Reversed Closure Sequence of the Mitral and Tricuspid Valves in

Congestive Heart Failure

PETER S. RAHKO. MD, FACC, JAMES A. SHAVER, MD, FACC,*
ROSEMARIE SALERNI, MD, FACC*
Madison. Wisconsin and Pirrsbtogh, Pennsylvania

Objectives Thepupw ofthI study WBPtacvslwte functiuaml

aad hemadyanmic IZ!‘WI that determine the adtral.trknsp~d and
aortir~palmoaacy valve ckmne seqwenoeia patknta with dilated
wdirmyopPthy.
Ruc&wmd. The physiologif factors &terminiq closure at-
quence of cardiic vntves in varhw forms of heari discnx have
hem found lo be complex. Few deta exist for dilated wdbmy-
apathy, frarticulatfy for diffemntiating the eff& ofa cmnhwkion
delay versus changes in ventrkular performance.
Melhods. A group of 64 patknta ware compared with 36
control sub@& Timing of vatve ctosare aud m
intewals were determined by cembi,,ed M.-mode ecbwa,dm
phy, phonocard!ogrnphy and spexcmdlography. Hemodynmnk
data fmm right heart &heterizatbn won svaPhk in 46 patients.
Rwh. In all wntrol subje&, the wtk valve &al before
the pulmonary valve aad the mitral valve &avJ hefare the
Irk&d valve. IO the study group, Jo padab (49%) had
revenrd so&-pulmonary satve c&we and 21(%%) al these
bad a lefl-sided can&tion delay. There *err 38 patimls (60%)
who had reverwd mitral-tricuspid valve cbsw, but this wu
unrelated to the prewce of a Ieft.dded mdwtioo d&y. The
prwenee of high vcatrkalar Rlllq pressures and pwr systolk

