J Characteristics presystolic the superior new thoughts on forgotten · 2007-08-28 · second sounds. Wenowreport three cases in which presystolic sounds were identified by phonocardiography

Characteristics of presystolic flow in the superiorvena cava: new thoughts on a forgotten sound

S J D Brecker, H B Xiao, M Mbaissouroum, D G Gibson

AbstractCross sectional, M mode, and Dopplerechocardiography, apexcardiography,and phonocardiography were used tocharacterise presystolic cardiovascularsounds in three patients with ventriculardisease. Although the aetiology wasdifferent (dilated cardiomyopathy,primary pulmonary hypertension, andchronic pulmonary thromboembolicdisease), in each case the presystolicsound was associated with a rapidchange in acceleration of blood and withflow reversal in the superior vena cava,and could only be recorded at the rightsternal edge or over the jugular veins.Such flow characteristics may beexplained by a raised ventricularend diastolic pressure with reducedcompliance. Use of these techniqueshelps to understand the cause of apreviously described but little recog-nised heart sound, and adds weight to theinterpretation of its presence in disease.

31

MethodsCross sectional and M mode echocardiogramswere taken with the patient in the standardleft lateral position with an AdvancedTechnical Laboratory 860C Imager with a3-5 MHz mechanical transducer. We recordedDoppler signals with a Doptek Spectrascanand a 2 MHz transducer. Peak transmitral andtranstricuspid flow velocities were identified bycontinuous wave and were recorded in pulsedwave mode with a 3 mm gate and 250 Hz wallfilter. Regurgitant flow was identified andrecorded in continuous mode. Apexcardio-grams were recorded from the point ofmaximal impulse by a Cambridge Instrumentstransducer with a time constant of fourseconds. Phonocardiograms were recordedfrom a Leatham microphone with a lowfrequency filter. Apexcardiograms, M modeechocardiograms, and Doppler traces wererecorded separately with simultaneouselectrocardiogram and phonocardiogram on aHoneywell (Ecoline 22) strip chart recorder ata paper speed of 10 cm/s.

Cardiac Department,Royal BromptonNational Heart andLung Hospital, LondonS J D BreckerH B XiaoM MbaissouroumD G GibsonCorrespondence toDr D G Gibson,Cardiac Deparunent,Royal Brompton NationalHeart and Lung Hospital,Sydney Street, Chelsea,London SW3 6NP.Accepted for publication9 February 1992

Heart sounds may be classified into thoseoriginating from the valves, which coincidewith the final halt of opening and closing of thevalves and are of high frequency, and addedventricular filling sounds, which are of lowfrequency and comprise the third sound, theatrial or fourth sound, or summation of thetwo.' Such added heart sounds occurring dur-ing diastole are important physical findingsand often indicate ventricular disease. Withnon-invasive techniques such as Doppler,phono, and echocardiography, we can studythe direction, position, and velocity of bloodflow within the heart at the instant an addedsound occurs. Such techniques have shownthat the onset of a third sound occurs at thetime of peak transmitral flow velocity (duringrapid filling), whereas a fourth sound usuallyoccurs at the beginning of flow at the start ofatrial systole.2 The mechanisms are thus dif-ferent from those responsible for the first andsecond sounds. We now report three cases inwhich presystolic sounds were identified byphonocardiography. The aetiology in eachcase was different, yet similar patho-physiological mechanisms have been iden-tified. This added sound has features differentfrom diastolic sounds usually present inpatients with ventricular disease, so we

analysed the mechanisms responsible.

Case reportsPATIENT 1

A 46 year old woman with previous good healthpresented with chest pain and breathlessness.Physical examination revealed tachycardia,raised venous pressure (20 cm above the sternalangle), ankle oedema, and cardiomegaly. Herchest x ray film showed pulmonary oedema,although her electrocardiogram and cardiacenzyme concentrations were normal. A diag-nosis of viral myopericarditis was made duringa prolonged hospital admission, and she wasdischarged on diuretics and angiotensinconverting enzyme inhibitors. She was re-admitted four months later with increasinglysevere dyspnoea, orthopnoea, and ankle swell-ing. Pansystolic murmurs were audible at theapex and lower left stemal edge. Chest x ray filmshowed an enlarged cardiac shadow withpulmonary venous congestion, and herelectrocardiogram remained unchanged fromthe previous recording with a heart rate of 120and PR interval of 0-16 s. She was referred forechocardiographic assessment. This showedbiventricular enlargement with global hypo-kinesis. No convincing added sound could beheard with phonocardiography from the leftstemal edge. An unimpressive low frequencydeflectionwas present at thepeakofrapid filling,but after the start of the succeeding P wave (fig

