FULL TEXT (全文ー英語のみ)

An Evaluation of the Clinical Experience with Atrial Pacing at the Newark Beth Israel Medical Center

Doug Berger

Department of Surgery, Newark Beth Israel Medical Center, July-August 1979.

(Note: Tables and Figures are not available on-line.)


Within the past few years, there has been an increased interest in the management of complex arrhythmias, and the recognition of the atrial "kick" as beneficial to many conduction disturbances (1,2). This is especially important for those with myocardial disease, as ventricular pacing has been shown to increase myocardial oxygen requirements, and aggrivate angina on some ischemic patients (3). furthermore, ventricular pacing results in irregular filling and output, aggrivates mitral insufficiency, and retrograde conduction may produce re-entry tachycardias and worsen atrial tachyarrhythmias (4,5).

The advantages of atrial and related modes of pacing are: 1.) preservation of atrioventricular synchrony; 2.) normal ventricular depolarization; and 3.) suppression of ectopic rhythms in both chambers. These advantages can bring about beneficial results such as increased cardiac output, reduction of angina and congestive failure, regular ventricular filling, and prevention of retrograde conduction, re-entry arrhythmias (3).

Despite these advantages, the implantation rate of atrial pacemakers is only about one percent worldwide (6). The failure to utilize more fully the benefits of atrial pacing is largely a result of past difficulties in securely positioning permanent transvenous electrodes of several designes.

As it is a relatively new field, acceptance of atrial pacing as a dependaple means of controlling conduction disturbances will surely spark improvements in existing atrial pulse generators and in the development of more sophisticated devices.

This study analyzed the effectiveness of atrial pacing of all types in the total clinical experience of paced patients at the Newark Beth Israel Medical Center (NBIMC), paying particular attention to the electrode malfunctions, in order to assess the security of long-term atrial pacing.


From November 1967 through April 1979, NBIMC implanted 100 patients with atrial leads. Data were obtained via clinic charts, threshhold charts, telephone monitoring records, computer lists, and the FDA files. Tabulation of results into statistical forms (tables, graphs, etc.), with the exception of threshold computations and the actuarial curves, were done manually by a computer.

Six additional cases of the Medtronic 1X366D electrode implanted after the study period are also analyzed in the relative electrode actuarial survival curves and threshold data.


Patient Population Characteristics

In November of 1967, the first patient to receive an atrial implant at NBIMC was a 78 year-old female with complete heart block. Up to this study (thru April 1979), 99 additional patients have had atrial and related modes of pacemakers implanted, with an average of approximately 1.6 operative interventions per patient. This represents about 60% implantation rate for all patients ever implanted with pacemakers at NBIMC. In the past three years however, this figure has risen to about 15%.

The age and sex distribution for this study group are presented in Figure 1. Patients ranged in age from 28 to 91 years with a mean age of 65.5 years. Exactly 67% of the patients were male with a mean age of 64, while females had a mean age of 68. Females ranged in age from 43-86 which is .68 the size of the male range of 28-91.

Males clustered around the 61-70 group while the female population rose steadily from the 41-50 group up to the 71-80 group, which had the greatest affinity for females, and then fell sharply in the 81-90 age group.

Indications for Pacing: Systems Used

Four different systems were used in this study population to overcome their disturbances. In original implants, 47% AAI, 30% used DVI, 18% used VAT, and 5% used RF (radio-frequency) (Table 16). Parsonnet (7) describes the nomenclature code for implantable cardiac pacemakers.

In diseases of the sinus node alone, the atrial conduction mechanism may be maintained by pacing the atrium with AAI mode. In patients with diseases of the AV node or conduction system alone, the mechanism is preserved by sensing the atrium and then synchronized ventricular pacing (VAT). When both systems are involved A-V sequential pacing of both chambers must be used (DVI) (2). In patients with episodes of atrial tachycardia, a radio-frequency unit was employed to overdrive the arrhythmias.

