JACC Vol . 22, No . 5November 1, 199-3:132434

Scopolamine Increases Vagal T_ and Vagt

-, Z_

In PatientsAfter MYocardial Infarction,

GAETANO M. DE FERRARI, MD,*-i MASSIMO MANTICA, MD,*t EMILIO VANOLI, MD,*STEPHEN S . HULL, JIL, PHD, .t PETER J. SCHWARTZ, MD, FACC*§Milan and Pavia, Italy and Oklahoma City, Oklahoma

Objectives . The goal of this study was to assess the hypothesisthat transdermal scopolamine would increase vagal activity inpatients after myocardial infarction .

Background. In postmyocardial infarction patients, low heartrate variability and reduced baroreceptor reflex sensitivity areassociated with increased mortality. Accordingly, there is anincreasing interest in a mechanist for shifting the sympathovagalbalance toward vagal dominance

Methods. The effects of transdermal administration of scopol-amine on heart rate variability and baroreceptor reflex sensitivitywere assessed in 20 patients (mean v;ge 59 ± 11 years) bypharmacologic washout 14 ± 3 days after myocardial infarction .Heart rate variability and baroreceptor reflex sensitivity weremeasured 24 h after application of the scopolamine patch andcompared with the values measured before scopolamine and afterapplication of a placebo patch. The following variables werederived from a 15-min electrocardiographic recording : the meanRR interval and its standard deviation, the mean square succes-sive difference, the percent of intervals differing >50 ms from thepreceding RR interval and the low and high frequency areasresulting from power spectral analysis.

The potential beneficial effects of augmented vagal activityduring acute myocardial ischemia have received increasingattention in the past several years (1-4) . Both experimental(5-7) and clinical studies (8-10) indicate a strong correlationbetween indexes of either impaired vagal reflexes or reducedvagal tone and a greater incidence of sudden cardiac death .Experiments performed in a conscious canine animal modelfor sudden cardiac death (11) have indicated that a depressedbaroreflex sensitivity (5,6) and a reduced heart rate variabil-ity (7) characterize subjects at high risk for the occurrence ofventricular fibrillation . The experimental findings are closely

From the *Centro di Fisiologia Clinica e 1perten sione, Istituto di ClinicaMedica Generale e Terapia Medica, Univ ersita' degli Studi di Milano, Milan,Italy ; tDivisione di Cardiologia, Ospedale Maggiore Policlinico, IRCCSMilan, Italy ; *Department of Physiology and Biophysics, The University ofOklahoma Health Sciences Center, Oklahoma City, Oklahoma and §Diparti-mento di Medicina, Universita' degli Studi di Pavia, Pavia, Italy .

Manuscript received January 5, 1993 ; revised manuscript received April28, 1993, accepted May 5, 1993 .

Address for correspondence : Peter J . Schwartz, MD, Istituto di ClinicaMedica Generale e Terapia Medica, Via F. Sforza, 35, 20122 Milan, Italy .

©1993 by the American College of Cardiology

Results. The placebo patch had no effect on the variablesmeasured. Scopolamine increased both heart rate variability andbaroreceptor reflex sensitivity significantly . Specifically, the meanRR interval and its standard deviation increased by 7 .1% (p =0.01) and 25% (p = 0.004), respectively . The mean squaresuccessive difference increased by 38% (p = 0 .0003) and thepercent of intervals differing >50 Ins from the preceding intervalby 100% (p =- 0 .001) . The ratio of low to high frequency r ofthe power spectrum decreased by 24% (p = 0.02), an. I : :a-receptor reflex sensitivity increased by 42% (p =: 1MY TO eeffects were also evident in patients with very low initial values.Side effects were minimal .

Conclusions . Transdermal scopolamine increased measures ofheart rate variability and baroreceptor reflex sensitivity in pa-tients with a recent myocardial infarction toward values associ-ated with a better prognosis . Pharmacologic modulation of theautonomic balance by scopolamine or related drugs deservesevaluation as a new and promising approach to reduce risk aftermyocardial infarction .

(J Am Coll Cardiol 1993,22 :1327-34)

1327

paralleled by clinical studies in patients after myocardialinfarction (8-10) . These considerations, together with thenegative effect on survival caused by the administration oftraditional antiarrhythmic drugs (12,13), have given impetusto the search for a mechanism to shift the autonomic balancetoward vagal predominance as a potential strategy in theprevention of sudden cardiac death .

Low doses of atropine and other muscarinic antagonistshave been shown to produce paradoxic vagomimetic effects(14,15) . Dibner-Dunlap et al . (16) observed that transdermaladministration of the muscarinic antagonist scopolamine inhealthy young men produced an increase in the mean RRinterval and its standard deviation (SD) and in the baro-receptor response to graded neck suction . However, therelevance of these data to patients after myocardial infarc-tion is uncertain because patients with myocardial infarctionare older and show a transient impairment in vagal activity(6,17,18) . Therefore, the goal of this study was to evaluatethe effects of transdermal scopolamine on indexes of vagaltone and reflexes in patients shortly after myocardial infarc-tion. Preliminary data have been presented elsewhere (19) .

