journal, paradoxical electrocardiographic effects of amyl

British Heart Journal, I972, 34, 85 I-857.

Paradoxical electrocardiographic effects of amylnitrite in coronary artery disease

zRichard E. Kerber and Donald C. HarrisonFrom the Department of Industrial Medicine, University ofIowa Medical Center, Iowa City, Iowa;and Cardiology Division, Stanford University School of Medicine, Stanford, California 94305,U.S.A.

The electrocardiographic response to amyl nitrite inhalation was correlated with treadmill stresstesting in 6i patients. Eleven patients developed ST depressions suggesting myocardial ischaemiaafter amyl nitrite inhalation on a total of I7 different occasions. All had positive treadmill tests;8 had coronary artery disease seen on angiocardiography; I had aortic stenosis, I had lupus, andi had angina with normal coronary arteries. This group was compared with i5 patients who hadangiographically proven coronary artery disease and positive treadmill tests but negative amyl

nitrite tests: the group with positive amyl nitrite tests responded to drug inhalation with a signifi-cantly greater heart rate increase (47% vs. 34%, P< o.oI), and a significantly greater fall insystolic blood pressure (42% vs. 31%, P< o'o5) and diastolic blood pressure (42% vs. 29%,P< o os). Atrial pacing studies in 4 patients suggested that the amyl nitrite induced ischaemicchanges were not due to tachycardia alone. No patient with a negative treadmill test had a

,positive amyl nitrite test; positive amyl nitrite tests became negative after revascularization sur-

gery, propranolol therapy, or clinical improvement. The paradox of ischaemic electrocardio-graphic changes caused by amyl nitrite is probably explained by the observed acute reduction ofcoronary perfusion pressure and abbreviation of diastole. Amyl nitrite may have clinical appli-cation in the diagnostic and therapeutic evaluation of coronary artery disease.

Amyl nitrite has long been recognized to bean effective agent in the treatment of anginapectoris (Brunton, I867). Its effects on thephonocardiogram and its haemodynamicactions have been well studied and provideuseful diagnostic information in a variety ofcardiac lesions (Beck et al., I96I; Cohn,Flamm, and Hancock, I968; Perloff et al.,I963). Electrocardiographic alterations of

,-several types after amyl nitrite inhalationhave been described (Goldberger, 1945; My-burgh, I969), and it has been suggested thatthe appearance of these changes may indicatethe presence of coronary heart disease (Con-tro, Haring, and Goldstein, I952; Varela deSeijas Aguilar et al., I953; Rizzon, Lovreglio,,and Bacca, I963). These studies were under-taken to evaluate further and attempt to ex-plain the electrocardiographic actions of amylnitrite, and to correlate these effects withtreadmill testing and coronary angiographyin the same patients.

Received 6 December 1971.

Subjects and methodsSixty-one adult patients were given amyl nitritea total of 75 times. The initial 30 amyl nitritetests were performed on any patients undergoingtreadmill testing who would consent to the amylnitrite inhalation. The remaining 45 tests wereperformed on subjects who had already beenshown to have positive treadmill tests. No attemptwas made to separate or select patients by age,sex, clinical condition, or other criteria.

After instruction and consent, the patient tookfour deep breaths without the drug, while anelectrocardiogram (leads II and Vs simultane-ously) was being recorded continuously; recordingwas continued for at least 2 minutes. This con-stituted a control period and evaluated the effectof hyperventilation alone. An amyl nitrite pearlwas then crushed and held under the patient'snose while the inhalations were repeated andsimilar electrocardiographic recordings made.Blood pressures obtained by the cuff methodwere recorded before inhalation, and at themaximal amyl nitrite effect (about 30 seconds).The test was performed in the erect position inall patients, and repeated in the supine positionin half the cases. Similar blood pressure and

on April 14, 2022 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.34.8.851 on 1 August 1972. D

ownloaded from

http://heart.bmj.com/

852 Kerber and Harrison

electrocardiographic observations were madeafter sublingual nitroglycerin administration in IOstanding patients, who had previously undergoneamyl nitrite inhalation.Maximal treadmill stress testing, using a modi-