It has been established by prior studies (l-7) that the major the closure sequence ofthe valves (and thus the nature of thx
components of the first heari sound are coincident with first heart scald) have been reported in several disease
coaptation af the mitral and t+uspid valves BE shown by statzs. With Ebstein’s anomaly, right bundle branch block or
M-mcde echocardiography and coincideat with the down- stopic rhythms originating from the left ventricle, the
stroke of the left and right atrial E waves as recorded by high closure sequence is maintained but the interval between
fidelity micromanometer catheters. It has further been mitral and tricuspid valve closure is prolonged (9.10). Rever-
shown in normal subjects that the mitral valve closes ap sal ofthe closure sequence has been shown in ~atiints paced
proximately 20 tu 30 ms before the tricuspid valve (2,8). from the right vent&k and in mitral stenosis~and Icft’atrial
Depending on the length of this closure interval, the first myxoma (1.1 I). For oatients with left bundle branch black. it
heart sound mey he perceived as single or split. Changes in m&t be assumed &at the mitral-tricuspid closure sequence
would be reversed. However, an aiegant study by Burggraf
(8) demonstrated that the relation is more complex than this.
In his study (8). 27% of patients had a normal closure
sequence, 31% had simullaneous cl~surc and 42% had
reversed closure. Except for one revott (12). there is no
information on the effecidleil and right ventricular function
on the mitral-tricuspid closure seqnence. Because Id ven-
tricular dysfunction is conunon it! patients with left bundle
branch block, simultaacous investigaticn of the eSects ofleA
ventricular dystiwtion and left-sided conduction delay might
add insight into the determinants oftbe Gid-tricuspid closure Digital 12-l& ECGs were avaiIabIe for all patients sod QRS
sequence. Therefore, the purpose of the present study was to duration and PR interval were measured by computer and
determine the mitral-tricuspid closure sequence in a group of reviewed for awuacy by ow of the investi@tors. Those
patients with dilated cardiomyopathy. Ftndings are compared patients who had a QRS dutatioa of ?I20 ms and met the
with those in nonoal subjects and are related to underlying criteria for I& bundle btaach block or an itwaventriadaz
right and left ventricular function. left-sided conduction delay conduction delay ofthe left bundle brooch type were coasid-
and hemodynamic state. The closure sequence of the a& ered as having a left-sided condwtion delay. No patient in the
and puLnonary valves was also determined. study had right bundle branch block.
The following time intervals and measurements of VCII-
tricular function were made (Fig. I):
Methods
1. The mitral closure intewal (Q-MVC) was defined as the
The study group coasisted of 61 patieals aged 18 to 66 interval from the onset of the ECG QRS complex (Q) to
years (mean 52) who had symptoms of congestive hean mitral valve closure (MVC).
failure and evidence of dilated cardiopatby. There were 54 2. The tricuspid closure interval (Q-TVC) was the interval
men and 10 women, who had New York Heart Association from the onset of the QRS complex (Q) to tricuspid valve
functional class II to IV congestive heart failure due to either clos!Jrc l-fw.
coronary artery disease (n = 21) or idiopathic dilated car. 3. The electromechanical coupling interval (&LVR) was
diomvooathv fn = 43). The averwe eiection fraction as the interval from the onset of the QRS complex (Q) to the
asses~eh bi ~dionuclide ventricul&a~hy was 19 i 7% onset of npid upward deflection oa the apexwdiagmm.
(range 6% to 33%). which has been shown to be an excelknt estimate of the
All patients underwent complete M-mode dnd two- iuterval from the Q wave to the wet of the kfl ventricolar
dimensional echocardiography using standard imaging tech- pressure increase (LVR! (lZ.lh!.
niques on a Hewlett-Pa&&d 77020 phased array cardiac 4. The aortic closure interval (Q-AVC) was the total time
ultrasound imaging system. Addition;1 M-mode imagine: was _ ._
of systole from onset &he ORS complex (0) to a&c valve
performed in wmo patients using an Irex System II. F&time clo& (AVC!.
interval measurements, all p&eats had a phoaocardiogmm 5. Left ventricular ejection time (LVET) was measured
(Hewlett-Packard or hex microphone) placed at the base, on the carotid pulse tracing a, $e interval fmm onset of the