0735-1097193IS6.00

1328

DE FERRARI ET AL .SCOPOLAMINE AND VAGAL ACTIVITY

Methods

The study was performed in 20 randomly selected pa-tients who gave informed consent to the study (one addi-tional patient declined to take part in the study, and one didnot perform the study because of high blood pressure on theday of the first test). Two subjects were women, and 18 weremen (mean age 59 ± 11 years, range 37 to 78 ; mean leftventricular ejection fraction 52 ± 12%, range 29% to 68%0) .The infarction was anterior in 8 patients and inferior in 12 .

Exclusion criteria were age 80 years ; atrial fibrillationor frequent atrial or ventricular arrhythmia that made anal-ysis of sinus cycles impossible ; atrioventricular block ; his-tory of insulin-dependent d',~ °tes mellitus and signs orhistory of central or periphei .. nellropathy ; blood pressure>160/100 mm Hg with diuretic gents; and serious, concom-itant noncardiovascular disease, such as renal or hepaticfailure, Patients were also excluded if they had postinfarc-tion angina, were directed tc card urgent mechanical orsurgical revascutarization or I . i heart failure unresponsiveto the sole administration of t : : iretic agents . Patients wereenrolled 8 to 15 days after myocardial infarction and werestudied during pharmacologic washout . Beta-adrenergicblocking agents ; calcium channel antagonists ; vasodilators ;antiarrhythmic, antidepressant and antihypertensive agents ;nitrates, and any drug that could have interfered with theautonomic assessment were withdrawn at least 5 half-livesbefore the beginning of the study . No patient was receivingdigitalis . Diuretic agents were allowed, if needed, to controlblood pressure and pulmonary congestion . Platelet-activeand noncardiovascular drugs were also allowed and kept atthe same dosage throughout the study .

The protocol consisted of the assessment of heart ratevariability and baroreceptor reflex sensitivity in controlconditions, 24 h after the application of one patch of scopol-amine (Transcop, Recordati) to the skin behind one car and24 h after removal of the active compound patch and itsreplacement with a placebo patch . The patients had noknowledge of the treatment given . The active patch con-tained 1 .5 mg of scopolamine and was designed for contin-uous release of 0.5 mg every 24 h. Although most of the testsin each patient were performed on 3 consecutive days,occasionally > 1 day elapsed between two tests for logisticreasons. However, the patch was always applied 24 h beforethe evaluation. The test with scopolamine was performed14 ± 3 days (range 10 to 18) after the occurrence ofmyocardial infarction.

The test was always performed at the same time of day,3 h after a light meal. Throughout the study, smoking andeine-containing beverages were not allowed. The pa-

tients were studied while supine at rest in a quiet, tempera-ture-controlled room, and they were asked to relax and toavoid talking and sleeping . The recording was started after10 to 15 min of acclimation and lasted 15 min . Heart ratevariability was assessed with a commercially available soft-ware (Predictor, Corazonix). Analog electrocardiographic

JACC Vol . 22, No. 5November 1 . 1993 :1327-34

(ECG) signal was digitized at 500 Hz . The QRS complex wasrecognized by cross correlation with the template of thenormal QRS complex chosen by the investigator . Prematureatrial and ventricular beats and the subsequent interval wereexcluded automatically by the analysis, and the automaticdetection was also visually checked by the investigator forproper identification. Time domain variables considered inthis study were the mean RR interval and its SD, the meansquared successive difference and the percent of intervalsdiffering >50 ms from the preceding interval. Power spectralanalysis was performed by an autoregressive algorithm(Predictor, Corazonix) with a direct current filter and modelorder 16 . The boundaries of the low frequency (average 0 .05to 0.16 Hz, with a peak at 0 .10 Hz) and high frequency(average 0 .18 to 0 .42 Hz, with a peak at 0.27 Hz) regionswere chosen manually for each test by an investigatorunaware of the treatment, and the area was automaticallyobtained by the software . A very low frequency peak wasgenerally not revealed by the analysis. A pilot study of thereproducibility of the time domain measures (the SD of themean RR interval, the mean square successive differenceand the percent of intervals differing >50 ms from thepreceding interval) among two recordings t :-:90 min fromeach other, performed with this protocol in 10 patients,shows a correlation (r value) from 0.974 to 0.982 .