Br Heart J 1992;68:31-37

on Novem

ber 1, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.68.7.31 on 1 July 1992. D

ownloaded from

http://heart.bmj.com/

Brecker, Xiao, Mbaissouroum, Gibson

Figure I Pulsed waveDoppler recordings oftransmitral flow velocityfrom patient 1, withsimultaneouselectrocardiogram (ECG)and phonocardiogram(PCG). Time intervals of100 ms are marked.(A) Phonocardiogramrecordedfrom lower leftsternal edge. The onlypossible diastolic deflectionrecordedfrom this sitecoincides with the peakforward transmitralflow,characteristic of asummation sound.(B) Jugular venousphonocardiogram shows aloud presystolic sound (S),140 ms after the P waveoccurring after peakforward flow when thevelocity is decreasing. Fullscale deflection is 4 kHz.

A loomsAECG

X4kHz X B

....................................k...,.. . . .......................

2nd

'11Ii-i' II1111 I111111

lB111111111 10111!21111 ECG11111111111{111~~~IFl-

Summation..Hz__.... .. .. .. .. .. .. . .. . .. .. . . .. . _..

140jPHONO Ms411[t111j1Si11jX 11 [101i114L 1X11ki1 j14ty li jkll tali:010~~~~~P

IA). The only possible explanation of thiswould be a summation sound. No other addedsound was recorded from this site, but on thejugular venous phonocardiogram, 140 ms afterthe start of the P wave, a second, loud highfrequency vibration was present (fig 1B). Thissound was localised to an area at the base of theneck and upper right sternal edge. It could notbe recorded by placing the microphone lowerdown the sternal edge or at the apex. The shortdiastolic filling time, localised nature, andtiming in relation to the P wave made it unlikelyto be a fourth heart sound, so a detailedDoppler, echo, and phono cardiographic studywas performed to identify its origin.The first and second heart sounds occurred

normally in association with mitral and aorticclosure identified from the correspondingechograms. Mitral and tricuspid regurgitationwere present, limiting the time available forfilling of the left ventricle to 220 ms, and theright to 260 ms. On the transmitral Dopplerrecording a single peak of forward flow wasidentified and the only possible added soundrecordable from the left sternal edge coincidedwith the peak velocity of left ventricular inflow,characteristic of a summation sound (fig 1A).The sound under study occurred 90 ms afterthis and 40 ms after peak forward tricuspid flowwhile the velocity of atrioventricular flow onboth sides of the heart was steadily decreasing.Premature closure of the tricuspid valve' wasexcluded by showing normal timing on theechocardiogram of the tricuspid valve, co-incident with the first heart sound. As themaximal amplitude of the sound was recordedwith the transducer placed beneath the rightclavicle, we considered the possibility of theorigin of the sound being within the superiorvena cava. We therefore recorded pulsed waveDoppler signals from the right supraclavicularfossa directed towards the right atrium (fig 2). Asimultaneous jugular venous pulse recordingshowed that the "x" descent was shallow, butthe "y" descent was dominant and deep, indi-cating predominant early diastolic antegradeflow into the right atrium characteristic of arestrictive filling pattern. Also the apexcar-diogram was strongly suggestive of restrictedventricular filling. Diastolic downwardflow wasrecorded from the superior vena cava to theright atrium on Doppler. On the spectraldisplay a rapid halt to forward flow could beseen with subsequent flow reversal (peakforward velocity 0 5 m/s; acceleration in aretrograde direction 7 7 m/s' = 0 79 g) occur-ring before ventricular systole, with retrogradeflow from the right atrium towards the superiorvena cava. The start of the high-pitched addedsound occurred at the time of peak velocity ofinflow to the right atrium and coincided withabrupt deceleration.