whiih was recorded simultaneously with the electmcardio- upstroke to the dicmtic retch.
graphic (KG) lib lead that best defined the onset of the 6. The preejeclioa period (PEP) was defbted as (@AVCl-
QRS complex, an apexcardiogram aad a carotid p&e trac- LVET.
in& Recordiis wore made at a paper speed of IuO m&s. 7. The isovolumetrtc contraction time (ICT) was defined
Pbonwardiographic bandpass liken were adjusted to opti- as PEP - (Q.LVR).
mize the quality of the first ati seccnd heart sounds. The 8. Fractional sbonening (FS) was determined from the
pulse recorder had a time constant of 3.3 s and a frequency M-mode echocardiogram BE(LVEDD - LVESD)/LVEDD.
band of 0.05 to I00 Hz. Adequate apexcardiograms for where LVEDD is left ventricular enddiastolic dimension
aaalyaIs were available in 48 patients. adequate M-mode and LVESD is kfl venlticular end-systolic dimension.
studies for kfl ventricular dimension ana!ysis were present 9. The mean velocitv of circumferential fiber shmieaiaa
in 61 and adequate twdimensional studies to evaluate right was defined as FSiLVk. This was cotrected for heart tati
ventricular size and function were avaiiable in 63. variation by multiplioation by %%Yheart rate fl7).
Ib timing of valve closure was determined from the Right ventricular futtctioa was evaluated from all avail-
M-mode echc-xdiogmm (Fig. I). The interval from onset of able views on the hvadimensional echocardiogram. Systolic
the QRS complex to valve closure ~8% determined for each tinction was semiqu&itativeIy sexed as bping namal (I+).
ofthe four valves. Valve closure could te detrmrined ir, all miId!y to moderately decreaxd (2+) 01 severely decreased
patients for the first heart sound and in 61 patients for the (3t). Right veahkular size was similarly scored ar normal
second heart sound. Valve closure was defined as the point (It), mildly to moderately iwcased (2+) OT severely in
of apoosition of two leaflets. If only one leaflet could be crwed(3t). Poiotsconsforeach variabie werecombined so
visu&ed, closure was defined as the point of abrupt tenni- that rightventrkularhllftionwasgradedfmm2+ (mxmalsii
nation of autetior or poatetior closing motion of the valve and function) to6t (severe dvsfunctionand ealmxement). The
leaflet (8). Aa average of five beats was used to detemxinc r&t ventricular fimction &es ‘were then coklated with
the closure interval of eat!. valve and the sequence oi valve closure sequence and clwxe intervals.
closure was determined froin these averaged &tes. A subgroup bf 46 patients underwent right hearI catheter-
The M-mode ech-phk tneasuremtnts were made ization as part of a clinical study (n = i7). a pretraasplant
ushut criteria of the Amelicao Sowtv cf Echocardiiohv evaluation (n = 14) or a research pmtocol (a = IS). All
(I3~~Systalic time intervals were d&mined from an .a&& studies were completed within 72 h of the noninvasive
of five beats and indexed for variations in heart rata by using studies aad patients were included in this subgroup only if
the regression equations developed by Weisster et al. (14). they remained in clinically stable condition. All pmtoe~ls
 k'iiyreI, Milral (4 and tricuspid (b) valve
echocardiognms recorded simultaneously with
t’lc electrocardiogram W-Xi) and base and apex
ph.mocardiogrmnr (PHONO) in a patient with
dilated irchemk cardmmyopathy and left bundle
branch block. The mitral CM,) and tricuspid (T,)
compnents of the first hart round are coinci-
dent with the terminat elnswe (0 oftbe& rzspec-
live valves. There is a marked delay (100ms) in
!hc closure of the mitrat valve (QMC). resulring
in a reversed sequenceof valve closure and the
first hean sound (T&i,). Pulmonary mtery
wedgr pressure is elevated to 23 mm Hg and is
I5 mm HS greater than the r&3,, avial (RA)
pressure (8 mm Fig). There is a prominent El
bump on the mitral valve echocardiogram(61.
and the delay in mitral valve closure is due to tie
slow terminal closing motion of the mitral valve
during ventricutogenicclosure. In this patient.
the clectromcctw~ical ccupling interval on the
agexcardiogram IQLVR interval) is 50 ms.
Therefore, the ventriculogenicclosure sequence
time is: QMC - (Q-LVR) = (1W ms - 50 ms) =
54”s. AB bumpofrhorterdumtionisprerenton
the tricuspid valve and the tricuspid ctowc n-
teal (QTC) remains normal at 6.0ms. A = mrtat
motion of mitral leaRet: A, = sonic closure
sound; P, = pulmonary closure sound:
S, = third heart round.
,) /