Assessment of baroreceptor reflex sensitivity was per-formed shortly after, according to the method of Smyth et al .(20). Blood pressure was continuously monitored noninva-sively by infrared digital plethysmography (Finapres 2300,Ohmeda). `This signal and a lead 11 ECG signal were digitizedand fed into a Compaq 286 computer with custom-madesoftware (G . Plnna, PhD, Montescano, Italy) for analysis ofbaroreceptor reflexes . The alpha,-agonist phenylephrine wasinjected (initial dose 200 fag) for 30 s to obtain an increase inblood pressure of 20 to 30 mm Hg . Baroreceptor reflexsensitivity was calculated as the slope of the linear regres-sion line relating systolic blood pressure changes to RRinterval changes . Regression lines with >_20 points and anr ~ 0.70 were accepted for analysis . The phenylephrineinjection was repeated until three linear regressions withthese characteristics were obtained (in most cases this wasachieved with the first three injections). The average of thesethree slopes was then taken as the baroreceptor reflexsensitivity value . The average duration of the whole testsession was 80 min (range 70 to 100) .

Statistical analysis. Results are expressed, unless other-wise stated, as mean value ± SD. Regression lines forbaroreceptor reflex sensitivity were calculated with theleast-squares method . Normality of the distributions wasassessed both graphically and with a goodness of fit test . Theoverall difference among mean values in the three groups(control, scopolamine, placebo) was evaluated with one-wayanalysis of variance for repeated measures or with aKruskal-Wallis test for nonnormal distributions (percent ofintervals differing >50 ms from the preceding interval andpower spectral densities). A subsequent direct comparison

.ACC Vol . 22, No. 5November I, 1993 :1327-34

0-1

CONTROL

CONTROL

SCOPOLAMINE

SCOPOLAMINE

PLACEBO

PLACEBO

was made, respectively, with a t test for paired samples orwith a Wilcoxon signed rank test . A p value < 0 .05 wasconsidered significant, with the exception of the multiplecomparisons among the three treatments, where the limit forsignificance was set at p < 0 .017, according to the Bonfer-roni correction .

Results

The scopolamine patch was well tolerated by the patients .The sole side effects were dryness of mouth in one patientand light-headedness in another patient . The latter patientalso reported a similar effect during the subsequent placebopatch application. No other side effect was reported .

The overall technical quality of the tests was good . Noheart rate variability recording had to be discarded becauseof excessive premature beats . The average correlation coef-ficient of the baroreceptor slopes was 0 .85 t 0.08. Threepatients had on?, two acceptable regression slopes in one ofthe two contre,l conditions. No major side effect was ob-served after phenylephrine injection . Two patients reporteda mild transient flushing ; one had modest horripilation ; and afew perceived the slowing of heart rate .

Heart rate vari?,Wfity. Analysis of variance for time do-main variables derived from the 15-min recording of RRintervals revealed significant differences in all of the vari-ables (Fig. 1). The placebo patch did not induce any signif-icant change in any of the variables .

Mean heart period (mean RR interval) during the scopol-

16

14

12

DE FERRARI ET AL .SCOPOLAMINE AND VAGAL ACTIVITY

TIA

CONTROL

CONTROL

SCOPOLAMINE

SCOPOLAMINE

PLACEBO

PLACEBO

1329

Figure 1 . Effect of scopolamine and placebo on the mean sinus cycle(RR), its standard deviation (StD), the mean squared successivedifference interval (MSSD) and the percent of sinus cycles differingfrom the preceding cycle by >50 ms (PNN50) . Bars represent meanvalue ± SEM for each variable . *Analysis of variance for repeatedmeasures for RR (p < 0 .002), SO (p < 0 .001) and MSSD (p < 0.001)and the Kruskal-Wallis test for pNN50 (p < 0 .001) .

amine test was greater compared with both control (average7.1%, p = 0.01) and placebo conditions (average 9 .9%, p =0.0001). This effect on the RR interval corresponds to anaverage decrease of 5 to 6 beats/min in heart rate . The SD ofthe mean RR interval increased correspondingly more . Dur-ing the scopolamine test, it was 25% greater than during thecontrol period (p = 0 .004) and 36% greater than during theplacebo period (p = 0.003) . This greater increase in the SD ofthe mean RR interval, compared with that of the mean RRinterval, indicates that the increase in variabil ;ty was linkedonly to a minor extent to the increase in mean heart cycle .The mean squared successive difference among heart cyclesincreased significantly during the scopolamine test comparedwith both the control (by 38%, p = 0 .0003) and placeboconditions (by 61%, p = 0 .0003). The percent of intervalsamong sinus beats differing >50 ms from the previousinterval was significantly greater in the scopolamine test thanin the control or placebo trial (p = 0 .001 and p < 0.001,respectively, by the Wilcoxon signed rank test) . Figure 2 isan example of the heart period histogram of a patient duringcontrol conditions ard after scopolamine .

