual determinant of battery drain. Without a method of reducing output after the critical period of early postimplantation rise in threshold,18 excess current and energy flow are wasted. Once threshold of stimulation is exceeded, consumption of both energy and charge can be a function of pulse duration. With current drain as low as 10 ^a, pacer life can be expected to exceed 4 years and become more a function of battery internal discharge than battery drain. If this expectation is accurate, then approximately 40 to 45 per cent of patients now requiring pacers would not survive long enough to require a second operation.,,J The need for a different energy source for pulse generators would be considerably reduced. Surgeons have recently realized that pacemaker output exceeds the usual chronic threshold and that much of the excess out-
Thoracic and Cardiovascular Surgery
put is required for only a short period after initial implantation. At this time, threshold physiologically rises but then returns to a predictable two to three times threshold at implantation. This realization has led to efforts to reduce pacemaker output after electrode stability has been reached. Though experimental efforts at self adjustment to threshold have been made,-" the major efforts are manual. Three new approaches have become available for clinical use recently. One, described in this manuscript, involves continuous variability of one of the three factors in output, in this instance pulse duration, while voltage and current are kept constant. The second involves use of the programmable pacemaker, the Cordis Omnicor. This unit can be noninvasively programmed through four output steps in which pulse duration is constant and voltage (and consequently current) is set as necessary. The third approach involves use of a full-output 10 Ma. unit, (i.e., perhaps 1 msec, second pulse duration) which is replaced, at its exhaustion, with a unit of substantially lower output (i.e., pulse duration of 0.5 msec, output of 5 Ma., or both). A number of manufacturers have such units available. It remains to be determined which of the approaches will be most effective. In all probability, the effectiveness will depend on the actual reduction of battery drain and not on the method by which this reduction is accomplished. Summary We have implanted two pulse generators in which voltage and current output are fixed but pulse duration can be varied noninvasively. The units are used with highly efficient electrodes. Threshold pulse durations and currents have been achieved which indicate that pulse generator longevities of 4 years may be achieved. By January, 1975, no failure of either the Model 5931 or 5961 generators had occurred.
Volume 69 Number 3
Factors in pacemaker longevity
389
March, 1975
REFERENCES 1 Sykosch, H. J.: Survey on Permanent Cardiac Pacing in 1972: West Germany, Switzerland and Austria, in Thalen, H, J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co. BV, p. 95. 2 Schaldach, M., and Kirsch, U.: In Vivo Electromechanical Power Generation, Trans. Am. Soc. Artif, Intern. Organs 16: 184, 1970. 3 Wolfson, S. K., Prusiner, P., and Gofberg, S. L.: The Bioautofuel Cell: A Device for pacemaker Power From Direct Energy Conversion Consuming Autogenous Fuel, Trans. Am. Soc. Artif. Intern. Organs 14: 198, 1968. 4 Laurens, P., Piwnica, A., Reidemeister, C , Chardack, W. M., and Gage, A. A.: Clinical Results of the Implantations of an Isotopic Pacemaker, in Thalen, H. J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co. BV, p. 198. 