Br HeartJ 1982; 48: 428-33 Two dimensional echocardiographic assessment of aortic valve morphology: feasibility of bicuspid valve detection Prospective study of 100 adult patients MICHAEL J ZEMA, MICHELE CACCAVANO From the Division of Cardiology, Department of Medicine, Brookhaven Memorial Hospital Medical Center, Patchogue, New York, USA SUMMARY Two dimensional echocardiographic criteria for bicuspid aortic valve recognition have greater specificity than previously proposed M-mode echocardiographic criteria. The potential clinical use of the two dimensional technique is, however, limited by the technical inability to image adequately the aortic valve leaflets in many patients. One hundred consecutive adult patients under- going two dimensional echocardiography were prospectively studied. Valve cusp number could not be determined because of dense calcification in eight patients. A bicuspid aortic valve was diagnosed in a single subject. A parasternal short axis view disclosed three commissures (diastolic "Y" configuration) in only 26 patients. Technically adequate parasternal short axis imaging was more likely in younger patients and in non-smokers. In patients not successfully imaged from the parasternal approach, an anteriorly tilted apical four chamber view showed a diastolic "Y" configuration in an additional eight cases. Considering the high prevalence in our population of incomplete two dimensional echocardio- graphic aortic valve leaflet imaging, angiographic and/or pathological studies must be performed to establish the correlation between these incomplete echocardiographic patterns and aortic valve anatomy if two dimensional echocardiography is to have widespread application in the diagnosis of the congenital bicuspid aortic valve. With the exception of mitral valve prolapse, the bicuspid aortic valve is the most common congenital heart lesion found in adults, seen in 0-7% to 2-0% of routine necropsies.12 Moreover, unlike the usually excellent prognosis associated with mitral valve prolapse,3 the subsequent complications of a bicuspid aortic valve include infective endocarditis,-7 pro- gressive valvular stenosis, and/or valvular regurgi- tation.8-'2 Fenoglio and coworkers7 have suggested that about one-third of patients over the age of 20 years born with a non-stenotic bicuspid aortic valve will develop aortic stenosis, one-third will develop aortic regurgitation (in most cases because of endocarditis), and the remainder will be free of any haemo- dynamically significant problem. The greatest single cause of death in subjects with a bicuspid aortic valve is infective endocarditis, with over 75% of these deaths occurring before the age of 50.7 Two dimensional echocardiography appears to be Accepted for publication 26 May 1982 more specific than M-mode echocardiography in the diagnosis of bicuspid aortic valve.'3 14 Practical appli- cation of two dimensional echocardiography to this problem, however, requires that technically adequate images of the aortic valve cusps be obtained. The objectives of this prospective study were, there- fore, (1) to determine the frequency with which the commissural margins of the aortic valve cusps could be imaged by two dimensional echocardiography, (2) ts determine the factors affecting the technical feasibilit of the above determination, and (3) to evaluate the usi of an anteriorly tilted apical four chamber view fo imaging the aortic valve commissural margins b diastole in certain subjects without an adequat parasternal echocardiographic window. Methods (A) PATIENT SELECTION One hundred (43 men and 57 women) consecuti adult patients aged 18 to 91 years (mean 54 years), wit no previous history of aortic valve replacement, we 428 on March 23, 2020 by guest. Protected by copyright. http://heart.bmj.com/ Br Heart J: first published as 10.1136/hrt.48.5.428 on 1 November 1982. Downloaded from
Transcript
BrHeartJ 1982; 48: 428-33
Two dimensional echocardiographic assessment ofaorticvalve morphology: feasibility of bicuspid valve detection
Prospective study of100 adult patients
MICHAEL J ZEMA, MICHELE CACCAVANO
From the Division ofCardiology, Department ofMedicine, Brookhaven Memorial Hospital Medical Center, Patchogue,New York, USA
SUMMARY Two dimensional echocardiographic criteria for bicuspid aortic valve recognition havegreater specificity than previously proposed M-mode echocardiographic criteria. The potential clinicaluse ofthe two dimensional technique is, however, limited by the technical inability to image adequatelythe aortic valve leaflets in many patients. One hundred consecutive adult patients under-going two dimensional echocardiography were prospectively studied. Valve cusp number could not bedetermined because of dense calcification in eight patients. A bicuspid aortic valve was diagnosed in a
single subject. A parasternal short axis view disclosed three commissures (diastolic "Y" configuration)in only 26 patients. Technically adequate parasternal short axis imaging was more likely in younger
patients and in non-smokers. In patients not successfully imaged from the parasternal approach, an
anteriorly tilted apical four chamber view showed a diastolic "Y" configuration in an additional eightcases. Considering the high prevalence in our population of incomplete two dimensional echocardio-graphic aortic valve leaflet imaging, angiographic and/or pathological studies must be performed to
establish the correlation between these incomplete echocardiographic patterns and aortic valveanatomy if two dimensional echocardiography is to have widespread application in the diagnosis of thecongenital bicuspid aortic valve.