1330 DE FERRARI ET AL .SCOPOLAMINE AND VAGAL ACTIVITY

00 :15 :00lilts : 915Wpt: 974Short 740.0Long 1115.0

5600 STATS

it . 924.0StO: 70.51SS0 :

10.0

64 .952 .2

F07 7 .!WA NN. 4 .

970.065.21

915 .065.6

i"Na..

10

Analysis of the frequency domain variables (power spec-trum density of the different areas) disclosed no overalldifference among the three groups in the total (0 to 0 .5 Hz)spectral density or in the tow frequency component . A trendtoward a difference was noted for the high frequency com-ponent (p = 0.06). The ratio of low and high frequencycomponents showed a significant difference among the threegroups (p < 0,05) . The placebo conditions had no effect onthis variable, whereas scopolamine induced a decrease thatwas of borderline significance compared with the valueduring control conditions (24%, p = 0.02) and was clearlysignificant compared with that during placebo conditions(p = 0.009) (Fig. 3). There was a moderate correlationbetween measures of heart rate variability during controlconditions and after the scopolamine test (for the SD of themean RR interval, r = 0 .59, p = 0.005).

HRV-ECG7L ANALYZE_____!

00 :15 :00ants :

995.Pt : 981Short 746 .0bang 1025 .0maxrrq :

98RN STATS

Mean : 907 .5StD :

43.3SO:

40.2PON: 19.3%

"AoYr :

4 .8WANN :

16,940 .0

A .

63.8

900 .0".7

!nft 11

Figure 3. Effect of scopolamine and placebo treatments on the ratioof the areas in the low and high frequency bands of the powerspectrum (LFIHF) . Bars represent mean value ± SEM . *p < 0.05,analysis of variance for repeated measures .

M .

CONTROL

Figure 2. Example of frequency (FREQ) histogram andrelated time domain measures provided by the studysoftware in one patient during control conditions (upperpanel) and scopolamine treatment (lower panel) . In thispatient there was a very modest effect on the mean RRinterval (Mean) but a marked effect on the standarddeviation (StD) (1-63% the mean squared successivedifference (MSSD) (+74%) and the percent of sinuscycles differing from the preceding cycle by >50 ms(PNNSO) (+170%%). HR = heart rate . The software alsoprovides the total number of intervals that occurredduring the recording (Ints) ; the number of those acceptedfor the analysis (Accpt), the duration of the shortest(Short) and longest (Long) interval ; the frequency ofoccurrence of the most common class of beats (MaxFrq),according to the selected bin width (BinWd, in theexample IC ms) ; and the following statistical variablesobtained from the RR intervals (1111 STATS): the mode

and median (Mod) of the distribution ; the coefficient ofvariance (CoVr) and the standard deviation of the 5-minaverage normal to normal (NN) intervals (SDANN) .

SCOPOLAMINE

JACC Vol . 22, No . 5November 1 . 1993 :1327-34

PLACEBO

JACC Vol . 22, No . 5November I, 1993 :1327-34

CONTROL

SCOPOLAMINE

Figure 4 . Effect of scopolamine and placebo treatments on thebaroreceptor reflex sensitivity (BRS) . Bars represent mean valueSEM . *p < 0 .001, analysis of variance for repeated measures .

Baroreceptor reflex sensitivity . Analysis of variance forthe baroreceptor reflex sensitivity disclosed I significantdifference among the three groups (p < 0 .001) . Similar to theother variables analyzed, no difference was found betweencontrol and placebo conditions (p = 0 .34). Scopolamineincreased baroreceptor reflex sensitivity by 42% comparedwith control values (from 8 .1 ± 4.6 to 11 .5 ± 6.4 ms/mm Hg,p = 0.0006). Baroreceptor reflex sensitivity was also signif-icantly greater in the scopolamine test than during placeboconditions (p = 0 .01) (Fig. 4). Figure 5 shows the goodcorrelation (r = 0 .84, p < 0 .001) between baroreceptor reflexsensitivity at baseline and after scopolamine. Figure 5 (right)identifies the patients with increased, and those few withdecreased, baroreceptor reflex sensitivity . Figure 6 is anexample of the linear regression of two baroreceptor reflexsensitivity tests in one patient during control conditions andafter scopolamine .

Correlation of all of the variables studied with the age ofthe patients was evaluated . This disclosed a moderate neg-ative correlation between age and baroreceptor rcflcx sensi-tivity (r = 0 .63, p < 0.005) and very modest negativecorrelations between age and the SD of the mean RR

E

0

0

Figure 5 . Correlation between thevalues of baroreceptor reflex sensi-tivity (BRS) observed during controlconditions and scopolamine treat-ment. Left, Linear regression showsgood correlation . Right, Solid circlesto the left and right of the identity liney = x indicate itients whose barore-ceptor reflex sensitivity increased 5~and decreased, respectively .