5 Greatbatch, W., Lee, J., Mathias, W, Eldrige, M., Moser, J., and Schneider, A.: The SolidState Lithium Battery: A New, Improved, Chemical Power Source for Implantation Cardiac Pacemakers, IEEE Trans. Biomed. Eng. (BME-I8) 5: 317, 1971. 6 Greatbatch, W.: Chemical Power Supplies for Implantable Cardiac Pacemakers, in Thalen, H. J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co. BV, p. 188. 7 Fontaine, G., Kevorkian, M, Bonnet, M., et al.: Definition de la mesure du suil d' entrainement electrique, Ann. Cardiol. Angeiol. 20: 491, 1971. 8 Furman, S.: Transtelephone Observation of Implanted Cardiac Pacemakers, J. Assoc. Adv. Med. Instrum. 7: 196, 1973. 9 Parsonnet, V., Zucker, I. R., Gilbert, L., Lewin, G., Myers, G.H., and Avery, R.: Clinical Use of a New Transvenous Electrode, Ann. N.Y. Acad. Sci. 167: 756, 1969. 10 Furman, S., Escher, D. J. W., and Parker, B.:
Pacemaker Longevity, Am. J. Cardiol. 3 1 : 111, 1973. 11 Furman, S., Parker, B., and Escher, D . J. W.: Decreasing Electrode Size and Increasing Efficiency of Cardiac Stimulation, J. Surg. Res. 1: 105, 1971. 12 Furman, S., Denize, A, Escher, D . J. W., and Schwedel, J. B.: Energy Consumption for Cardiac Stimulation as a Function of Pulse Duration, J. Surg. Res. 6: 441, 1966. 13 Jaros, G., Marchand, P., Milner, M., and Obel, I. W. P.: A Long-Term Study of Threshold Values and Output Characteristics of Cardiac Pacemakers, Thorax 22: 63, 1967. 14 Ruben, S.: Sealed Zinc Mercuric Oxide Cells for Implantable Cardiac Pacemakers, Ann. N. Y. Acad. Sci. 167: 627, 1969. 15 Chardack, W. M., Bakken, E. E„ Bolduc, L., Giori, F. A., and Gage, A. A.: Magnetically Actuated Pulse Width Control for Implantable Pacemakers: Its significance for Follow-up of Patients and the Reduction of Current Drain, Ann. Cardiol. Angeiol. 20: 345, 1971. 16 Furman, S., Escher, D. J. W., and Parker, B.: Failure of Triggered Pacemakers, Am. Heart J. 82: 28, 1971. 17 Furman, S., and Escher, D. J. W.: Principles and Techniques of Cardiac Pacing. New York, 1970, Harper & Row, Publishers, p. 40. 18 Davies, J. G., and Sowton, G. E.: Electrical Threshold of the Human Heart, Br. Heart J. 28: 231, 1966. 19 Furman, S., Escher, D. J. W., and Parker, B.: Results of Long-Term Pacemaker Implantation, in Dreifus, L. S., and Linoff, W., editors: Cardiac Arrhythmias, New York, 1973, Grune & Stratton, Inc., p. 599. 20 Preston, T. A., and Bowers, D. L.: Report of a Continuous Threshold Tracking System, in Thalen, H. J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co., BV, p. 295.
Seymour Furman, M.D., Julius Garvey, M.D., and Philip Hurzeler, Ph.D., Bronx, N. Y.
P J- robably the single most significant problem with implantable cardiac pacemakers is the limited functional life provided by the mercury-zinc cells powering the unit. Presently, most pulse generators are replaced at intervals of two years or less.1 Although alternative energy sources such as biogalvanic energy,- fuel cell,3 nuclear,4 and lithium5 or sodium halide cells have been investigated and used for units in man, the most common power source for cardiac pacemakers is the mercury-zinc cell.0 The limited capacity of this cell is the cause for frequent reoperation. The major approaches taken for the prolongation of pulse generator longevity are as follows: (1) development of new energy sources; (2) careful follow-up so that pulse generators are removed for causes such as From the Cardiothoracic Service, Division of Surgery, Montefiore Hospital and Medical Center, Bronx, N. Y. Supported in Part by U. S. Public Health Service Grant HE 04666-13. Received for publication Aug. 26, 1974. Address for reprints: Dr. Seymour Furman, 111 East 210th Street, Bronx, N. Y. 10467.