The breakdown of electrocardiographic indications for pacing for the initial implants of the 100 patients are as follows: 62% had sick sinus syndrome (SSS); 23% complete heart block (CHB); 6% CHB and SSS, 4% SSS and other arrhythmias; 2% incomplete A-V block, 2% other arrhythmias, and 1% had SSS and incomplete A-V block.

Device Population

Pulse generator distribution by mode and company is depicted in Table 1. The most widely used company was Cordis, which dominated both the AAI and VAT modes. Cordis, Medtronic, and American Optical with percentages of 53%, 21%, and 19% respectively, contributed to 93% of all the original pacers implanted. American Optical was used only in the DVI mode and contributed to almost two-thirds of these pacers.

Table 2 displays the use of electrodes by company. Medtronic-Smyth and American optical transvenous electrodes were implanted both in 41% of original cases respectively, completely dominating the others.

Patient Longevity

Figure 2 depicts the survival curve for atrial patients. Table 3 is the raw data for the construction of the curve, and Table 4 lists the statistical probabilities. Experience indicates that in 5 years, 49% of the patients died. The first three month interval proved to be the most crucial as 10 of 31 (32%, Table 3) of the failures occured here. The gross figure of the curve seems to take a sigmoidal shape characteristic of biological plots.

Electrode Longevity

Figure 3 depicts the electrode longevity of atrial patients including the myocardial electrodes. Table 5 is the raw data, and Table 6 lists the statistical probabilities; the curve is very similar to that for the patients, with 41% failing in the first 5 years. The first interval in this group also saw a high number of failures as 12 of the 28 failures occured here. This curve is sigmoidal in shape.

Relative electrode survival by company is presented in Figure 4. Table 7 is the raw data and Tables 8-10 are the statistical probabilities. "All other" electrodes include 6 Cordis-J, 3 Biotronik-Shalbach, and 4 Medtronic 1X366D (the Medtronic plot contains 10 electrodes- 6 implanted after the study period).

Early failures hit hardest on "all others" the worst with 84% surviving in the first three month interval. Medtronic-Smyth had 90% surviving here, American Optical 94%, and the Medtronic 1X366D hasn't had a failure up to 9 months now. In the long run, 39% of the American Optical electrodes were surviving after 5 years. By 2 1/2 years, there were no more Medtronic-Smyth electrodes in any of the living patients, however, the survival up to that point was 73% compared to American Optical (71%) and all others 56%.

Electrode Complications and Management

Table 2 lists every complication which resulted in repositioning or disconnection of the atrial electrode. Of the 100 original patients, there were 2 replacements (neither of which has had a complication) making 102 fresh electrode insertions. Seven of the 100 original pacemakers were repositioned totaling 109 electrode lives. Since 5 electrodes were myocardial, there were 104 transvenous implants. None of the myocardial electroded have malfunctioned. Twenty-eight of the 104 transvenous lives (26.9%) had to be repositioned or removed, 7 (6.7%) were repositioned and 21 (20.2%) were removed (two of which were replaced, leaving 19 (18.3%) abandoned).

Eleven of 104 (10.5%) insertions resulted in early malfunctios with the high threshold being the major cause. Table 2 depicts this data and Figure 5 illustrates the time distribution. Seven of these resulted in repositioning, 3 were removed, and 1 replaced. Eight of 11 early malfunctions were Medtronic-Smyth electrodes.

Eight of the 104 (7.7%) lives resulted in late electrode malfunction. Four of these were high thresholds (range:6-100 mos), and 4 sensing failures (6-61 mos.). This totals 19 electrode malfunctions (11 early, plus 8 late), one of which (early dislocation) was a recurrence (high threshold). This should not be confused with the 19 abandonments, 9 of which were abandoned for non electrode related malfunctions. Table 13 (not available on line) summarizes the total (early and late malfunctions) along with the non-electrode related malfunctions.