10 1

0

PLACEBO

y=2.10+1 .09 x, r=0.84

5

10

15

20

25control ORS (msec/mm Hg)

DE FERRARI ET AL .

1331SCOPOLAMINE AND VAGAL ACTIVITY

interval, the mean square successive difference and thepercent of intervals differing >50 ms from the precedinginterval (r = -0 .49 to -0.42, p = 0 .025 to 0 .06). No othersignificant correlation was found . No correlation was foundbetween ejection fraction and the measures of vagal tone andreflexes . No difference was also found during any studyconditions between patients with inferior versus anteriormyocardial infarction . However, it is possible that the ab-sence of difference may be due to the relatively small samplesize of this study .

In the overall group of patients, scopolamine did notcause significant changes in blood pressure (126 ± 17 vs .132 ± 18 mm Hg during control conditions for systolic bloodpressure). The standard 12-lead ECG disclosed no significantdifference in the PR interval (172 ± 28 ms during controlconditions and 178 ± 24 ms during the scopolamine test) orin QT interval corrected (QTc) for heart rate (QTc 420 ± 34before and 412 ± 30 during the scopolamine test) . In onepatient, Wenckebach second-degree atrioventricular (AV)block occurred during phenylephrine injection in two of fourscopolamine tests but in none of the control injections .

An echocardiographic evaluation was performed in ninepatients during both placebo and scopolamine administra-tion, and no difference was found in either dimension orfunction of the left ventricle (ejection fraction was 54 ± 12%during control conditions and 53 ± 10% during scopolaminetreatment) .

DiscussionThis study indicates that transdermal administration of

scopolamine significantly increases measures of heart ratevariability and baroreceptor reflex sensitivity in patientswith a recent myocardial infarction . These markers of vagaltone and reflexes are shifted toward values associated with alower mortality in previous clinical studies (8-10). Pharma-cologic modulation of the autonomic nervous system maythus result in a shift of the sympathovagal balance toward a

1332

DE FERRARI ET AL .SCOPOLAMINE AND VAGAL ACTIVITY

1400E

1300

1200

1300

12001

1100dI

1000 ~

CONTROL

y n 103 + 6,62 X , to 0,96

120

130

140

150

160

170

100

190mmHg

6. Example of baroreceptor reflex sensitivity in one patientduring control conditions and after scopolamine treatment. Sensi-tivity is expressed by the slope of the linear regression betweensystolic blood pressure (absdm) and the subsequent RR interval(ardhate) . In this patient, a marked increase in baroreceptor reflexsensitivity was observed . The blood pressure increase during thetwo tests was significantly greater than the average increase (20 to30 mm ft but, as easily observed, the response was still linear.

vagal dominance and may deserve evaluation as a newapproach in the prevention of sudden death after myocardialinfarction.

BackgrawW. Scopolamine has been shown to have vag-omimetic effects in healthy young volunteers (16) . However,the possibility of a similar effect in patients after myocardialinfarction was uncertain because 1) the age range of thesubjects in the study by Dibner-Dunlap et al . (16) (22 to 34years, mean age 27) was completely different from that ofpatients with myocardial infarction . Age exerts a consider-able influence on markers of vagal activity and specifically

JACC Vol. 22 . No . 5November 1 . 1993 :1327-34

on baroreceptor sensitivity (8,9,21) ; 2) myocardial infarctioncreates a derangement of the autonomic balance, resulting inan impairment of the parasympathetic component (6,17,18) .Thus, the observation in healthy young adults cannot nec-essarily predict the results in older subjects with myocardialinfarction . The finding that scopolamine may also increaseheart rate variability and baroreceptor reflex sensitivity inpatients shortly after myocardial infarction may have clinicalimplications because higher values of both markers arecorrelated with a better prognosis (8-10) .

Baseline findings . The baroreceptor reflex sensitivity inthis group of patients was very similar to that in a previousstudy by our group (8) . Similar to that study, we found nocorrelation between baroreceptor reflex sensitivity and ejec-tion fraction, and a moderate negative correlation betweenage and baroreceptor reflex sensitivity . A modest correlationwith age was also found for heart rate variability indexes .The absence of a stronger correlation was expected becausethe presence of myocardial infarction is likely to perturb therelation between age and vagal effects .

Effects of scopolamine. The mean RR interval increasedafter scopolamine administration by a value comparable tothat (13%) observed in healthy young men by Dibner-Dunlapet al. (16) . The increase in SD was also similar to that in thelatter study (31%) despite the great difference in baselinevalues (35 vs . 105 ms) . The findings are strengthened in thisstudy by the marked and significant effect of scopolamine ontwo time domain variables considered as indexes of tonicvagal activity to the heart (22) : the mean square successivedifference and the percent of intervals differing >50 ms fromthe preceding interval . Among the patients with a SD of themean RR interval below the mean value of the whole group(n = 10), o of 10 increased their heart rate variability, oftenconsiderably. Therefore, the effect of scopolamine appearsto be particularly marked among patients considered to be athigher risk .