fication of the multistage method of Bruce et al.(I963) was performed at least once on all patients.Most patients underwent this test on the sameday as the amyl nitrite study; the two tests wereseparated by at least I5 minutes, and longer ifnecessary, to allow electrocardiograms to returnto the control condition. The electrocardiographicrecordings from the two tests were interpretedby independent readers; horizontal ST segmentdepression of at least i mm, which persisted for atleast o-o8 sec was defined as 'positive' (indicatingmyocardial ischaemia) for both tests (Master andRosenfeld, I967). The amyl nitrite test was com-pared with both control and hyperventilationtracings when being interpreted. 'J'-point de-pression was classified separately. Patients receiv-ing digitalis were excluded (Kawai and Hultgren,I964).Four patients who had previously undergone

treadmill and amyl nitrite testing were studiedusing atrial pacing during the course of cardiaccatheterization. A Goetz bipolar pacing catheterwas placed at the junction of the superior venacava and right atrium, and pacing at rates up toI50 impulses a minute accomplished whileexternal electrocardiograms (leads II and Vssimultaneously) were recorded.

Coronary angiography was performed on 30 ofthese patients. Significant coronary artery diseasewas defined by us as at least one lesion in a maincoronary artery, resulting in 50 per cent or greaternarrowing of the arterial lumen. Most patientshad much more severe disease than this. In 2patients who died before angiography, necropsyevaluation of the coronary arteries was available.

ResultsNineteen patients had negative treadmilltests. Nine of these patients underwent diag-

FIG. I Electrocardiographic effects of amylnitrite in a 40-year-old woman with coronaryartery disease. Ischaemic ST depressions arepresent on both amyl nitrite and treadmilltesting. HV= hyperventilation, AN= amylnitrite, TM= treadmill, CAD = coronaryartery disease.

CONTROL HV AN TM

V5 191

CONTROL H V ANI

V5

J.F: (CAD)

F I G. 2 Electrocardiographic effects of amylnitrite in a 42-year-old woman with coronaryartery disease. The initial amyl nitrite (ANi)test and treadmill test show ischaemic STdepressions. The second amyl nitrite test(AN2), performed 6 weeks later when thepatient's angina was worsening, shows moresevere ST segment depression than the firstamyl nitrite test; a second treadmill test wasnot performed. She died shortly after under-going coronary artery bypass surgery. HV=hyperventilation, AN= amyl nitrite, TM=treadmill, CAD= coronary artery disease.

nostic coronary arteriography and 6 wereshown to have significant coronary arterydisease. All of these I9 had negative amylnitrite tests; no patient with a negative tread-mill test had a positive amyl nitrite test. Thisgroup will not be further considered in thispaper.

Twenty-eight patients had ischaemic ST

F I G . 3 Electrocardiographic effect of amylnitrite in a 45-year-old man with coronaryartery disease. ANi shows flat ischaemic STdepression in II and Vs, while TMi shows'J'-point depression. Both tests are negativeafter coronary artery bypass surgery. HV=hyperventilation, AN= amyl nitrite, TM=treadmill, CAD = coronary artery disease.

CONTROL HV

i POST OPERA1E.O. (CAD)

AN2TM

ANI TMI AN2

:3

V5

A.S. (CAD)

A



j.com/

Br H


ownloaded from


Effects of amyl nitrite in coronary artery disease 853

CONTROL H V ANI TM,

PROPRANOLOLS.S. (ANGINA)

FIG. 4 Electrocardiographic effects of amyl nitrite in a s8-year-old man with clinically typicalangina and normal coronary arteries. The first amyl nitrite and treadmill tests show ischaemicST depressions. (Maximal amyl nitrite rate 126 per minute.) While on propranolol, the maximalamyl nitrite rate was only go/min, and the test (AN2) is negative. The second treadmill test(TM2) remains positive. Atrial pacing to a maximal rate of 120/minute fails to induce ischaemicST depression. HV= hyperventilation; AN= amyl nitrite; TM= treadmill.

depression on treadmill tests, but negativeamyl nitrite tests. Eighteen of these patientsunderwent coronary arteriography; I5 wereshown to have significant coronary arterydisease (Group A). Eleven patients (Group B)Wdisplayed ischaemic ST depression on amylnitrite inhalation which was reproducible onseveral occasions; a total of 17 positive amylnitrite tests was obtained in these i i cases(Fig. i-6). Of these ii patients, 8 had coron-ary artery disease, shown angiographically(Fig. I-5). One patient had aortic stenosiswith normal coronary arteries, one sufferedfrom clinically typical angina pectoris withangiographically demonstrated normal coron-*ary arteries (Fig. 4), and one had systemiclupus erythematosus (Fig. 6). Of these iipatients, I0 also had positive treadmill tests;one had 'J' depression only on treadmilltesting (Fig. 3).