were approve 1 by the InstitutionsI Research Boards of the blood pressures were measured by using either a peripheral
University of Pittsburgh and University of Wisconsin and all arterial line or central aortic pressure using fluid-filled cath-
patients gave informed consent. Hemodynamic measure- eters. As a hemodynamic index of left ventricular function,
ments were made with a 7F triple-lumen balloon flotation AP/At was determined, where AP = (diastolic biw prer-
catheter. Thermodilution cardiac output measurements were sure) - (mean pulmonary artery wedge pressure) and AT =
peifomted. requiting that three samples have <IS% V&I- isavolumetric contraction time (IS).
tton. Pressures were recorded as electronic mean values or A 8roup of 36 patients (I8 men and !? women) with an
as the average value of a complex respiratory cycle when average a8e of 31 t I I years who had a normal ECG, no
phasic values were determined. The systolic and diastolic clinical evidence of heart disease and a normal echacardio-
Table I. Ekclromschanical Intervals and Measuremenrrof Left
Ventricular Sire and Function

pressure was in&ssed to 21.6 2 10.7 nun Hg(range 2 to 43),

the cardiac index was reduced to 2 + 0.5 litersimin veerm2
(range I.? to 3.0) and AP/At WBEreduced to 526 + 286 nun
H& (range 177 to I250).
In each control subject. closure of the mitral valve
preceded closure of the tricuspid valve end closure of the
ventrkubu performance and bemodynamb. Ekctmme- aortic valve preceded closure of the pulmonary valve. The
chanical intervals and measurements of left ventricular size closure sequence for patients with heart failure was differ-
and function ate shown in Table I. Compared with control ent. In 38 patients (60%). the mitral-tricuspid valve closure
subiects. the patients with heart failire had u marked pro- sequence was reversed (p < O.C@Ol vs. control subjects). For
Ion&on of isovolumetric contraction time, mitral v&e the 61 patients in whom it could be evahmted. the aortic-
closure interval and preejection period. Left ventricular pulmonary valve closure sequence was revetsed in 27 (44%).
ejection time was significantly shorter in the patients than in A total of I8 patients (28%) had both aortic-ptdmonaty and
the control subjects. There was no significant difference in mitral-tricuspid valve closure reversal.
the electromechanical coupling interval (Q-LVR), aortic Elect of left vettlricular mnduetii d&v. The conse-
closure interval (QAVC) or tricuspid closure interval (Q- quences of a left ventricular conduction dela~on electrome-
TVC) between the groups. As expected. the study group had chanical intervals are shown in Table 2. In patients with a
a significant increase in left vsntricular dimension and a conduction delay, there was significant prolongation of iso-
significant decrease in !eft ~ventri;co!arrywlir performance volumetric contraction time. prcejection time. left ventricu-
compared with values in control subjects. Heart rate at rest lar eiection time and aotiic closure interval. However, there
was significantly higher in the study group. wa~no significant difference in electromechanical coupling
The subgroup of 46 patients who underwent right heart (Q-LVR) or the closure intervals of the mitral and tricuspid
catheterization had significant hemodynamic abnormalities. valve. There was also no diierence in left ventricular size,
The tneen central veilous pressure was increased to 9.4 + function or hemadynamic values between the groups except
FIgwe 2. Two examples of patients with reversed closure of the valve closttro interval was significantly shorter in the re-
mitral (M”) and tricuspid Valves. Bolh have poor left ventricular versed sequence group than in those with normally se-
function and high tU!ing pressures, which result in a prolongation of quenced valve closure.
mitral cIo9ure (C) due to a large R bump (B). Roth also have
Although absolute filling pressures were not signilicatIlly
relatively well preserved right vsntriculw function and low filling
pressures in the right ventricle. Thus, there is a wide disparity diScrent in the two valve closure sequence groups, the
between right. and left-sided filling pressures. In a, the patient has a difference between mean pulmowy artery wedge pressure
lelt.sidcd conduction delay, but the Q-LVR interval is normal at and central veoou pressure was signiticaotly er=ater in the
U as. In h, the QXS duration is normal, as is Q.LVR (!6 es).
Therefore. the primary delay occurs during irovolumetric cootrac-
lion, which is prolonged to 105 ms ina and IZOms in b(a is recorded
et SO mm/s, b at IO0 mm&. ECG = electrocardiogram: MVC = Function and’HemuJynamm ata by Valve Closure Sequence
mitral valve closure; other abbreviations as in Figure I.
M”. TV TV. MY
,n = 261 tn - 381 p Yak
Q&w ml$ 49 z 9 Jo+9 0.r)
for APIAi. which was reduced in patients with a conduction
ICT tmrl 8bz26 %*38 0.46
delay. The closure sequence of the aortic and pulmonary PEP br, 137 I M 110+38 0.72
valves woe highly dependent on the presence or absence of L.“ET ,mr, 232 f 38 23o + 42 0.83
the ventricular conduction delay. With only three excep- HR tteak.kllin> 85 + I9 85 ? 17 0%
tions, there was a reversal of closure of the sonic and LVEDD km, 7.1 t (1.8 1.3 i 1.1 0.49
pulmonary valves only when a left-sided conduction delay FS 1%) 12.2 * 1.2 14.8 + a.9 0.M
was present. Conversely, the conduction delay had no mwc (circ”mfrrenn~$ 0.45 f 0.19 0.54 + 0.28 0.16
mC”P tmm “I, 9.9 + ,.6 9.0 f 0.1 0.6,
significant efkct on the closure sequence ofthe tricuspid and
RPAWP tmm HEI 18.9 t II.0 23.2 ? 10.D 0.17
mitral valves (Fig. 2).
CoaPar! by ti!ral-tricuspid valve closure sequence.
Electromechanical intervals, left ventricular function and
hemodynamic values did not diier betweb. those with
normal and those with reversed milt&tricuspid valve CID
sure seauencc (Table 3). However. the mitral valve closure
inrerval‘was si&ftca& prolonged. whereas the tricuspid
Tnbterl. Patients Having a Reversed Closure Seqaence of ths
Mitral and Tricuspid Valves by Hemdynamic Subgroup tn = 461