382
battery depletion or electronic malfunction, rather than routinely at an elective period 7 - s ; and (3) reduction of current consumption."' 10 The first approach is beyond the scope of this paper and the second is one in which we have been actively engaged with significant results. The present report is concerned with reduction of current consumption and a projected increase in pacer longevity. Theoretical considerations The current and energy requirements for consistent pacing are based on the material and size of the electrode used,11 the impulse duration,12' 13 and several myocardial factors beyond control of the surgeon. The material and electrode size and impulse duration are controllable and do affect the current and energy requirements for pacing. 1. Although the threshold of stimulation can be measured in a variety of variables, the single most useful parameter to determine pacer longevity is measurement of charge in microcoulombs; that is, the threshold as a function of the product of current
Volume 69 Number 3
Factors in pacemaker longevity
383
March, 1975
(in milliamperes) and duration of current flow (in milliseconds). Charge is the major criterion, because it is the measure of the total flow of current from a cell (in this instance, the Mallory* RM-1 mercuric-oxide-zinc cell). For this reason, the cell capacity is rated in milliampere hours (charge), in this case 1,000 per cell. The output voltage of a cell is an indication that charge remains, but it does not define how much charge remains. One characteristic of the mercury-zinc cell is that whether it is fully charged or partially or substantially discharged, the battery voltage remains almost constant, declining only when the cell is on the verge of exhaustion.11 Consequently, reducing either the duration of current flow or the actual current flow itself decreases the expenditure of charge per impulse. Charge consumed with each stimulus decreases, as a direct function of pulse duration, even though threshold current rises somewhat with the decrease of pulse duration. The decreases in pulse duration are far greater, at threshold, than the increase in current required, once pulse durations below 0.5 msec, are used. Consequently, charge (the product of pulse duration and current) is reduced as pulse duration of the pacemaker is decreased, even if current amplitude is kept constant (Fig. 1). 2. There is a close and direct relationship between the threshold of cardiac contraction measured in current or charge and the electrode surface area.11 As electrode size increases, current and charge thresholds increase linearly, for both are a function of the current density at the electrode endocardia! surface. Electrodes of similar configuration and the same metal but of different surface areas require the same current density (milliamperes per square centimeter) to produce a cardiac contraction. Therefore, smaller electrodes require less total current and consequently less charge than larger electrodes. The total energy at threshold is also decreased with reduction in elec*T. R. Mallory Co., Tarrytown, N. Y.
<■> 0.1
0.02 0.05 0.1 Q2
0.5
1
Pulse Duration, Milliseconds
Fig. 1. The value of charge (microcoulombs) plotted against pulse duration indicates that, at threshold, charge is minimal and rises as pulse duration increases. This occurs because pulse duration decreases much more rapidly than the current threshold increases. The chronic threshold curve is identical to the acute curve, although at a higher level (log scales, clinical data).
trode surface area. Energy is proportional to current- x resistance x time (pulse duration) : E = PRT. With reduction in electrode size, electrode electrical resistance is increased. In constant-voltage pulse generators (Medtronic Models 5931 and 5961), an increase in electrode resistance is accompanied by a decrease in current flow and, consequently, a reduction in energy expended by the pulse generator. The increase in resistance is more than overcome by the square of the reduction of current (see formula). The coupling of a low-threshold electrode (with small surface area) and a variable-pulse duration pulse generator allows a major over-all reduction in current drain from the battery. Materials and methods Since March, 1972, seventeen Medtronic Model 5931 asynchronous units (earlier numerical designation was 1254) and sixteen Medtronic Model 5961 ventricular inhibited pulse generators with constant current and voltage have been implanted. The output voltage is 5.2 and current is 10.4 Ma. into a 500 ohm load. Rate declines and pulse width increases with battery exhaustion.1'' In the standard unit, pulse duration is
The Journal of
3 84
Furman,
Garvey,
Hurzeler
Thoracic and Cardiovascular Surgery
0.2msec/div L.
I
i
Fig. 2. Oscilloscopic display of the pulse generator artifact indicates the wide variability of the pulse duration (adjusted noninvasively) and the consequent generator output. A, Pulse duration 0.7 msec. B, Pulse duration 0.5 msec. C, Pulse duration 0.2 msec. D, Pulse duration 0.05 msec. Incidentally, the final duration (0.05 msec.) was the threshold of stimulation in the patient. The amplitude of the impulse remains constant. (Retouched for emphasis.)