Table 14 reviews the abandoned atrial wires. nineteen of the 104 (18.3%) fresh implants had to be abandoned and charged to ventricular pacing, except one patient whose system was removed completely. Sixteen of these were pacing wires which makes 16 of the 83 minus 2 (epicardium)= 81 (19.8%) original pacing wire implants (AAI, DVI, and RF; see Table 1) abandoned. Three sensing wires (VAT) our of the 17 minus 3 (epicardium) = 14 sensing implants (21%) were abandoned. The abandonments ranged from 2 days to 100 months. Thirteen of the 19 were American Optical units. Of the bifocal electrodes, 4 were Cordis-J, and only 2 were Medtronic-Smyth.

The mode changes involved are: 8 AAI to VVI; 4 DVI to VVI; 1 DVI to VOO; 1 DVI removed; 2 RF to VVI; 2 VAT to VVI; and 1 VAT to VOO. Table 15 summarizes the reasons for abandonment. high threshold contributed to the largest amount (5 of 19, 26%).

Repositioned electrodes are reviewed of Table 16. These were all the result of an early malfunction and ranged from 1 to 90 days and 6.4% of all the electrode lives were repositioned. Seven (7.4%) of the 95 original transvenous implants and 6.7% of all electrode lives were repositioned. All were pacing wires; this means that 8.6% of all pacing wires were repositioned while no sensing wires were. None needed to be repositioned again, however, one was abandoned to a ventricular mode.

The reasons for repositioning were high threshold, 3; dislodgement, 3; and poor sensing, 1. Five of the 7 repositioned electrodes were Medtronic-smyths. Possible extenuating factors contributing to malfunction were identified in 12 patients. In 6 instances there was concurrent arteriosclerotic heart disease, of which 2 had experienced an early dislodgement and the other 4 high thresholds; 2 had rheumatic heart disease and experienced high thresholds; 2 had undetermined heart disease and had poor sensing; 1 had hypertension and had high thresholds; 1 became dislodged after 3 monthe in a fall from a revolving door accident.

Early malfunctions by company were compared and tested (chi-square) for statistical difference (Table 17). No difference was found between Medtronic-Smyth (17%); American Optical T.V. (4.8%); Cordis-J (0%); and Medtronic 1X366D (0%) or any combination of them (at a 95% conficence level). This means that at least 95% of the time these differences can be obtained randomly.

Age-sex Distribution of Managed Patients

Figure 6 displays the age-sex distribution of the patients with electrode removal or repositioning. The ages of the males ranged from 30 to 82 years-old and the females ranged from 57 to 78 years-old. The female ages, male ages, and combined-average age at implant are essentially the same as for the total atrial population at 67, 64, and 65 years-old respectively. The gender percent is also significantly different, the males comprising 65% of the distribution. Both sexes clustered around the 61-70 year-old age group.

Threshold Data

Acute threshold data was compiled and averaged for each type of electrode (Table 17). Data for 6 of the 10 Medtronic 1X366D electrodes were obtained after the time period of this study as it is a new model whose exceptional performance is in need of numerical evaluation. Mean current readings at 1.5 ms duration indicated that the 1X366D electrode (.7 mA) was significantly lower (p<.05) than all the others. Medtronic-Smyth: .957 mA, American Optical T.V.: 1.45mA, and Cordis-J: 1.85 mA. All these values are significantly different from each other except for Cordis and American Optical T.V.. Thresholds of electrodes other than those used transvenously cannot be compared to those used transvenously and will not be discussed here.

Voltage thresholds also showed the 1X366D electrode to be significantly superior to all except the Cordis-J, but this is due to the fact that only 2 Cordis electrodes had available information. The p wave showed no significance between electrodes except for American Optical, which was significantly lower than the others: 64% at the nearest value. Peak-to-peak values were all statistically indifferent. The energy values gave the 1X366D electrode a value that is only 55% of its nearest competitor, the Medtronic-Smyth (.75 mJ to 1.37 mJ), while American Optical (2.23mJ) and Cordis-j (3.56mJ) were far out of the running.