Frequency domain analysis of heart rate provided afurther means to investigate the autonomic regulation of theheart . The ratio of low frequency to high frequency spectraldensities has been proposed as an index of sympathovagalbalance (23) . The decrease in this ratio, observed in thescopolamine test, indicates that the autonomic balance wasmodified toward vagal predominance.

Vagal reflexes were also significantly increased 24 h afterapplication of the scopolamine patch . Baroreceptor reflexsensitivity increased by 422% compared with control values .More specifically, eight patients with a baseline baroreceptorreflex sensitivity <6 ms/mm Hg showed an average increasein sensitivity of 53% ; five patients with a baseline baro-receptor reflex sensitivity <4 ms/mm Hg showed an averageincrease of 74%; and three patients with a baseline baro-receptor re6exx sensitivity <3 ms/mm Hg showed an averageincrease of 790/6 . This is an important point in the studybecause it indicates that the effects of scopolamine were alsoclearly present in patients considered to be at very high risk(8,9) (i .e ., baroreceptor reflex sensitivity <3 ms/mm Hg) .

JACC Vol . 22, No . 5November 1, 1993 :1327-34

Mechanisms. The first observation that low doses ofatropine cause a reduction in sinus rate was made as early as1915 (14) . However, both central (24,25) and peripheral (26)sites of action have been suggested, and a precise explana-tion for this paradoxic vagomimetic effect is still lacking (27) .Relevant here is the fact that, compared with most othermuscarinic antagonists, scopolamine has a greater centraleffect (28) .

Characteristics and limitations of the study . Analysis ofheart rate variability was performed with relatively shortperiods of recording so that the assessment could be coupledwith that of baroreceptor reflex sensitivity . The result was atotal duration of the test of <90 min that would not cause thepatient discomfort and thereby alter the results . This rela-tively short duration makes a study of this type feasible as aclinical tool, at least for selected patients, We decided not tobase our study on repeated Holter recordings because thiswould have doubled the duration of the protocol owing tothe lag time for scopolamine absorption and the half-life ofthe drug (16) . Moreover, it was possible to strictly controlthe recording cots -litions, including recording time, roomtemperature and isition and activity of the patients,whereas these coi ., ions may vary widely during repeatedHotter recordings and have a profound influence on heartrate variability . Finally, stationary signals, such as thoseobtained during controlled recording in this study, are bettersuited for both time and frequency domain analysis (29) .

For of baroreceptor reflex sensitivity weutilized noninvasive measurements with the digital plethys-mographic technique . This technique has been validated forcontinuous beat to beat blood pressure recording (30). Apreliminary analysis of the baroreceptor reflex sensitivityvalues obtained with intraarterial methods and with Finapresin 376 patients after myocardial infarction, as part of themulticenter ATRAMI (Autonomic Tone and Reflexes AfterMyocardial Infarction) study, shows very good correlation(r = 0 .92 [311). On the basis of these results, the use ofnoninvasive assessment of baroreceptor reflex sensitivitywas preferred, also because the measurements were re-peated three times. Noninvasive blood pressure measure-ment is clearly a less stressful condition for the patientcompared with intraarterial recording .

The absence of any significant difference between thecontrol and placebo conditions indicates that 1) the effectwas indeed due to scopolamine ; 2) no relevant spontaneouschange in vagal activity occurred during the period of thestudy; 3) the effect was reversible ; and 4) the placebo patchhad no effect on the assessed variables. In addition, thepresent results are in agreement with those of two similarstudies performed at the same time by two other groups(32,33) .

This study enrolled unselected patients after myocardialinfarction. As a consequence, several patients had goodventricular function and would thus be considered at lowrisk by the traditional risk stratification criteria . However,autonomic variables, notably baroreceptor reflex sensitivity,

DE FERRARI ET AL .

1333SCOPOLAMINE AND VAGAL ACTIVITY

have no correlation with ejection fraction, and they appearcapable of identifying patients at very high risk also amongthe subgroup with a very good ejection fraction (8,9) . Wedecided to study patients without drugs for this preliminarystudy, to avoid potentially confounding agents . However, itwill be useful to confirm these effects in patients taking theirroutine drug treatment .