variable from a minimum of 0.15 to a maximum of 1.7 msec. About half of the units that we implanted were standard; the other half were specially designed to have the shortest pulse duration—0.05 msec. The units are unipolar, enclosed in titanium (a nonmagnetic metal), and capable of having the pulse duration changed noninvasively. A magnetically activated gear train connected to a variable potentiometer is incorporated within the unit. The pulse duration is varied by rotation of two external bar magnets (contained in a gearoperated controller) over the implanted pulse generator. Their rotation by operation of the controller's hand crank magnetically couples with and rotates the magnet-bearing potentiometer. The pulse duration increases or decreases correspondingly (Fig. 2). Because the orientation of the potentiometer determines whether rotation in a single direction will lengthen or shorten the pulse duration, it is important to place each unit with the lettering (manufacturer's data)
facing outward. Reversal of the implant will require reversal of the direction of the controller rotation. The Model 5961 unit has been designed so that the fluctuating magnetic field does not inhibit the generator"1 (Fig. 3). The indication for implantation of the Model 5931 unit was complete heart block, either at primary implantation or on replacement of an already existing electrode. In 6 cases, the pulse generators of earlier asynchronous units were replaced. The average pulse duration required for longterm pacing was 0.4 msec. In 11 other instances, Medtronic unipolar No. 6907 electrodes were used for primary implantation. One of these electrodes was injected 8 months after implantation, and both the generator, though functioning well, and the electrode were replaced at another institution. Two pacemakers have exhibited rate instability. One, a replacement unit, continuously increased in rate and was removed after 9 months. The other pacer, a carefully
Volume 69 Number 3 March, 1975
Factors in pacemaker longevity
385
Fig. 3. With the magnet in position, pacemaker sensing is lost (first and second QRS complexes). The rotating magnetic field superimposes a sine wave on the electrocardiogram but does not, in Model 5961, inhibit pacer output. The pacer stimuli persist at the automatic rate.
Fig. 4. The threshold of cardiac stimulation is a linear function of the size of the cathode in contact with the endocardium. The Medtronic No. 5816 electrode (top) has a cathode tip surface area of 87 sq. mm. The No. 6901 electrode (bottom) has an area of 12 sq. mm. The ratio is 7:1, and the threshold relationship is about 4:1.
observed primary implant, has increased slowly in rate and continues to function after 22 months. The remaining 14 units are functioning well, 2 to 24 months after implantation. Sixteen patients underwent initial implantation with Model 5961 (ventricular-inhibited) pulse generators and No. 6907 electrodes. One unit was removed because sensing of ventricular signals was inadequate. However, the unit was functioning normally. One patient died of metastatic carcinoma of the lung 1 month following implantation. Another patient died 8 months after implantation, although normal pacemaker function had been demonstrated on the day preceding death. Electrodes
Until energy sources superior to the mercury-zinc cell are available, it will be necessary to limit current drain from the
battery to achieve increased generator longevity. The single major element common to all of the techniques of reducing battery drain is the use of an efficient electrode. The electrodes described have a small surface area, require a small amount of current at threshold, and consequently, since the current output of the pulse generator is fixed, have very low thresholds as a function of time (pulse duration). Pulse generator output should be matched to the requirements of the electrode system used. In constant-voltage pulse generators, such as those described, a smaller surface area with a higher electrode resistance reduces pacer output, whereas a lower electrode resistance increases it. The Medtronic electrode No. 5816, available since 1965, has a surface area of 87 sq. mm. The acute threshold at implantation is 2.0 Ma. at 1 msec. The average threshold for cardiac stimulation of a
The Journal of
386
Furman, Garvey, Hurzeler
Thoracic and Cardiovascular Surgery
I
I
I 5961 J
(5816 Tip In 2 0 % . / Ringer SoluttonL?'
8. 3 0 " S
.02
.05
.1
.2
.5
1 2
Pulse Duration, Millisec.