Outcome of Original Modes

Table 18 displays the outcome of the 100 original atrial implants by mode. 47% are still alive with their original mode; 25% were lost to death;18% were abandoned to a ventricular mode with no preference by mode, except for RF (however the RF sample size is very small); 5% have been changed to a related atrial mode; 4% have been lost to follow-up; and 1% was removed completely.


Since 1967 the NBIMC implanted atrial pacemakers at a rate of 6%. In the past 3 years, however, this figure has risen to 15% indicating a greater interest in controlling complicated arrythmias, controlling disturbances previously paced ventricularly, and the need for research and improvements in this area.

Analysis of the age-sex distribution of all atrial patients (Figure 11) compared with those that were managed (Figure 6), showed the distribution of Fig. 1 to have the greatest percent of males (29%) to fall in the 61-70 age group. Fig. 6 also shows the greater percent of males (39%) within this age group.

In contrast, the greatest percent of females (39%) fell into the 71-80 group with 24% in the 61-70 age group. Figure 6 shows the greatest percentage of females (60%) to fall in the 61-70 age group, while 30% fell into the 71-80 age group. This in itself isn't significant, however, if one takes a broader look at the distribution, it is evident that a greater percentage of females (figure 6) lie in the 61 and over region as compared with figure 2 (90% in figure 6 to 73% in figure 2).

Males on the other hand, showed little deviance in this respect (67% in figure 6 to 64% in figure 1). This can possibly indicate that males may be greater prone to have a malfunction at an earlier age than females. This suggests that diffuse fibrosis of age (increase in the collagenous fibers in the endocardium) occurs faster in males. This initiates destruction of trabeculae, which may then lead to migration and/or perforation of the electrode. It may also be that this fibrosis encapsulates the electrode and contributes to a greater stability, or eventually produce too high a threshold; the former working against this theory and the latter for it. In early malfunction, it may be that the trabeculae are very small or non-existant, more so in older patients which would explain their greater frequence of repositioning and /or discontinuation.

Comparison of the atrial actuarial curve to a cumulative survival curve of all patients at NBIMC from 1961-1976 (N=1215) indicated that after 5 years 5% more (49% to 44%) of the atrial patients had died (7). A statistical difference test has not been done at the time of this writing. However, a lower confidence limit on the atrial atient extends down to 36%, a 15% interval; the upper confidence limit here, and the lower confidence limit of at least 2-3% on the cumulative curve most surely overlap. The 3 year survival percentages are almost identical at approximately 70% surviving each.

This is compared to a non-paced survival of patients with complete heart block, who only had 10% surviving after 5 years (7). consequently, atrial patient survival is on a level comparable to ventricular patient survival and much superior to non-paced patients.

Analysis of the final outcome of the original modes (Table 18) shows 50% of all DVI implants have been lost to death compared with 17% for AAI, 5.6% for VAT, and 20% for RF. The mean date of implant for DVI was less than 5 years from the end of the study period and patients averaged 68.6 years-old at implant. This is compared to 50% of all patients that died after 5 years (Figure 2) and that the total average at implant was 65.5 years, which indicates that DVI is not greater prone to fail than the overall survival.

67% of all patients with VAT implants still have their units in; only 1 of 18 (5.6%) were lost to death, and 3 of 18 (17%) were abandoned to ventricular pacing. This is not unresonable as the mean date of implant is within the prior 3 1/2 years with an average age at implant of only 62 years-old. Figure 2 shows 68% survival at 3 1/2 years, almost exactly the same as for the VAT patients. Consequently, there does not seem to be any correlation between mode and longevity.

The possible relationship between subsequent electrode malfunction and initial electrophysiologic measurements (stimulating threshold) was evaluated. The 8 early malfunctions of the 46 Medtronic-Smyth lives averaged thresholds greater than the 34 which did not malfunction (Table 19). Data was very lacking for Cordis-J and Biotronik-Shalbach, but these too had greater stimulating thresholds.