Three additional patients developed 'J'-point depression after amyl nitrite, as well ason treadmill testing (Fig. 7).

Atrial pacing studies were performed on 4patients: one with ischaemic ST depressionsinduced by both amyl nitrite and treadmilltests (Fig. 4); one with 'J' depression on bothamyl nitrite and treadmill test (Fig. 7); and 2with positive treadmill tests but negative amylwaitrite tests. The latter 2 were chosen becausethe maximal heart rates achieved on theirnegative amyl nitrite test (i I0 and ioo beats/min) were considerably slower than the corre-sponding maximal treadmill rates (130 and130). These 4 patients were paced- to ratesequal to or exceeding their maximal amylnitrite induced rates and/or maximal tread-

mill rates. In no case did atrial pacing resultin ischaemic ST changes.Amyl nitrite inhalation produced hypoten-

sion and tachycardia in all subjects. This effectwas maximal in about 30 seconds after thefinal inhalation, and was generally gone in iminute. No patient developed chest pain afteramyl nitrite. These effects were invariablygreater in the standing than in the supineposition, and the observed electrocardio-graphic changes were always more prominent

FIG. 5 Electrocardiographic effects of amylnitrite in a 48-year-old man with coronaryartery disease. ANi and TMi show ischaemicST depression, especially in lead II. Afterpropranolol therapy, AN2 shows equivocal STdepression in lead II when compared with thehyperventilation (HV) tracing, while Vs nolonger shows any ST depression greater thanthe control and hyperventilation tracings. Thetreadmill test (TM2) remains positive. HV=hyperventilation; AN= amyl nitrite; TM=treadmill; CAD = coronary disease.

CONTROL HV ANI TM,~~~~~~~~11 _I 1

A hi ,3 PAC I NG

TM2

I

iPROPRANOLOL

R.C. (CAD)



j.com/

Br H


ownloaded from



standing than supine. For purposes of statisti-cal analysis, the I5 amyl nitrite tests in GroupA patients (coronary artery disease, positivetreadmill test, negative amyl nitrite test) werecompared with the I7 positive amyl nitritetests obtained in Group B patients (Fig. 8-I I).The mean ages and sex distribution withinthese two groups were not statistically differ-ent. Before amyl nitrite, there were no signifi-cant differences between the two groups inheart rate and systolic and diastolic bloodpressure. After amyl nitrite inhalation, theGroup B patients developed significantlyfaster maximal heart rates and lower systolicand diastolic blood pressures. The product ofheart rate and systolic blood pressure, however,was not significantly different between the twogroups, either before or after amyl nitrite.Three patients with positive amyl nitrite

tests underwent repeat amyl nitrite testingwhile receiving oral propranolol. In all, theblood pressure changes were unaffected, butthe mean maximal heart rate attained wasreduced from 135 to III (I7% decline) whilethe patient was receiving propranolol. In I ofthese 3 patients, the repeat amyl nitrite testperformed during the administration of pro-pranolol became negative (Fig. 4), while in asecond, the repeat test was equivocal (Fig. 5).

Sublingual nitroglycerin was administeredto IO standing patients with coronary arterydisease, including 2 with positive amyl nitritetests. In no instance did the blood pressureor heart rate changes of nitroglycerin approachthose induced by amyl nitrite, and no ischae-mic electrocardiographic changes occurred.

HV AN9 TM1 AN2 TM2

fI

5

PREDNISONE

D. N. (SLE)FIG. 6 Electrocardiographic effects of amylnitrite inhalation in a 29-year-old womanwith systemic lupus erythematosus and pre-sumed inflammatory disease of small coronaryarteries. Both ANi and TMi show ischaemicST depressions. Intensive prednisone therapyresulted in some clinical improvement, but bothtests remain positive.

ary artery disease or aortic stenosis; in viewof a positive treadmill test, he is consideredto have the syndrome of angina with normalcoronary arteries (Kemp, Elliott, and Gorlin,I967). Thus, there is an excellent correlationbetween anatomical or physiological lesionsknown to produce myocardial ischaemia, andelectrocardiographic ST depression inducedby amyl nitrite inhalation. This supports thepresumption that the observed electrocardio-graphic changes are in fact a result of myo-cardial ischaemia. The failure to elicit chest