Fig. 5. The limiting factor in stimulation with generators that have variable pulse durations is the increase in threshold voltage with decreasing pulse duration. The output voltage of these pacemakers is fixed at 5.5 v., corresponding to a pulse duration threshold of about 0.03 msec, acutely and 0.08 msec, chronically. The chronic curve is identical to the acute curve, although at a higher level (log scales, clinical data). chronically implanted electrode is 4.9 Ma. at a pulse duration of 1 msec.17 Medtronic No. 6901 (bipolar) and No. 6907 (unipolar) electrodes implanted over the last 4 years have been evaluated both acutely and chronically at pulse generator replacement. The average threshold at implantation was 0.5 Ma. at a pulse duration of 1 msec, and the average long-term threshold at the same pulse duration was 1.6 Ma. (Fig. 4 ) . With reduction to one seventh of the electrode size, that is, from the area of a Medtronic No. 5816 (87 sq. mm.) to a Medtronic No. 6901 or 6907 (12 sq. mm.) electrode, the electrode resistance rises from 300 to 1,000 ohms. Threshold is reduced to one quarter and pulse generator output to one half. The pairing of these electrodes with variable-pulse duration generators produces the combination of a small surface area, low current threshold, high resistance electrode with a low-output pulse generator capable of pulse duration variation (Fig. 5). (Pacer output at 5.5 v. through 1,000 ohms is 5 to 6 Ma., whereas at 500 ohms it is 10 to 11 Ma.)
6907
E
zo-
% £
22
c
/Electrode ' 6907
18-
C
16-
3 °
12
14
i / '5931 500 .a.
24
"
1
' 5961 500^
'
8 6
O
0.2 0.4 0.6 0.8
1.0
1.2
1.4
1.6 18
20
Pulse Duration (msec) Fig. 6. Battery current drain versus pulse duration. Both pulse duration and electrode cause variation in pulse generator output. The Model 5961 generator at the previously standard pulse duration of 0.85 msec, with the widely used No. 5816 electrode, requires about 28 j«a. The same pulse generator with a No. 6907 electrode, set at 0.3 msec, requires 10 /ia. The Model 5931 unit is asynchronous, without a sensing circuit, and under the same conditions drains 6 /
Volume 69
Factors in pacemaker longevity
Number 3
387
March, 1975
T
S \
35 ua • 23 Months ■ .
\\ C 20
25ua-33 Months
\
Q ♦.
c a>
15
s\
3 io O
\\
\\ J_ 0 2 4 6 810121416
20 24
28 32 36 40 44 48 52 56 60
64 68
72 76 80
Months
Fig. 7. Two generators, one with a drain of 35
At the first patient visit after 1 month, threshold is rechecked. If it has settled to the anticipated chronic level, it is reset between 0.2 and 0.3 msec. All adjustments of pulse duration are performed with a Hewlett-Packard storage oscilloscope and Polaroid photography recording of the impulses. The continuous current drain from the battery of a Model 5961 (ventricularinhibited) generator is 10 ^a* if the unit is pacing continuously at its full output (5 to 6 Ma. per impulse) through a No. 6907 electrode at a pulse duration of 0.2 to 0.3 msec. The drain of a Medtronic Model 5842 (ventricular-inhibited) generator with a pulse duration of 0.9 msec, is 28 /ua through a No. 5816 electrode and 18 fi& through a No. 6901 electrode. With this unit, the saving in charge is 8 to 18 /*a compared to a conventional generator; the amount of charge saved depends on the electrode with which the pacer is used (Fig. 6). Because a constant-voltage pacemaker (10 Ma. output), used with a No. 6907 electrode, will produce normal acute thresholds of 0.05 msec, and because a 0.2 msec. pulse duration is compatible with stable long-term pacing, we have implanted 41 ' A microampere is 1/1,000 of a milliampere.