PATIENT 2A 29 year old woman with primary pulmonaryhypertension was referred for assessment forheart lung transplantation. She had been born atwin by caesarean section. At school she hadbeen breathless on mild exertion and had

c

.o

32

on Novem



/B


ownloaded from


Characteristics ofpresystolicflow in the superior vena cava: new thoughts on aforgotten sound

I111111111111111111111111111111111Illlllllllllllllllllllllll

ECG

2kHz

$s.~~~% M. va0-%*

Is ils isWAsFigure 2 Pulsed wave Dopplerflow of the superior vena cava recordedfrom the rightsupraclavicular fossa directed towards the right atriumfrom patient 1. Simultaneouselectrocardiogram (ECG) andjugular venous phonocardiogram (PCG) are displayed.Loud presystolic "superior vena cava sound" (S), the start of which coincides with thepeak velocity into right atrium at a point of abrupt deceleration. Full scale deflection is2 kHz.

avoided strenuous activities. Symptoms pro.

gressed and cyanosis was noted. After cardiaccatheterisation at another hospital, a diagnosiEof primary pulmonary hypertension had beermade and she was referred for assessment anc

consideration for heart-lung transplantationOn examination she was centrally cyanosed anc

her pulse was 80 beats per minute and regulaiwith a normal volume and waveform. Thtvenous pressure was raised to 20 cm with Egiant "a" wave clearly visible (fig 3). Bloocpressure was 110/70 mm Hg, the righiventricular impulse was prominent, and or

auscultation, a gallop rhythm together with Esoft pansystolic murmur were audible at thtapex. The pulmonary component of the seconcheart sound was accentuated. The electrocardiogram showed right axis deviation anc

chest x ray film showed enlarged pulmonar3arteries with peripheral pruning. At cardia(

catheterisation, pulmonary artery pressure wasrecorded as 170/1 10 mm Hg. A diagnosis ofprimary pulmonary hypertension with a patentforamen ovale was made.

Echocardiography showed a small leftventricular cavity and a hypertrophied andmildly dilated right ventricle. Tricuspidregurgitation was present with a peak rightventricular atrial gradient of 1 10mm Hg beingrecorded. Phonocardiography showed a pre-systolic sound recorded over the upper rightsternal edge 130 ms after the start ofthe P wave.The transtricuspid Doppler signal (fig 4A)showed that forward atrioventricular flow hadceased and eliminated the possibility of thisbeing a fourth heart sound. Thus despiteelectrical and mechanical evidence of atrialactivity no atrial transtricuspid flow could bedetected, implying a fixed end diastolic rightventricular volume into which no further fillingcould occur. Superior vena caval flow wasexamined with Doppler from the supra-clavicular fossa (fig 4B). Flow reversal could bedetected, with clear retrograde flow from rightatrium to superior vena cava. The peak velocityof this retrograde flow (0-55 m/s, acceleration,9-6 m/s2 = 0-98 g) occurred 130 ms after thebeginning of the P wave. The start of the addedpresystolic sound coincided with the peakretrograde velocity, when the accelerationchanged to deceleration.

PATIENT 3A 57 year old West Indian woman with chronicpulmonary thromboembolic disease wasreferred for assessment because of progressivedyspnoea for eighteen months. Her pulse ratewas 90 beats per minute and regular and bloodpressure was 160/100 mm Hg. The venouspressure was raised with a prominent "a" wavevisible. A prominent right ventricular impulsewas palpable, a fourth heart sound was audible,and the pulmonary component of the secondsound was prominent. Chest x ray film showedenlarged proximal pulmonary arteries and anenlarged cardiac shadow. The electrocardio-gram showed right axis deviation, P pulmonale,and inverted T waves over leads V1-V4. Atcardiac catheterisation the following pressureswere recorded (mm Hg): right atrium, a 17,mean 13; right ventricle 75/14; main

c pulmonary artery 70/30, mean 45. Pulmonarys angiography showed filling defects in the right1 main and both lower lobe pulmonary arteries.

Echocardiography showed the right ven-tricle to be dilated with reversal of septal

1 motion. Prolonged tricuspid regurgitationr shortened the right sided filling time to less thane 200 ms. A single peak was present on thea Doppler recording of forward tricuspid flow,i the start of which coincided with a fourth heartt sound recorded at the right sternal edge. AI louder presystolic sound was recorded over thea right jugular veins. Simultaneous jugulare phonocardiography and recording of superiorI vena caval flow on Doppler (fig 5) showed that

this presystolic sound coincided with the startI of retrograde flow in the superior vena cava

y after atrial systole (peak retrograde velocityc 0 34 m/s; acceleration 15-4 m/s2 = 1 57 g).