DiscussionThe initial studies of Contro et al. (I952) andothers (Varela de Seijas Aguilar et al., I953;Rizzon et al., I963) accepted both ST seg-ment depression and T wave inversion as evi-dence of amyl nitrite induced myocardialischaemia. These criteria were too broad, asshown by T wave inversion in a high percen-tage of young healthy military recruitsinhaling amyl nitrite (Myburgh, I969). Twave inversions alone were therefore notaccepted as indicative of myocardial ischaemiain this study. An additional reservation withregard to the validity of the earlier studies wasthe absence of anatomical or radiologicalevaluation of the coronary circulation in theascription of coronary artery disease. Coron-ary arteriography was therefore performed onas many patients as possible, including io ofthe I i patients with positive amyl nitrite tests.In the latter group, only i of the IO patientsstudied angiographically did not have coron-

F I G . 7 Electrocardiographic effects of amylnitrite inhalation in a 53-year-old man withcoronary artery disease. 'J7'-point depressionis present in both control and hyperventilationtracings and is accentuated by both amylnitrite and treadmill testing (maximal amylnitrite rate I40/minute). Atrial pacing toISo/minute induces only minimal 'J'-pointdepression, less than the AN and TM tests.HV= hyperventilation; AN= amyl nitrite;TM= treadmill; CAD = coronary arterydisease.

V5

CONTROL HV AN TM PACING

A.S (CAD)



j.com/

Br H


ownloaded from



HEART RATE INCREASES AFTER140 AMYL NITRITE INHALATI N

GROUP

120.U,

tU

w

P80

I.-

w 60

40

CONTROL. AMYL NITRITE

PN-O.O5 PIC 0.01

FIG. 8 Heart rates before and after amylnitrite inhalation in erect patients. Group Apatients have coronary artery disease, positiveatreadmill tests, and negative amyl nitrite tests;Group B patients have positive amyl nitritetests. There is no significant difference inresting heart rate before amyl nitrite inhala-tion, while after the drug the maximal heartrate achieved is significantly faster in Group Bpatients.

pain during any of the positive amyl nitritetests remains unexplained, but may be relatedto the very brief duration of the amyl nitriteeffects (usually less than one minute). Similarfailure of chest pain to accompany ischaemicST depression has also been reported in ahigh percentage of patients undergoing maxi-mal treadmill stress testing (Ellestad et al.,;I969).

How does amyl nitrite, a vasodilating agent,produce myocardial ischaemia ? The inhala-tion of amyl nitrite results in a conspicuousfall in systemic and pulmonary vascular re-sistance and a precipitous drop in systemicarterial pressure (Mason and Braunwald,

:'I965). A reflex venoconstriction and tachy-cardia ensue, and the cardiac output risesrapidly because of the decreased systemic re-sistance (afterload) and increased venous re-

turn (pre-load), the latter resulting fromintense venoconstriction. However, the ten-dency for the reflex venoconstriction to in-

X)crease the pre-load and work of the ventricleis overbalanced by the greater effect of thereduced afterload; this results in overall re-

ductions of intraventricular systolic pressureand volume and diminished myocardialtension and oxygen demand (Mason, Zelis,and Amsterdam, I97I). This is the physio-logical basis for the use of amyl nitrite in the

relief of angina pectoris. On the other hand,the fall in diastolic blood pressure results ina conspicuous fall in coronary perfusion pres-sure. At the same time, tachycardia results inan abbreviation of diastole and coronary bloodflow. These factors probably assume addi-tional importance in a diseased coronary ves-sel with critical luminal stenosis. It has,therefore, been suggested that the amyl nitriteinduced rapid fall in blood pressure might,paradoxically, worsen ischaemia in instanceswhere the decline in coronary blood flow isgreater than the decrease in myocardial oxy-gen need (Mason et al., I97I).

This investigation presents evidence tosupport this hypothesis. It is also supportedby comparison of the Group B patients (posi-tive amyl nitrite tests) with Group A patients(coronary artery disease, positive treadmilltest, negative amyl nitrite test) with regard toamyl nitrite induced heart rate and bloodpressure chahge. These changes were signifi-cantly greater in the patients with positiveamyl nitrite tests than those with negativetests. On the other hand, the product of sys-tolic blood pressure and heart rate - a roughindicator of myocardial oxygen requirements- declined about equally in both groups. Thus,amyl nitrite appears to have reduced myo-cardial oxygen demand to the same degree inboth groups of patients; those who developgreater hypotension and tachycardia with thedrug presumably incur a fall in coronaryblood flow which transiently outweighs thereduced myocardial oxygen demand and re-sults in a temporary period of myocardialischaemia.