With the exception of mitral valve prolapse, thebicuspid aortic valve is the most common congenitalheart lesion found in adults, seen in 0-7% to 2-0% ofroutine necropsies.12 Moreover, unlike the usuallyexcellent prognosis associated with mitral valveprolapse,3 the subsequent complications of a bicuspidaortic valve include infective endocarditis,-7 pro-gressive valvular stenosis, and/or valvular regurgi-tation.8-'2 Fenoglio and coworkers7 have suggestedthat about one-third of patients over the age of20 yearsborn with a non-stenotic bicuspid aortic valve willdevelop aortic stenosis, one-third will develop aorticregurgitation (in most cases because of endocarditis),and the remainder will be free of any haemo-dynamically significant problem. The greatest singlecause of death in subjects with a bicuspid aortic valve isinfective endocarditis, with over 75% of these deaths
occurring before the age of 50.7Two dimensional echocardiography appears to be
Accepted for publication 26 May 1982
more specific than M-mode echocardiography in thediagnosis of bicuspid aortic valve.'3 14 Practical appli-cation of two dimensional echocardiography to thisproblem, however, requires that technically adequateimages of the aortic valve cusps be obtained.The objectives of this prospective study were, there-
fore, (1) to determine the frequency with which thecommissural margins of the aortic valve cusps could beimaged by two dimensional echocardiography, (2) tsdetermine the factors affecting the technical feasibilitof the above determination, and (3) to evaluate the usiof an anteriorly tilted apical four chamber view foimaging the aortic valve commissural margins bdiastole in certain subjects without an adequatparasternal echocardiographic window.
Methods
(A) PATIENT SELECTIONOne hundred (43 men and 57 women) consecutiadult patients aged 18 to 91 years (mean 54 years), witno previous history of aortic valve replacement, we
428
on March 23, 2020 by guest. P
rotected by copyright.http://heart.bm
j.com/
Br H
eart J: first published as 10.1136/hrt.48.5.428 on 1 Novem
referred to the Division of Cardiology, BrookhavenMemorial Hospital Medical Center, for two dimen-sional echocardiography. The age distribution is shownin the Table. The clinical indication for the echocardio-graphic examination covered a wide spectrum ofcardiac disease. Though certain patients were referredwith diagnoses of aortic stenosis and/or regurgitation,none was suspected of having a congenital bicuspidaortic valve in particular on clinical grounds alone.Patients who had smoked the equivalent of 20cigarettes a day for 10 years or more were classified as
smokers. Forty-nine patients were smokers and 51were non-smokers. Weight in relation to height(ponderal index) [height (in)/3N/ weight (lb)] rangedfrom 9-2 to 13-9 (mean 119).
(B) EQUIPMENTUltrasound examination was performed with a com-
mercially available mechanical two dimensionalimaging system (Picker Echoview System 80CI,Northford, Conn.) with a 2-25 MHz transducerimaging 30 and 60 degree sectors. Image rates were 45and 60 images per second when using the 60 and 30degree sectors, respectively. The wide angle was usefulfor spatial orientation and aortic valve plane local-isation, while the 30 degree angle provided betterimage quality and structural detail. Images were
recorded on a 0-5 in videotape via a Sanyo VTC 1000cassette recorder (Sanyo Corporation, Compton, CA)for analysis.in real time, slow motion, and single frameformat. Individual frames were photographed on
Polaroid film (Polaroid Corporation, Cambridge, Ma).