33

on Novem



/B


ownloaded from



PHONOS St 2nd

Figure 3 Recording ofjugular venous pulse (JVP) from patient 2, with simultaneouselectrocardiogram (ECG), phonocardiogram (PCG), and respiratory trace (R),frompatient 2. Giant "a" wave coincides with presystolic sound (S). First and second heartsounds are indicated.

DiscussionThe invention of the stethoscope by Laennec in1826 gave physicians the opportunity to useclinical skills in the diagnosis of cardiacdisease.4 Within 10 years of this, controversydeveloped about the physical processes under-lying the generation of heart sounds. The firstand second heart sounds are generally thoughtto arise from closure of the atrioventricular andsemilunar valves.' Diastolic sounds seem to bemore closely related to ventricular filling, andmay be right or left sided. Physiological andpathological third sounds are associated withthe peak of rapid filling at the time of peakvelocity of inflow. Studies with a miniatureaccelerometer of ventricular wall dynamics indogs, have shown that the third sound issimultaneous with sudden reduced accelera-tion.5 Similar techniques have been applied tohumans with third sounds, and this worksupports the concept that the third sound isproduced by a sudden intrinsic limitation oflongitudinal expansion of the left ventricular

wall during early diastole.6 Fourth soundsusually occur at the beginning of atrio-ventricular flow after atrial systole, whenacceleration of blood starts.2 Apart from mitraland tricuspid opening snaps, the only othercommonly recognised diastolic sound is apericardial knock. This is an early diastolicsound heard along the left sternal edge inconstrictive pericarditis, and corresponds tothe sudden end of ventricular filling and thepremature diastolic plateau of the ventricularvolume curve.7 It occurs simultaneously withthe nadir of the "y" descent on the recording ofjugular venous pressure. We have alreadypublished a case report of a sound duringisovolumic relaxation due to intracavity flow.8This sound occurred at the time of peak flowvelocity when acceleration had stopped anddeceleration had started. Thus the generalmechanism of diastolic sound productionseems to be related to the dissipation of energyassociated with sudden changes in acceleration.The presystolic sounds in our patients had

none of the usual characteristics by which werecognise common diastolic sounds. In case 1the sound occurred after the peak forwardtransmitral flow and was not associated withatrial systolic flow into the ventricle. Weconfirmed that it was not due to prematuretricuspid closure3 (the normal PR intervalmade this unlikely). Premature mitral closurehas been described in a patient with aorticregurgitation and first degree heart block.9Again, examination of the Doppler trace andmitral echogram confirmed that this could notbe the explanation. The other unusual featurewas the specific localisation over which thesound could be recorded. This was limited toan area beneath the right clavicle in the para-sternal region. The diastolic sound in thispatient corresponded in time to a change inacceleration associated with subsequent flowreversal in the Doppler recording from thesuperior vena cava. Acceleration of blood hadceased and deceleration had started, and bloodflow in the superior vena cava subsequentlybecame retrograde. The sudden change ofacceleration to deceleration requires the opera-tion of a force and dissipation of energy thatappears as mechanical vibrations generatingthe added sound and may occasionally even bepalpable in the venous pulse. The change inacceleration responsible for the added sound inthis case is analogous to that responsible for athird heart sound, but comes from differentcirculatory structures. The question arises whysuch a change in acceleration and abnormal flowreversal should occur. This patient's apexcardiogram strongly suggested restrictedventricular filling. The patient's previousmyocarditis had progressed to cause sub-stantial ventricular impairment and it seemsreasonable that such a ventricle could displayrestrictive properties. Hence, during rapidearly diastolic filling a point is reached whereatrioventricular blood flow rapidly deceleratesyet venous return from the great veinscontinues. In this situation subsequentretrograde flow into the superior vena cavawould occur. Possibly the start of atrial systole

34

on Novem



/B


ownloaded from



II ~~~ECG111111111111111111111OO

2kHz

A

Figure 4 Pulsed wave Doppler recording of(A) transtricuspidflow velocity, and (B) superior venacavalflow,from patient 2, with simultaneouselectrocardiogram (ECG), phonocardiogram (PCG),and respiratory trace (R). (A) Phonocardiogramrecordedfrom the upper right sternal edge reveals apresystolic sound (arrowed), occurring after peakforward transtricuspidflow, whenflow had ceased.(B) "Superior vena cava sound" (S) occurs at point ofmaximal retrograde velocity, associated with an abruptchange in acceleration. Full scale deflection is 2 kHz.