FIG. 9 Systolic blood pressure before andafter amyl nitrite inhalation in erect patients.Group B patients achieve a significantly lowerminimal systolic blood pressure after amylnitrite.

SYSTOLIC BLOOD PRESSURE CHANGES150 AFTER AMYL NITRITE INHALATION

C) TGROUP A.EEm!GROUP 5.

Q.~~~~--0.o00

Cl)



j.com/

Br H


ownloaded from



DIASTOLIC BLOOD PRESSUREW t AFTER AMYL NITRITE INHAl

60

20I-is

elCONTROL AMYL NITP.O.5 PIO.O

FIG. IO Diastolic blood pressure biafter amyl nitrite inhalation in ereciGroup B patients achieve a significaminimal diastolic blood pressure aftenitrite.

Beta-adrenergic blockade with 1modified the amyl nitrite effect bthe reflex tachycardia, resultinabbreviation of diastole and a loIof coronary blood flow. In addmyocardial oxygen demands are j

propranolol (Wolfson et al., I966).positive amyl nitrite tests becameequivocal after propranolol wasOn the other hand, atrial pacing akrates equivalent to the amyl nitri

FIG. II The product of systolic blpressure and heart rate in erect patiand after amyl nitrite inhalation. 7causes almost equal falls in both gropatients.

Q SYSTOLIC PRESSURE X HEART R,AFTER AMYL NITRITE INHA

£12' 0

4000-

I-CO)

a)1E 1

CHANGES maximal rates did not result in electrocardio-LATION graphic changes suggesting ischaemia. This is

GROUP A . probably because the diastolic blood pressureGR. a. rose slightly during pacing (the mean di-

astolic blood pressure rose from 70 to 87mmHg, a i6% increase) as the stroke volumefell. Thus, though diastole was abbreviated,coronary perfusion pressure rose rather thanfell, and the net reduction in coronary bloodflow was less than with amyl nitrite.

Nitroglycerin was reported to exacerbateelectrocardiographic changes caused by exer-cise in I7 of I58 patients tested by Russek,Urbach, and Zohman (1955). However, we

'RITE were unable to produce ischaemic electro-0s cardiographic change in standing, non-exer-efore and cising patients by administering sublingualt patients. nitroglycerin. The difference between thesezntly lower results and amyl nitrite's effects is probablyor amyl explained by the observation that nitro-

glycerin's direct action on the veins producingvenodilatation and reduced ventricular pre-load appears to outweigh its action on thearterioles of decreasing systemic vascular

y reducing.ll resistance (afterload) (Mason et al., 1971). InY .r g contrast, inhaled amyl nitrite enters the circu-ig m less lation much more rapidly, and its predomin-iger period ant effect is on the arterial bed. Hypotension1ition, total and tachycardia are much less with nitro-reduced by glycerin than with amyl nitrite (Mason et al.,Thus, two 1971).negative or From a clinical viewpoint, amyl nitriteinstituted. probably has application in the therapeutic

one to heart evaluation of coronary artery disease. Positiveite induced amyl nitrite tests were repeatedly positive inpatients who were displaying a partial or un-satisfactory response to medical therapy for

tsood coronary artery disease (Fig. 2) or, in one

'ents before case, to corticosteroid therapy for systemicvhe drug lupus erythematosus (this patient is believed)ups of to have inflammatory disease of the small

coronary vessels (Brigden et al., I960), butpATE CHANGES coronary angiography had not been done)4LATION (Fig. 6). In other instances, the test becameGROUP A *O negative after surgical revascularization (Fig.GROUP B. 3) and propranolol therapy (Fig. 4 and 5),

paralleling the results of treadmill testing andclinical improvement in these patients. Asnoted, there is good correlation between ana-tomical coronary artery disease and a positiveamyl nitrite test; the percentage of false posi-tive tests appears to be low, and probablyapproximates that of treadmill testing (Masterand Rosenfeld, I967). On the other hand, I5patients were encountered who had angio-graphically proven coronary disease and posi-tive treadmill tests, but negative amyl nitritetests. Thus, the number of false negative

RITE amyl nitrite tests is substantial, which limits)3 the diagnostic value of the procedure. It may



j.com/

Br H


ownloaded from



be worth while to perform it for diagnosticpurposes in the occasional patient who is un-able to undergo exercise stress testing becauseof a physical disability (amputation, arthritis,etc.); a positive result strongly suggests thattreadmill testing and angiographic studieswill be abnormal.athe authors are indebted to Mrs. Utra Nair andMr. Warren Clark for their enthusiastic co-opera-tion in performing the treadmill and amyl nitritestudies, and to Dr. E. William Hancock for hisinvaluable advice and criticism in planning thisstudy.