(C) TECHNIQUEAll examinations were performed in the supine and leftlateral decubitus positions by one of the authors (MZ).Parasternal short axis views of the heart and greatarteries were obtained by directing the beam in a planeperpendicular to the long axis of the left ventricle andtilting the transducer superiorly. An apical fourchamber view was obtained by placing the transducerat the point ofmaximal impulse, with the patient in theleft lateral decubitus position and directing the ultra-sonic beam from apex to base, perpendicular to theplane of the atrial and ventricular septa. On tilting thetransducer slightly more anteriorly, the aortic valvecould be recorded.'5 Visualisation of aortic leaflet
429
commissures by this latter technique was systematic-ally attempted only in subjects in whom adequate para-sternal short axis imaging was not technically possible.The aortic valve was examined from multiple inter-spaces and the transducer carefully swept through theentire space since cardiac excursion often temporarilyremoved the aortic valve from the plane of visual-isation. Both parasternal short axis and tilted apicalfour chamber echocardiographic views could often berecorded more easily with held expiration; in somepatients, however, apical images were obtained bestduring held inspiration. Real time visualisation ofvalvemotion was useful, inasmuch as it was easier torecognise structures and their interrelations as theymoved through cyclical excursion than if seen only instatic display. The static images, moreover, weredegraded optically because stop action pictures showonly a single field of a complete two field video frame.The aortic root images were reviewed for location ofcommissural edges, number of cusps visualised, andshape of the aortic valve orifice in systole. Still framesfor display as well as data for statistical analyses werederived only from diastolic echocardiographic images.
(D) STATISTICSThe two tailed Student's t test was used to determinethe significance of differences in mean ± standarddeviations between two groups of subjects. The effectof certain modifying factors upon adequate visual-isation of aortic valve leaflets was determined using x2test uncorrected tor continuity.'6 Multiple regressionanalyses were not performed. An alpha level of 005was established as the level of significance.
Results
(A) PARSTERNAL SHORT AXIS VIEWTwo dimensional echocardiography disclosed threeapparent commissures (diastolic "Y" configuration)(Fig. 1A) in 26 patients, suggesting a tricuspid aorticvalve.'3 15 Two commissures (diastolic "V" configur-ation) (Fig. lB and C) were imaged in an additional 25subjects. Only a single commissure ending at a centralnodulus Arantii was seen in diastole in four cases (Fig.2A). Either two or three sides of the previouslydescribed "inverted triangle configuration"'5 (Fig. 3systole) were noted in systole in all the above cases. Onepatient had a bicuspid aortic valve diagnosed by bothsystolic and diastolic leaflet configuration'3 15 (Fig. 4Aand B), and had an aortic ejection sound on phono-cardiography and only two sinuses of Valsalva visibleon supravalvular aortography.
(B) FACTORS AFFECTING TECHNICALFEASIBILITY OF PARASTERNAL SHORT AXIS VIEW
In 73 of the 100 patients, determination of the numberof cusps was not possible because (1) failure to obtain
on March 23, 2020 by guest. P
rotected by copyright.http://heart.bm
j.com/
Br H
eart J: first published as 10.1136/hrt.48.5.428 on 1 Novem
Fig. 1 (A) Parasternal short axis view diastolicframe showing three aortic cusps (" Y"configuration); (B) and (C) parasternal short axisview diastolic frames showing "V" configuration(two aortic leaflet commissures). R, right coronarycusp; N, non-coronary cusp; L, left coronary cusp;NA, noduli Arantii.
A
Equal cusps NAand NLcomm ssures
Diastole
B
Unequal cuspsnnd equalc ommissures
NA
Diastole
Fig. 2 (A) Parasternal short axis view diastolic frame showingonly a single commissure. NA, nodulus Arantii; (B) parasternalshort axis view diastolic frame showing highly calcified aorticvalve cusps.
an adequate tomographic short axis view at the level ofthe aortic valve in 36 patients; (2) though a technicallyadequate tomographic section was obtained optimalvisualisation not realised in 29 patients (Fig. iB, IC,and 2A); and (3) the valve cusps densely calcified ineight patients (Fig. 2B). Technically adequate para-sternal short axis imaging was more likely in non-smokers (37/51 (73%) vs 27/49 (55%) (p<0O05)) and inyounger patients (48-9±18-0 y vs 57-2±13-0 y)(p<0 05). Sex (37/64 (58%) women vs 20/36 (56%)women (NS)) and weight in relation to height(12-+1±0) vs (11-9±0-8) (NS) were not factors.