group with reversed sequencing of mitnl-tricuspid valve

closure. To further evaluate this observation. txttients were
classified into hemodyoamic subsets on the basis of right-
and left-sided filline oressures. As shown in Table 4. the
group with the grcak~t difference in filling pressuw bad the and leasl likely when rhet’e was were dysfunction. As right
highest probability of having a reversed valve closure se- ventricular function declined, there was a progressive in-
quence. There was also a weak but significant correlation crease in central venous pressure aad the tricuspid valve
between the difference in tight- and lefl-sided filling pres- closure interval. Pulmonary artery wedge pressure and the
sures and the difference b&een !he tricuspid aad~n&al mitral valve closure interval also increased, but the change
valve closure intervals (Fig. 3). occurred at milder levels of right ventricular dysfunction.
Elf& of right ventricular function. The relation between Thus. the difference between central venous pressure and
right ventricular function and valve closure sequence is wedge pressure and the difference in valve closure intervals
shown in Table 5. There was a wide variation in function was greatest when right ventricular function was moderately
among the patients. Reversed valve clorure was most corn- decwwd because the changes in the intervals were not in
man when there was moderate right ventncular dysfunction phase with each other.
There was oo correlation between the PR interval on the
ECG and the valve closure intewais. Also, there was no
F&m 3. Graph showing the correlation between the difference in ditference in valve closure sequence between those with zed
man pulmonary artery wedge pressurea,d mean central venous without coronary artery disease.
pressure(MPAW - MCVP) YCISUS the dillereace between the
t&spa and mitral valve closure intervals (Q-TVC and Q-MVC).
The datapoints we codedto correspondto hemodynamiccateawier
shown in Table 4. 0 = mean cetttral vettous prewre ~10. mean
Discus&m
pulownwy wedgepressure516: n = meancentnl W”O”S pressurp The principal finding of this stud; is that reversal of the
510. mean pulmonary wedge pressure >t6: x = mean centmt closure sequence of the mitral and tricuspid valves is B
~eoeus pressure >lO. mean pulmonary wedge pressure >I6 tali
common finding in patients with di!ated carGomyopathy and
valuer in mm Hg).
congestive heart failure. This change in valve closure w-
quence is not due to the presence of a left-sided conduction
delay. but appears to be most closely related to the relative
severity of riaht- and left-sided abnormalities of ventricular
fun&~ and-hemodynamics. In contrast. the closttre se-
quence of the aortic and pulmonary valves is highly depen-
dent on the presence of a left-sided conduction delay.
The results of this study agree with and extend priorwork
by Burggmf (8) and others (2.12). The mitral and tricuspid
valve closure intervals in nomxd subjects in this study are
almost identical to those reported by Barggti ((3 and
Brooks et al. (2). Burg@ (8) also found no consistent
relation between the presence of a IeA ventricular canduc-
lion delay and reversal of mitral-tricuspid valve closure
s:quence~but, as we have, did show thit the mitral valve
clorurc interval was prolonged in pntients with reversed
mitral-tricuspid valve closure. Although Burggmf(8) did not
evaluate relative differences in vsnaicular function or filling
pressures. many of his patients appear to have had left
ventricular dysfunction. Because petients with reversed
mitral-tricuspid valve closure had a longer QRS duration, he The site of the conduction delay in patients with left
believed that “the degree of completeness” of the left bundle branch block is difficult to ascertain (20). Electm
ventricular conduction delay explained reversed mitral- physiologic mapping studies and prior nonin&& studies
tricuspid valve closure and that differences in ventricular (20-23) wggest that both proximal and distal dysfunction of
function were nut important (8). the conduction system can occur. If a left-sided conduction
Hultgren et al. (19) also studied the effect of a letl delay is discrete and proximal. it would be expected that
ventricular conduction defect on mechanical events of the elecrromechardcal coupling would be prolonged and that this
cardiac cycle in patients with and without left ventricular delay would be the primary cp.useof a delay in mihxl valve
dysfunction. In bothgroups, the mitral valveclosureintewal closure. This has been shown to be the case in patients with
and the isovolumetric contraction time were prolonged to au left bundle branch block and near normal ventriculv func-
extent similar to that in our study. However, in contrast to tion but was uncommon in our study group with dilated
our study, is which ihe electromechanical coupling time was cardiomyopathy (21). Only two patients had a prolonged
nororal, they (19) observed a significant increase in this electromechanical coupling interval >2 SD above control
interval. They also compared 32 patients with both a left values. However. common to all patients with a left ventric-
ventricular conduction delayand left ventricular dysfunction ular conduction delay and congestive heart failure was a
with a group of 20 patie& with a similar degree of left marked proiongation of isovoiumetric contraction time,
ventricular dysfunction but no left ventricular conduction which was sienilicantlv water than that found in those with
delay. Again, in contrast to our study, abnormalities in congestive heart failure alone (106 f 37 vs. 80 2 25 ms; p <
electromechanical intervals were rarely found in this latter 0.008). These data suggest that the predominant site of
group. From these data, the authors (19) concluded that the conduction delay in patients with dilated cardiomyopathy is
major abnormalities of the cwdiaz cycle in patients with u a peripheral or arboriz.ation block (23). However, incomplete
left ventricular conduction delay are-due to ihe conduction proximal block with dyssyncbrowus contraction in an al-
defect and not to the associated left ventricular dysfunction. ready compromised left ventricle is an alternate explanation
It is not possible to explain the diierences in b&the results for a normal electromechanical coupling interval end a
and conclusions between their study and the current report. significant increase in isovolumetric conbaction time (22).
In the present study, patieots with both a left ventricular Arborization block or greater dyssynchronous contra&a
conduction defect and severe congestive heart failure were may also be responsible for a relative prolongation of left
closely matched in terms of severity of left ventricular ventricular ejection time. It appears that the combined
dystimction (Table 2) with those with severe congestive increase in isovolumetric contraction time and ejection time
hean failure and no left ventricular conduction delay. It is is primarily responsible for the reversed sequence of pulme
possible the study of Hultgren et al. (19) had a more nary and aortic valve closure in these patients with left
heterogeneous patient cohort in each group. bundle branch block. However, prolong&n of isovolumet-