Potential clinical implications . Increasing evidence sug-gests that vagomimetic interventions may decrease the like-lihood of malignant arrhythmia in the setting of acute myo-cardial ischemia (1-4) . Vagal nerve stimulation markedlydecreased the incidence of ventricular fibrillation (34) in aconscious canine model for sudden cardiac death that com-bined a previous myocardial infarction, elevated sympa-thetic activity and transient myocardial ischemia (11). In thesank experimental model, the muscarinic agonists rnetha-choline and oxotremorine provided incomplete but signifi-cant protection against sudden death (35) . These drugs, atvariance with the beta-blocker propranolol, were completelydevoid of negative inotropic effects (35) . Oxotremorine alsomarkedly decreased the incidence of life-threatening ventric-ular arrhythmias induced in a feline model by the combina-tion of acute myocardial ischemia and sympathetic hyper-activity (36) . Thus, vagomimetic agents may represent aninteresting approach, particularly when beta-blockers arecontraindicated and when preservation of inotropism isessential .

However, the potential clinical application of muscarinicagonists must await the availability of compounds thatdisplay acceptable selectivity for cardiac muscarinic (M 2 )receptors (27) . Pharmacologic modulation of neural activitywith scopolamine may effectively increase vagal activitywith limited side effects and may thus deserve furtherevaluation in the search for new strategies in the preventionof sudden cardiac death after myocardial infarction .

We are grateful to Pinuccia De Tomasi for careful attention to the preparationof the manuscript and to Riccardo Sarri for expert photographic support .

References

I. Verrier RL . Neurochemical approaches to the prevention of ventricularfibrillation . Fed Proc ;9,86 ;45 :219 11-6 .

2. Corr PB, Yamada KA, Witkowski FX . Mechanisms controlling cardiacautonomic function and their relation to arrhythmogenesis . In : FozzardHA, Haber E, Jennings RB, Katz AM, Morgan HE, editors . The Heartand Cardiovascular System . Vol . 11, New York: Raven, 1986:1343-403 .

3 . De Ferrari GM, Schwartz PJ . Autonomic nervous system and arrhyth-mias . Ann NY Acad Sci 1990,601 :247-62 .

4 . De Ferrari GM, Vanoli E, Schwartz PJ . Vagal activity and ventricularfibrillation . In : Levy MN and Schwartz PJ, editors . Vagal Control of theHeart . Mount Kisco, NY : Futura . In press .

5 . Billman GE, Schwartz PJ, Stone HL . Baroreceptor reflex control of heartrate : a predictor of sudden cardiac death, Circulation 1982 ,,66 :874-80,

6 . Schwartz PJ, Vanoli E, Stramba-Badiale M, De Ferrari GM, Billman GE,Foreman RD. Autonomic mechanisms and sudden death . New insightfrom the analysis of baroreceptor reflexes in conscious dogs with andwithout a myocardial infarction. Circulation 1988;78:%9-73.

1334

DE FERRARI FT AL .SCOPOLAMINE AND VAGAL ACTIVITY

7. Hull SS, Evans AR, Vanoli E, et al . Heart rate variability before and aftermyocardial infarction in conscious dogs at high and low risk of suddendeath . J Am Coll Cardiol 1990 ;16 :978-85 .

8, La Rovere MT, Specchia G, Mortara A, Schwartz PJ . Baroreflex sensi-tivity, clinical correlates and cardiovascular mortality among patientswith a first myocardial infarction : a prospective study. Circulation 1988 ;78:816-24,

9. Farrell'TG, Odemuyiwa 0, Bashir Y, et al . Prognostic value of baroreflexsensitivity testing after acute myocardial infarction . Br Heart J 1992 ;67 :129-37.

10 . Kleiger RE, Miller JP, Bigger JT Jr, Moss Al, and the MulticenterPost-Infarction Research Group . Decreased heart rate variability and itsassociation with increased mortality after acute myocardial infarction .Am J Cardiol 1987 ;59 :256-62.

11, Schwartz PJ, Hillman GE, Stone HL . Autonomic mechanisms in ventric-ular fibrillation induced by myocardial ischemia during exercise in dogswith a healed myocardial infarction . An experimental preparation forsudden cardiac death . Circulation 1984 ;69:780-90.

12 . Akhtar M, Breithardt G, Camm AL et al . CAST and beyond, Implicationsof the Cardiac Arrhythmia Suppression Trial . Circulation 1990 ;81 : 11234,

13 . Yusuf S, Too KK . Approaches to prevention of sudden death : need forfundamental reevaluation. J Cardiovasc Electrophysiol 1991 ;2(suppl):S233-9 .

t4 . Wilson FN . The production of atrioventricular rhythm in man after theadministration of atropine . Arch Intern Med 1915 ;16 :989-1007 .

15, Das G, Cardiac effects of atropine in man : an update. Int J Clin Pharmacol1989 ;27 :473-7.

I6, Dibner.Dttnlap ME, Eckberg DL, Magid NM, Cintron-Trevino NM . Thelong-term increase of baseline and reflexly augmented levels of humanvagal •c ardiac activity induced by scopolamine . Circulation 1985 :71 :797-804.