Unequal cuspst N NA
aind N
Diastole Systole
Fig. 3 Diagram ofequal and unequal sized aortic valve cusps;on the right the aortic valve is opened. R, right coronary cusp; N,non-coronary cusp; L, left coronary cusp; NA, noduli Arantii.
(C) ANTERIORLY TILTED APICAL FOUR CHAMBERVIEWIn 40 subjects in whom maximally only one aorticleaflet commissure was imaged from the parasternalechocardiographic window, an anteriorly tilted apicalfour chamber view disclosed a diastolic "Y" configur-ation in eight (Fig. 5A), suggesting a tricuspid aorticvalve. A diastolic "V" configuration with visualisationof only two commissures was seen in five subjects (Fig.5B). In eight additional subjects only a partial "V"'configuration was imaged. The noduli Arantii and verycentral portions of the commissural edges of the valve
430
A(r z25i
AI
Systde
Systole
on March 23, 2020 by guest. P
rotected by copyright.http://heart.bm
j.com/
Br H
eart J: first published as 10.1136/hrt.48.5.428 on 1 Novem
Fig. 4 (A) Parasternal short axis view diastolic frame ofbicuspid aortic valve; (B) parasternal short axis view systolicframe ofnon-stenotic bicuspid aortic valve. A, anterior aorticleaflet; P, posterior aortic leaflet.
leaflets were not often visualised. The angle formed bythe imaginary medial extension of portions of thecommissures imaged (Fig. 5C) often exceeded thecommissural angle seen in the same patient, using theconventional parasternal short axis approach (Fig. IB).
Discussion
A bicuspid aortic valve occurs in over 50% of patientswith anatomically isolated aortic stenosis and in about25% of patients with clinically pure aortic regurgi-tation.2 It is not known why one congenitally bicuspidvalve becomes severely scarred, calcified, and stenotic;another only mildly scarred and regurgitant; another
NA
A(In = 8
Bn=5)
the site of infection; and another remains free of thesecomplications. Therefore, it is probably advantageousfor purposes of adequate follow-up and progress todetect all such valves as early in life as possible.Similarly, it would be very helpful to have a means ofeffectively excluding the presence of this commoncongenital cardiac defect.
Although the presence of an aortic ejection sound onauscultation is suggestive of a non-stenotic bicuspidaortic valve,'7 such sounds may be heard in cases ofaortic root dilatation of any aetiology,'8 and whenconfined to the cardiac apex may easily be confusedwith late tricuspid valve closure. '7M-mode echocardiography has previously been
reported as specific and reasonably sensitive in thedetection of a bicuspid aortic valve.'7 19 Use of themaximum eccentricity index for accurate indentifi-cation of abnormal valve anatomy has, however,recently been questioned.13 Transducer angulationcannot be precisely standardised, and transducerposition and angulation on the chest wall may bedictated by differing thoracic shapes and alteredorientation of cardiomediastinal structures. Because ofthis, the M-mode ultrasonic beam may intersect thebicuspid aortic valve cusps at successive locations,yielding widely variable values for the eccentricityindex.
In cases of angiographic and/or surgically provedbicuspid aortic valve, technically optimal two dimen-sional echocardiographic parasternal short axis imagesat the level of the great arteries confirmed the presenceof only two aortic leaflets opening and closing duringthe cardiac cycle in 65% (28/43),20 94% (33/35),21 100%(13/13),13 and 100% (11/11)'1 of cases in four reportedseries. Data also suggest that diastolic images havegreater sensitivity than systolic short axis images in thedetection of aortic valve leaflet number.'5 Systolic
Fig. 5 (A) Aortic valve with "Y"diastolic configuration as oftenincompletely imaged on anteriorlytilted apical four chamber view; (B)aortic valve with "V" diastolicconfiguration on anteriorly tilted apicalfour chamber view; (C) aortic valvewith "V" diastolic configuration asoften incompletely imaged onanteriorly tilted apicalfour chamberview. R, right coronary cusp; N, non-coronary cusp; L, left coronary cusp;NA, noduli Arantii; IVS,interventricular septum; LA, leftatrium; LV, left ventricle.