ventricular-inhibited, unipolar pulse generators with fixed pulse durations of 0.5 msec. Eleven have been in place for over 2 years. These were of two series: Medtronic Model 5843, modified for a reduced pulse duration, and Medtronic Model 5945. No instances of electronic malfunction or battery exhaustion have occurred. Five of the patients have died of causes unrelated to pacer function. One Medtronic Model 5843 unit was replaced by a somewhat higher output pacer (Model 5961) and another by a higher sensitivity Cordis 143K ventricular-inhibited generator. Current drain with continuous pacing is 14 /xa, somewhat higher than the Model 5931 and 5961 units pacing at pulse durations of 0.2 to 0.3 msec. A pulse generator draining about 35 /^a continuously will have an average longevity of about 23 months, whereas a pulse generator operating similarly but with a drain of 25 fxa will have an average longevity of 33 months (Fig. 7). This demonstrated increase of 50 per cent in units now 4 to 5 years old suggests that similar decreases in battery drain might lead to proportional increases in generator longevity. Discussion Threshold current and voltage rise progressively at short pulse durations (i.e., 0.01
The Journal of
388
Fwman,
Garvey, Hurzeler
msec). Total energy (current x voltage x time) consumed at threshold (in microjoules) is highest at ultrashort pulse durations (0.01 to 0.05 msec.) and lowest at pulse durations between 0.2 and 1.0 msec. However, total consumption of electrical charge (microcoulombs) rises rapidly at pulse durations greater than 0.5 msec, but reaches a low, stable plateau at 0.05 to 0.1 msec (Fig. 1). This is true for both acute and chronic pacing, except that the curve of the graph of energy and charge consumption at threshold in chronic stimulation lies above acute levels. A stimulus duration in excess of 1 msec. results in highly inefficient use of electrical charge for both short- and long-term stimulation in man. Considerable savings can be effected by simply reducing impulse duration from 1.0 to 0.5 msec, and maintaining the current output constant. Most pacemakers (including the Medtronic Models 5931 and 5961) have higher current and voltage outputs than required, in order to exceed the expected acute and self-limited rise in threshold after implantation. As voltage and current remain constant, reduction of pulse duration reduces charge, the actual determinant of battery drain. Without a method of reducing output after the critical period of early postimplantation rise in threshold,18 excess current and energy flow are wasted. Once threshold of stimulation is exceeded, consumption of both energy and charge can be a function of pulse duration. With current drain as low as 10 ^a, pacer life can be expected to exceed 4 years and become more a function of battery internal discharge than battery drain. If this expectation is accurate, then approximately 40 to 45 per cent of patients now requiring pacers would not survive long enough to require a second operation.,,J The need for a different energy source for pulse generators would be considerably reduced. Surgeons have recently realized that pacemaker output exceeds the usual chronic threshold and that much of the excess out-
Thoracic and Cardiovascular Surgery
put is required for only a short period after initial implantation. At this time, threshold physiologically rises but then returns to a predictable two to three times threshold at implantation. This realization has led to efforts to reduce pacemaker output after electrode stability has been reached. Though experimental efforts at self adjustment to threshold have been made,-" the major efforts are manual. Three new approaches have become available for clinical use recently. One, described in this manuscript, involves continuous variability of one of the three factors in output, in this instance pulse duration, while voltage and current are kept constant. The second involves use of the programmable pacemaker, the Cordis Omnicor. This unit can be noninvasively programmed through four output steps in which pulse duration is constant and voltage (and consequently current) is set as necessary. The third approach involves use of a full-output 10 Ma. unit, (i.e., perhaps 1 msec, second pulse duration) which is replaced, at its exhaustion, with a unit of substantially lower output (i.e., pulse duration of 0.5 msec, output of 5 Ma., or both). A number of manufacturers have such units available. It remains to be determined which of the approaches will be most effective. In all probability, the effectiveness will depend on the actual reduction of battery drain and not on the method by which this reduction is accomplished. Summary We have implanted two pulse generators in which voltage and current output are fixed but pulse duration can be varied noninvasively. The units are used with highly efficient electrodes. Threshold pulse durations and currents have been achieved which indicate that pulse generator longevities of 4 years may be achieved. By January, 1975, no failure of either the Model 5931 or 5961 generators had occurred.