I I I I I f I i I III1 I I I III I

Ftgwc 4. Simultaneous M-mode schewdio-

gram of the mitral valve WV). phonewdl+
grams aI the apex nod base (PHONO) and
apexcardiwem (APEX). There is a prominent
B bump on the mitral valve (B). In this padat.
the mitral valve beginsto close(C) at about the
onset of the QRS sequenceon the electrow.
diogmm (EC(I). Closure continues during the
Q-LVR interval and then abruptly stops..This is
the end of ?.tiiogenicclewe. AS i.9ovolemetB
contrwtion (ICl’l b+ns. the B bump forms and
remains ahnosf to the end of isovolumetrk
contraction.This is the ventriculogenlcpxli~n
of mitral valve closure (44 msl PEP = pre-
ejection period; other abbreviationsas in Fig.
1
ems and 2.
ric contraction. time with left bundle branch block does not A second abnormality of hearts with dilated cardiomyop
explain changes in atrioventriculnr (AV) valve closure SC- athy with or without left venrricuiar conduction defects is a
quence in dilated cardiomyopathy (Table 2. Fig. 2). marked increase in left ventricular end-diastolic volume
 associated with elevated end-diastolic pressure. A common
denominatorfor delayed mitral valve dosore in this setting is
a prolongation of ventticulogenic closure of the mitral valve
(Fig. 4). After atrial contraction ioto the volume-loaded left
ventricle (which is operating on the steep portion of its
compliance curve), there is onset of premature atriogenic
&ware of!he mitral valve. This closure is interrupted by the
onset of the left ventticular contraction, which results in a
prominent B bump on the mitral valve M-mode echacardio-
_
mam (24). Thereafter. the remainder of valve closure reties
an the rapidity of ventricular contraction (ventriculogenic
closure). The mitral closing motion will be progressively
prolonged as the rate of rise of left ventricular pressure
(APlAt) deteriorates as a result of leh ventricular dysfunction
with or without a conduction delay (Table 2, Pig. 4).
The net effect of slowed ventriculogenic valve closure
beginning after partial atriogenic closure is not only delayed
closure of the mitral valve, but also decreased intensity of
the mitral component of the first heart sound. Similarobser-
vations have been made for the tricuspid valve. Delayed
ventriculogenic tricuspid valve closure was found to be most
common when right ventricular end-diastolic pressure ex-
ceeded 9 mm Hg (Fig. 5) (25). Although direct measurement
of right ventricular end-diastolic pressure was not available
in this sNdy, the changes in central venous pressure are
consistent with this observation.
No previous studies have attempted to evaluate the role
of right ventricular dysfunction and the resultant increase in
right-sided filling pressure on AV valve closure sequence. In
this study, reversed AV valve closure was most common
when moderate right ventricular dysfunction was present
and less common when the right ventricle was either normal and pressure aad a significant decreax in the rate of rise of
or severely compromised with a marked increase in the right left ventricular prerwe. the closare sequence may revef~
ventricular filling pressure. Futlhermore, reversed valve as a result of prolonged terminal ventticulogcaic closure of