RteferencesBeck, W., Schrire, V., Vogelpoel, L., Nellen, M., and

Swanepoel, A. (I96I). Hemodynamic effects ofamyl nitrite and phenylephrine on the normalhuman circulation and their relation to changes incardiac murmurs. American Journal of Cardiology,8, 34I.

Brigden, W., Bywaters, E. G. L., Lessof, M. H., and, Ross, I. P. (I960). The heart in systemic lupus

erythematosus. British Heart Journal, 22, I.Bruce, R. A., Blackmon, J. R., Jones, J. W., and Strait,

G. (I963). Exercising testing in adult normal sub-jects and cardiac patients. Pediatrics, 32, 742.

Brunton, T. L. (I867). On the use of nitrite of amyl inangina pectoris. Lancet, 2, 97.

Cohn, K. E., Flamm, M. D., and Hancock, E. W.(I968). Amyl nitrite inhalation as a screening test

, for hypertrophic subaortic stenosis. AmericanJournal of Cardiology, 2I, 68I.

Contro, S., Haring, 0. M., and Goldstein, W. (I952).Paradoxic action of amyl nitrite in coronary pa-tients. Circulation, 6, 250.

Ellestad, M. H., Allen, W., Wan, M. C. K., and Kemp,G. L. (I969). Maximal treadmill stress testing forcardiovascular evaluation. Circulation, 39, 517.

Goldberger, E. (I945). The effects of amyl nitrite onthe downward T wave of the electrocardiogram.American Heart Journal, 30, 60.

Kawai, C., and Hultgren, H. N. (I964). The effect ofdigitalis upon the exercise electrocardiogram.American Heart Journal, 68, 409.

Kemp, H. G., Elliott, W. C., and Gorlin, R. (I967).The anginal syndrome with normal coronaryarteriography. Transactions of the Association ofAmerican Physicians, 80, 59.

Mason, D. T., and Braunwald, E. (I965). The effectsof nitroglycerin and amyl nitrite on arteriolar andvenous tone in the human forearm. Circulation, 32,755.

Mason, D. T., Zelis, R., and Amsterdam, E. A.(I97I). Actions of the nitrites on the peripheralcirculation and myocardial oxygen consumption:significance in the relief of angina pectoris. Chest,59, 296.

Master, A. M., and Rosenfeld, I. (I967). Two-stepexercise test: current status after 25 years. ModernConcepts of Cardiovascular Disease, 36, I9.

Myburgh, D. P. (I969). The effect of amyl nitrite onthe electrocardiogram of normal subjects. SouthAfrican Medical Journal, 43, 5I7.

Perloff, J. K., Calvin, J., De Leon, A. C., and Bowen,P. (I963). Systemic hemodynamic effects of amylnitrite in normal man. American HeartyJournal, 66,460.

Rizzon, P., Lovreglio, V., and Bacca, F. (I963).Modificazioni E.C. grafiche indotte dal nitritod'amile a carico della ripolarizzazione ventricolare:loro valore diagnostico quale test di insufficienzacoronarica e loro interpretazione. Folia Cardio-logica, 22, I47.

Russek, H. I., Urbach, K. F., and Zohman, B. L.(I955). Paradoxical action of glyceryl trinitrate(nitroglycerin) in coronary patients. Journal of theAmerican Medical Association, I58, I017.

Varela de Seijas Aguilar, J., Losada Trulock, E. R.,Martin Perez, G., and Herrero Botas, A. (I953).Sobre la objectivacion de la anoxia miocardicamediante electrocardiogramas obtenidos duranteel ejercicio y tras la inhalacion de nitrito de amilo.Revista Espafiola de Cardiologia, 7, 230.

Wolfson, S., Heinle, R. A., Herman, M. V., Kemp,H. G., Sullivan, J. M., and Gorlin, R. (I966).Propranolol and angina pectoris. American Journalof Cardiology, I8, 345.

Requests for reprints to Dr. Donald C. Harrison,Cardiology Division, Stanford University MedicalCenter, Stanford, California 94305, U.S.A.



j.com/

Br H


ownloaded from


journal, paradoxical electrocardiographic effects of amyl

Documents