Volume 69 Number 3
Factors in pacemaker longevity
389
March, 1975
REFERENCES 1 Sykosch, H. J.: Survey on Permanent Cardiac Pacing in 1972: West Germany, Switzerland and Austria, in Thalen, H, J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co. BV, p. 95. 2 Schaldach, M., and Kirsch, U.: In Vivo Electromechanical Power Generation, Trans. Am. Soc. Artif, Intern. Organs 16: 184, 1970. 3 Wolfson, S. K., Prusiner, P., and Gofberg, S. L.: The Bioautofuel Cell: A Device for pacemaker Power From Direct Energy Conversion Consuming Autogenous Fuel, Trans. Am. Soc. Artif. Intern. Organs 14: 198, 1968. 4 Laurens, P., Piwnica, A., Reidemeister, C , Chardack, W. M., and Gage, A. A.: Clinical Results of the Implantations of an Isotopic Pacemaker, in Thalen, H. J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co. BV, p. 198. 5 Greatbatch, W., Lee, J., Mathias, W, Eldrige, M., Moser, J., and Schneider, A.: The SolidState Lithium Battery: A New, Improved, Chemical Power Source for Implantation Cardiac Pacemakers, IEEE Trans. Biomed. Eng. (BME-I8) 5: 317, 1971. 6 Greatbatch, W.: Chemical Power Supplies for Implantable Cardiac Pacemakers, in Thalen, H. J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co. BV, p. 188. 7 Fontaine, G., Kevorkian, M, Bonnet, M., et al.: Definition de la mesure du suil d' entrainement electrique, Ann. Cardiol. Angeiol. 20: 491, 1971. 8 Furman, S.: Transtelephone Observation of Implanted Cardiac Pacemakers, J. Assoc. Adv. Med. Instrum. 7: 196, 1973. 9 Parsonnet, V., Zucker, I. R., Gilbert, L., Lewin, G., Myers, G.H., and Avery, R.: Clinical Use of a New Transvenous Electrode, Ann. N.Y. Acad. Sci. 167: 756, 1969. 10 Furman, S., Escher, D. J. W., and Parker, B.:
Pacemaker Longevity, Am. J. Cardiol. 3 1 : 111, 1973. 11 Furman, S., Parker, B., and Escher, D . J. W.: Decreasing Electrode Size and Increasing Efficiency of Cardiac Stimulation, J. Surg. Res. 1: 105, 1971. 12 Furman, S., Denize, A, Escher, D . J. W., and Schwedel, J. B.: Energy Consumption for Cardiac Stimulation as a Function of Pulse Duration, J. Surg. Res. 6: 441, 1966. 13 Jaros, G., Marchand, P., Milner, M., and Obel, I. W. P.: A Long-Term Study of Threshold Values and Output Characteristics of Cardiac Pacemakers, Thorax 22: 63, 1967. 14 Ruben, S.: Sealed Zinc Mercuric Oxide Cells for Implantable Cardiac Pacemakers, Ann. N. Y. Acad. Sci. 167: 627, 1969. 15 Chardack, W. M., Bakken, E. E„ Bolduc, L., Giori, F. A., and Gage, A. A.: Magnetically Actuated Pulse Width Control for Implantable Pacemakers: Its significance for Follow-up of Patients and the Reduction of Current Drain, Ann. Cardiol. Angeiol. 20: 345, 1971. 16 Furman, S., Escher, D. J. W., and Parker, B.: Failure of Triggered Pacemakers, Am. Heart J. 82: 28, 1971. 17 Furman, S., and Escher, D. J. W.: Principles and Techniques of Cardiac Pacing. New York, 1970, Harper & Row, Publishers, p. 40. 18 Davies, J. G., and Sowton, G. E.: Electrical Threshold of the Human Heart, Br. Heart J. 28: 231, 1966. 19 Furman, S., Escher, D. J. W., and Parker, B.: Results of Long-Term Pacemaker Implantation, in Dreifus, L. S., and Linoff, W., editors: Cardiac Arrhythmias, New York, 1973, Grune & Stratton, Inc., p. 599. 20 Preston, T. A., and Bowers, D. L.: Report of a Continuous Threshold Tracking System, in Thalen, H. J. Th., editor: Cardiac Pacing, The Netherlands, 1973, Van Gorcum & Co., BV, p. 295.