Li. l

B

would increase the pressure gradient under-lying this retrograde flow. Normal venouscompliance dictates a considerable phase delaybetween flow and increase in pressure. In thiscase the flow and pressure changes in thesuperior vena cava were in phase, indicative ofreduction in compliance or increased stiffness ofthe venous system. This would be expectedwhen the venous pressure itself is greatly raisedand provides further support for the venousorigin of the sound under discussion. Thehigh-pitched nature of the sound might be ameasure of the stiffness of the vein.

In our second case the timing of the addedsound was later and occurred 130 ms after the

start of the P wave. At first sight, this has thecharacteristics of a fourth heart sound, beingcoincident with the "a" wave. Forward tri-cuspid flow and hence ventricular filling hadstopped at this point, however, making thepossibility of atrioventricular origin of thissound unlikely. In this case the sound followeda rapid reversal of flow in the superior venacava, occurring at the point of peak velocity offlow away from the atrium. The magnitude ofthe change in acceleration was greatest in theretrograde direction, which may explain thedifference in timing from the first case. In bothcases 1 and 2, however, the change fromacceleration to deceleration implies a change in

ECGw

35

_.p

on Novem



/B


ownloaded from



11i 11aI11 11111111111111111111111 111 t1 :1

III1=

wPHONO s - !I lllllll 11111111111111

Figure S Pulsed wave Doppler recording of superior vena cavalflow from patient 3,with simultaneous electrocardiogram (ECG), jugular phonocardiogram (PCG) andrespiratory trace (R). The presystolic sound (S) occurs at the start of retrogradeflow inthe superior vena cava, again associated with a change in acceleration. Full scaledeflection is 2 kHz.

the direction of the force acting on the blood.In the third case the sound was only recorded

directly over the jugular veins. Atrial systolecaused both antegrade flow into the rightventricle (with a fourth heart sound recordedover the lower left sternal edge), and rapidreversal of flow in the superior vena cava. Thesound over the jugular veins occurred 120 msafter the P wave and coincided with the start ofacceleration and retrograde flow.The concept of presystolic sounds of venous

origin is not new. In 1908, Sir JamesMackenzie heard "a clear, sharp sound preced-ing the first sound" over the jugular veins of a

patient with tricuspid stenosis.'0 He concludedthat contraction of a hypertrophied rightatrium "sent back a large wave into the jugular,and with such force that it caused the valves inthe jugular and subclavian veins to close with a

snap." Groedel and Miller recorded jugularvenous phonocardiograms describing presys-

tolic sounds interpreted as being of venous

origin-" . . . the great veins may be importantas sound conductors, even in the directionopposite to the flow of blood."" Eight yearslater Dock described the recording of loudsounds during atrial systole over the jugularveins of patients with raised right atrial pres-sure."2 Twelve cases were described with

various pathological conditions. In two casesthe diagnosis was tricuspid stenosis, and it waspostulated that reflux of blood into the greatveins set up vibration in the tensed veins socausing a loud sound. This sound was notaudible, however, below the clavicle. Thepostulated mechanism was violent retrogradeflow at the height of atrial systole associatedwith high right ventricular end diastolicpressure, or tricuspid stenosis. We can find noreference to these sounds in publications since1956, although our experience suggests thatthey are not uncommon. In our patients wepostulate a similar mechanism to that sug-

- gested by Dock, and have been able to show the* sudden change in acceleration of blood assoc-

iated with flow reversal. We thus extend Dock'sideas by suggesting that it is the dissipation ofenergy associated with this sudden accelerationchange that generates vibrations within thegreat veins and hence the sound. In the first twocases, the sound can be shown to occur at a timeof peak velocity, when acceleration has ceased.In the first case this occurred at the point ofpeakforward velocity offlow from the superior venacava to the right atrium. In the second case thesound coincided with peak reverse velocityfrom the right atrium to the superior vena cava.In our third case the sound occurred at the startof rapid acceleration of blood in a retrogradefashion, more in keeping with the mechanismof a classical fourth heart sound. Thus thispreviously described but little recognised classof sound can be shown to have the same basicmechanisms as classical ventricular third andfourth heart sounds.We believe a suitable name for this added

sound is a "superior vena cava sound."Alternative titles such as "fifth sound" may beconfusing and "presystolic sound" tells us littleabout the mechanism producing it. It may bethat different classes of this sound will befurther defined according to their precise timerelations to the flow reversal-for example,

t "forward flow," "reverse flow," orl"acceleration" sounds. Nevertheless, for thetime being the term "superior vena cavasound" seems attractive and defines the site offlow reversal.