closure was most likely when there was a large difference the mitral valve. As IcR ventricular Iimctioo becomes se-
between left and right ventricular tilling pressu~s (Table 4, verely compromised and more long-standing, rlx right yea-
Fig. 3). The relation between these iindinns is shown in tricle begins to fail. When right ventricular dysfunction is
Table 5. As right ventricular function de&o&es, there is a
prolongation in closure inter&s of the mitral and tricuspid ventriculoge~~ closure of the tricuspid valve-is not delayed.
valves. The difference between mitral and tricuspid valve Thus, the combination of severe let? ventricular dyshmction
closure intervals is greatest when right ventricular function and mild to moderate right ventricular dysfunction results in
is moderately deoressed. and this level of dvsfunction is the largest squation of left- and right-sided filling prasswas
associated with tie great& difference between-left and right and is the combination with the highest likelihood olrevers-
ventricular filling pressures. As right ventricular dysfunction iag the mitral-tricuspid valve cl&urc sequeace. As right
progresses further, the tricuspid valve closure interval con ventricular dysfunction becomes awe severe aad filling
tinues to prolong, whereas the mitral closure interval re- p~%sures increase to higher levels, pmlonged terminal yea-
mains relatively fixed. Thus, reversed valve closure se- triculogenic closure of th tricuspid valve may occur. In
quence becomes less common. many cases, this may be enough of a delay to restore a
Conclusions. From the nsults of this study and previous normal mitml~tticurpid valve closure sequence.
work, a unified hypothesis on the closure sequence of the The presence of a l&sided cond&tioa defect may
AV valves may be developed (Fig. 6). In almost all reported slightly increase the chances of a reversed mitral-tricusakl
studies, the nomml mitral valve closes 20 to 30 tns before the valve closure sequence by prolongation of the isovolumeiric
tricuspid valve. As left ventricular function deteriordter. the contraction interval. However, this does not appear to be an
valve closure sequence tends to remain normal as loas as important factor in dilated cardiomyopathy. The etiology of
left-sided filling pressnres rema-n nomad. However, once dilated cardiomyopathy also does not appear to bc an
there is an increase in left ventricular nddiastnlic volume important factor.
 Refereaces
1. Shaver IA, Salemi R. Ausculta ion of thr heart In: Hurst 1%‘. Schlam
RC, Rx&y CE, Sonneablick LH. Wenw NK, eds. The Hean. 7th ed
New York: McGraw-Hill. 1990:177-96.
2. Brooks N. Leech G, LearhamA. Factors responsiblefor normal spliuing
of &rsrthean sound: hii speedechopimnoEardographicarudy 01 valve
mwemc”l. Br Hean I 1979:42:695-702.

1I. ~&&” HH, Leo TF. The tricuid conpomnt of rhe 6rsl bean round
in mitral stenosis. Circulation 1958;18:1012-7.
12. P&ash R, Maonhy K. Amww WS. The first hean soud: a phow
cchwudiiphic -l&on with miti, ttiurpid ad aortic valvular
C”III11. Catket cadiovw Diagn 1!37&2:)81-7.
13. Sahn DI. Deh!aria.&, Kieln 1. Weyman A. RecommaMons rewdina