17 . Schwartz PJ, Zaza A, Pala M, Locati E, Beria G, Zanchetti A . Baroreflexsensitivity and its evolution during the first year after a myocardialinfarction. J Ain Coll Cardiol 1988 ;12 :629-36.

18, Lombardi F, Sandrone G, Pernpruner S, et al . Heart rate variability as anindex of sympathovagal interaction after acute myocardial infarction . AmJ Cardiol 1987 ;60:1239-45 .

19. De Ferrari GM, Mantica M, Curcuruto P, Vanoli E, Schwartz PJ,Transdermal scopolamine modifies autonomic balance after myocardialinfarction (abstract) . Circulation 1991 ;84(suppl lI):II4266,

20. Smyth HS, Sleight P, Pickering GW . Reflex regulation of arterial pressureduring sleep in man . Circ Res 1969 ;24 :109-21 .

21 . Garibbin B . Pickering TG, Sleight P, Peto R . Effect of age and high bloodpressure on baroreflex sensitivity in man, Circ Res 1971 ,,29 :424-31 .

22. Bigger TJ, Jr, Schwartz Pi . V control and risk stratification . In : LevyMD, Schwartz PJ, editors, Vagal Control of the Heart . Mount Kisco, NY :Futura, 1993 . In press,

JACC Vol. 22, No. 5November 1, 1993 :1327-34

23. Malliani A, Pagani M, Lombardi F . Cerutti S . Cardiovascular neuralregulation explored in the frequency domain . Circulation 1991 ;84 :482-92 .

24. Katona PG, Lipson D, Dauchot PJ. Opposing central and peripheraleffects of atropine on parasympathetic cardiac control. Am J Physiol1977 ;232 : H 146-H51 .

25. Gilbey MP, Jordan D, Richter DW, Spyer KM. Synaptic mechanismsinvolved in the inspiratory modulation of vagal cardio-inhibitory neuronesin the cat . J Physiol 1984;356:65-78.

26 . Kottmeier CA, Gravenstein JS . The parasympathomimetic activity ofatropine and atropine methylbromide. Anesthesiology 1968 :29:1125-33 .

27 . De Ferrari GM, Zaza A . Drugs acting on cholinergic receptors . In : SinghBN, Dzau UJ, Van Houtten P, Woosley RL, editors . CardiovascularPharmacology and Therapeutics. New York- Churchill Livingstone . Inpress .

28 . Heller Brown J . Atropine, scopolamine, and related antimuscarinic drugs .In : Goodman Gilman A, Rall TW, Nies AS, Taylor P, editors . ThePharmacological Basis of Therapeutics . New York: Pergamon Press,1990 :150-65 .

29. Orsi G, Pinna G . Effect of record length on the power spectral densityestimation, In : Cerutti S, Minuco G, editors . Spectral Analysis and HeartRate Variability Signal . Pavia: La Goliardica Pavese, 1991 :45.58,

30. Parati G, Casadei R, Groppelli A, et al. Comparison of finger andintro-arterial blood pressure monitoring at rest and during laboratorytesting. Hypertension 1989 ;13 :647-53 .

31 . La Rovere MT . Mortara A, Bigger JT Jr. et al . for the ATRAMIInvestigators. Reliability of non-invasive assessment of baroreflex sensi-tivity after myocardial infarction . ATRAMI, an international multicenterprospective study (abstract). Eur Heart J 1993 ;14(suppl):253 .

32 . Casadei B, Pipilis A, Conway J, Sleight P . Effect of vagal stimulation bytransdermal scopolamine in patients post myocardial infarction (abstract) .Circulation 1991 ;84(suppl 11):11-555 .

33 . Vybiral T, Glaeser DH, Morris G, et al . Effects of low dose transdermalscopolamine on heart rate variability in acute myocardial infarction . J AmColl Cardiol 1993 :22:1320-6.

34 . Vanoli E, De Ferrari GM, Stramba-Badiale M, Hull SS Jr, Foreman RD,Schwartz PJ. Vagal stimulation and prevention of sudden death inconscious dogs with a healed myocardial infarction, Circ Res 1991 ;68:1471--81 .

35 . De Ferrari GM, Salvati P, Grossoni M, et al . Pharmacologic modulationof the autonomic nervous system in the prevention of sudden cardiacdeath. A study with propranolol, methacholine and oxotremorine inconscious dogs with a healed myocardial infarction . J Am Coo Cardiol1993 ;22:283-90.

36, De Ferrari GM, Vanoli E, Curcuruto P, Tommasini G, Schwartz PJ .Prevention of life-threatening arrhythmias by pharmacologic stimulationof the muscarinic receptors with oxotremorine . Am Heart J 1992 ;124 :883-90 .