Phonocardiography has become less popularin recent years. Nevertheless, its use withDoppler and echocardiography makes itpossible to describe and explain a class ofheart sound, virtually unrecognised since itsdescription by Mackenzie over 80 years ago.Physicians should be alert to the presence ofsuch sounds, as with increased understandingof their genesis, such physical signs maycontinue to provide valuable information aboutventricular physiology.

We dtank Dr Martin St John Sutton, Dr Jane Somerville, andDr Paul Oldershaw for allowing us to discuss their patients in theabove case reports.

S J D B is supported by a British Heart Foundation juniorresearch fellowship.

1 Leatham A. Auscultation ofthe heart. Lancet 1958;ii:703-8.2 Vancheri F, Gibson D. Relation of third and fourth heart

sounds to blood velocity during left ventricular filling.

36

on Novem



/B


ownloaded from



Br Heart J 1989;61:144-8.3 Lee CH, Xiao HB, Gibson DG. Presystolic tricuspid valve

closure: An alternative mechanism of diastolic soundgenesis. Cardiology 1990;77:340-5.

4 Laennec RTH. Traiti de l'auscultation mediate. 2nd ed.Paris: Brosson et Chaude, 1826.

5 Ozawa Y, Smith D, Craige E. Origin of the third heartsound: I. Studies in dogs. Circulation 1983;67:393-8.

6 Ozawa Y, Smith D, Craige E. Origin of the third heartsound: II. Studies in human subjects. Circulation1 983;67:399-404.

7 Tyberg TI, Goodyer AVN, Langou RA. Genesis of apericardial knock in constrictive pericarditis. Am JCardiol 1980;46:570-5.

8 Lee CH, Gibson DG. Isovolumic relaxation sound: a newclass of added heart sound? Br Heart J 1991;65:357-9.

9 Traill TA, Fortuin N. Presystolic mitral closure sound inaortic regurgitation with left ventricular hypertrophy andfirst degree heart block. Br Heart J 1982;48:78-80.

10 Mackenzie J. Valvular defects (continued). In: Mackenzie J,ed. Diseases ofthe heart. London: Oxford University Press,1908:229-34.

11 Groedel FM, Miller M. Studies on the acoustic phenomenaover the vessels of the neck in the healthy and diseasedheart. Exper Med Surg 1944;II:193-215.

12 Dock W. Loud presystolic sounds over the jugular veinsassociated with high venous pressure. Am J Med 1956;20:853-9.

ABSTRACTS IN CARDIOLOGY

Prediction of sudden death in hypertrophiccardiomyopathy

Abnormalities of heavy chain myosin accountfor perhaps 50% of hypertrophic car-diomyopathy (HCM). It now seems that thesite and nature of the aminoacid changeinfluence survival. Why should this be? Onepossibility is that the substitutions interfere to adifferent degree in myofibrillary organisationwithin the myocyte. It is probably significantthat the substitution leading to a change incharge has the greatest deleterious effect onsurvival. Abnormal myofibrillary organisationprobably leads in turn to misshapen cells and

abnormalities in cell to cell organisationproducing disarray and an ideal substrate forarrhythmias. Family history remains a cheapway of identifying those withHCM at high riskof sudden premature death but a good case canbe made for determining the exact gene abnor-mality. The understanding ofhowHCM affectsmyocyte structure will be further forwardedwhen the other genes unrelated to heavy chainmyosin are discovered. These are excitingtimes for those interested in HCM.

M J DAVIES

37

on Novem



/B


ownloaded from


Documents

J Characteristics presystolic the superior new thoughts on forgotten · 2007-08-28 · second sounds. Wenowreport three cases in which presystolic sounds were identified by phonocardiography