In contrast, the 6 malfunctions (segregation into early and late malfunctions was not done here) of the American Optical T.V. electrodes was lower than the electrodes which did not malfunction. 2 of the 6 units had early malfunctions, whose thresholds were even lower (current and voltage ranged from .65 to .75 in all 4 cases).

The P-wave values for Smyth and American Optical electrodes also were greater for those electrodes which did malfunction. P-wave data for Cordis-J and Biotronikwere not available. The other electrodes: the Medtronic 1X366D, Cordis myocardial. Cordis-J, and American Optical epicardial electrodes did not have any malfunctions, so thresholds could not be compared here. Statistical tests to show differences were not done on these graphs, however, the small sample sizes indicate that threshold differences are probably not significant. However, since 5 of the 8 Smyth electrodes malfunctioned within the first week, the greater stimulating thresholds may suggest that good contact between the electrode and the endocardium was not achieved. Excessive tissue reaction would be more likely to apply to increases in threshold after 2 weeks. This may indicate that initial stimulating thresholds should fall below a set of standardized values (dependent on electrode size and shape and method of measurement) in order to decrease the probability of malfunction.

In order to completely assess the effectiveness of atrial pacing at NBIMC, a comparative analysis of electrode malfunction was done between this study and a 3-center study of ventricular implants in which NBIMC participated (8). The combined statistics for the 3-center study are: 978 original transvenous implants; 18 replacements (996 fresh insertions); and 60 repositioning (1056) electrode lives. This is a sample size that is approximately 10 times the size of the atrial population (Table 20).

The atrial malfunction rate of 18.9% is 2.55 times greater than the ventricular malfunction rate of 7.4%. The early malfunction rate (30 days or less) however, is only 3.05% greater for the atrial patients (8.65% to 5.6%). This indicates that the initial stability of the atrial electrode is comparable to that of ventricular electrodes. The low rate of late ventricular electrode malfunction (2.5%) compared to atrial (9.6%) may suggest that once initial stability has been reached, it is unlikely that a ventricular electrode will malfunction. It may also be that the greater network of trabeculae in the ventricle contributes to a greater stability. The greater percent of original ventricular malfunctions reaching final stability after one transvenous operation (80.6% to 44.4% for atrial) may be a misleading figure as it is easier to convert an atrial implant to ventricular than it is to change a transvenous ventricular implant to myocardial.

This fact is also reflected in the number of atrial abandonments which contributed to 50% of initial patient management (2.8% for ventricular) and the overall abandonment rate (9.6% for atrial and 0.7% for ventricular. Consequently, it does not seem as though abandonments or initial malfunctions reaching stability are relevant parameters in comparing atrial to ventricular pacing. More indicative of pacing effectiveness is perhaps the early 5 malfunction rate discussed previously, which showed only a slight discrepancy (and probably not statistically different).

Effectiveness of patient management can be best compared by the successfulness of repositioned electrodes; 6.7% and 5.7% of atrial and ventricular electrode lives respectively have had repositioning, 85.7% of the initial atrial repositions reached final stability compared with 78.9% of the ventricular electrodes. These figures are probably not statistically different and indicate comparable effectiveness of patient management.


Six Medtronic 1X366D electrodes implanted after the time period of the atrial study were included in the threshold data (Table 7) and relative electrode longevity (Figure 4) in order to assesss the effectiveness of this new electrode. The Medtronic 1X366D tined endocardial atrial pacing lead is a unipolar, J-shaped lead designed for sensing and/or pacing in the right atrial appendage of the heart. The distal part of each lead is shaped in a "J" configuration with an arrangement of three polyurethane tines near the tip as shown in Figure 9 (9). The tines are disigned to stabilize the lead in the right appendage by lodging in the trabeculae network (Figures 11 and 12) (10).

Prior to the development of this lead, the Medtronic-Smyth Tined Atrial "J" lead was widely used. The circumference of this lead, however, is relatively large (11) and the tines had to be trimmed prior to insertion (Figure 10). This limits its use to the larger veins such as the external jugular in many cases. The Medtronic 1X366D electrode is smaller, allowing for easier insertion and more favorable thresholds. Figure 4 shows the relative longevity of the ten 1X366D electrodes. Since the first implant in January 1977 up until August 1979, more of the electrodes have malfunctioned.

Stimulating thresholds (Table 17) were shown to be significantly different for the 1X366D electrode compared to the Medtronic-Smyth, American Optical T.V., Cordis-J electrode, except for the Cordis-J voltage value (which was greater than all the others being tested for significance but due to the small sample size it was not significant).Stimulating thresholds for the 1X366D electrodes were the lowest of all threshold means at .70mA, .623 volts, and .752 millijoules.


This comprehensive study of atrial pacing by the NBIMC has revealed much about its effectiveness and the prognosis for future progress in this area. The main points concluded on in this paper are summarized below:

1.) Age-sex distributions of atrial patients indicate that males may be greater prone to electrode malfunctions at an earlier age due to a possible sex difference in the rate of fibroplasia.

2.) Atrial patient survival is on a level comparable to ventricular patient survival, and much superior to non-paced patients.

3.) There is no correlation between atrial mode and longevity.

4.) High acute stimulating thresholds may indicate that tight apposition between the transvenous electrode and endocardium has not been achieved and lead to an early malfunction.

5.) For comparable parameters (early malfunctions and effectiveness of repositioning), atrial transvenous pacing is at the same performance level as ventricular transvenous pacing.

6.) Long-term stability tends to favor transvenous ventricular electrodes, possibly due to a difference in the trabeculae network.

7.) Due to the ease of abandoning the transvenous atrial wire to transvenous ventricular (compared to converting a transvenous ventricular wire to myocardial), the abandonment and stability rates cannot be compared between the two modes.

8.) The low malfuncton rate, ease of insertion, and low threshold values of the Medtronic 1X366D electrode may make atrial pacing, as a whole, considerably more effective (lower stimulating thresholds allow for a longer pulse generator life due to a decreased need of energy output) and has stimulated much interest in the new electrode.


For their kind efforts in assisting my collection and analysis of data, I want to express my warmest thanks to Carol Balbach, Cheryl Benjamin, Jane Cort, Jack Driller, Hugh Vangelder, Dr. George Myers, Edith Moscow, Marge Manhardt, Esther Shilling, Kay Tocci, Alex Villanueva, and Ellen Wolf; and to the Catherization laboratory techinicians, Bill, Doug, and Ralph; and to Hazel who saved me hours of typing. Special thanks to Dr. Victor Parsonnet who laid the entire framework for the paper to be built on and spent hours of his time reviweing the progress of research and his guidance on how to proceed.


1.) Parsonnet, V and Driller, J: Results of a program to develop an improved atrial pacing electrode. (unpublished) 1979.

2.) Smyth NPD: Atrial programmed pacing. PACE 1:104, 1978.

3.) Goldman, BS et. al: Permant transvenous atrial pacing. The Canadian J. of Surgery2:138, 1978.

4.) DeSanctis, RW: Diagnostic and therapeutic uses of atrial pacing. Circulation 43:748, 1971.

5.) Furman, S: Therapeutic uses of atrial pacing. American Heart Journal 86:832, 1973.

6.) Parsonnet, V: Round Table: World survey on long-term follow-up of cardiac pacing. In, Cardiac Pacing, p. 555.

7.) Parsonnet, V: Pacemaker Implantation. In, Blades Diseases of the Chest, 4th Edition, CV Mosby Co, St. Louis, p. 699, 1978.

8.) Parsonnet, V, et. al: Early malfunction of transvenous pacemaker electrodes- a three center study, 1979.

9.) Medtronic 1X366D Technical Manual.

10.) Kleinert, M: Clinical uses of a new transvenous atrial lead. American Journal of Cardiology 40:237, 1977.

11.) Medtronic 1X366D Phase One Study Protocol.