C(n = 8i
431
on March 23, 2020 by guest. P
rotected by copyright.http://heart.bm
j.com/
Br H
eart J: first published as 10.1136/hrt.48.5.428 on 1 Novem
valvular "doming" noted on two dimensional echo-cardiographic parasternal long axis imagng is present inonly 46% (6/13), 3 (16/35)21 of angiographic and/orsurgically proven cases of bicuspid aortic valve, and isconfined to those cases with peak systolic transvalvulargradient greater than 33 mmHg.13 The echocardio-graphic finding of valvular doming, moreover, is notspecific for bicuspid aortic valve and is present insubjects with trileaflet aortic valvular stenosis and asignificant transvalvular gradient. 3 The usefulness oftwo dimensional echocardiography in the diagnosis ofbicuspid aortic valve appears, therefore, to bedependent upon the frequency with which thecommissural margins of the aortic valve cusps can beimaged during diastole using the parasternal short axisview.Our results indicate that two dimensional echo-
cardiographic recording of a parasternal short axis viewat the level of the great arteries could indentify theprobable number of valve leaflets using previouslydescribed echocardiographic criteria'3 15 in only 27% ofsubjects. In addition, false positive and negativediagnoses of a bicuspid aortic valve have been reportedpreviously using these criteria. 13 20 Our figure of27% issomewhat lower than that reported in the only otherlarge series of prospectively examined subjects'5 andmay be because there are more subjects with denselycalcified valves and/or more smokers among ourpatients. Age and smoking history were important vari-ables affecting the technical quality of the parasternalshort axis images obtained. This may be attributable tochanges in the thoracic skeleton (calcification) and alsoto hyperinflation of the lungs (increased residualvolume) which occurs with both age22 and cigarettesmoking.23 Though the equipment used in this studyproduced a relatively good image, use of the latestgeneration ultrasonography devices, with higherfrequency transducers, might well have resulted ingreater sensitivity in the detection of aortic valve cuspmorphology because of superior spatial resolution.
Systematic use of an anteriorly tilted apical fourchamber view for aortic valve leaflet imaging washelpful in subjects who smoked and did not have aparasternal echocardiographic window which wastechnically adequate. This has not, however, been ourexperience when applying this technique to non-smokers and subjects with an adequate parasternalechocardiographic window. In these subjects, atangential section of the aortic root and cusps almostalways results, which does not allow accurate identifi-cation of cusp morphology. For this reason, our studyprocedure systematically used the anteriorly tiltedapical four chamber view only in subjects in whom aconventional parasternal short axis view at the level ofthe great arteries was not technically good enough. Inthese cases, perhaps, cardiac rotation, possibly related
Zema, Caccavano
to chronic obstructive pulmonary disease, may havepermitted adequate short axis imaging ofthe aortic rootfrom the apical position.
In many other cases, the commissures between thenon-coronary and left coronary cusps were not imaged.On the basis of previously reported pathologicaldata,2 "I the "V" configuration formed by the remain-ing commissures and the noduli Arantii (n= 25) shouldbe diagnostic of a tricuspid aortic valve, regardless ofwhether equal cusps and commissures, unequal cuspsand equal commissures, or unequal cusps and com-missures are present (Fig. 3A-C). In the latter case, itis essential to obtain an exact cross-section of the aorticroot and to angle the beam slightly in all directions so asto find the miniimal angle present between the twolimbs of the imaged "V". If this is not done, this typeof tricuspid aortic valve could be misdiagnosed asbicuspid. Since the raphe of a congenitally bicuspidvalve rarely extends to the free margin of the conjoinedcusp or to the cephalid level reached by the two truecommissures,220 visualisation of the noduli Arantii,which are often echo reflective relative to the con-tiguous commissural margins of the aortic valve cusps,should also help in establishing the tricuspid nature ofthese valves.
Considering the high prevalence in our population ofincomplete two dimensional echocardiographic aorticvalve leaflet images, angiographic and/or pathologicalstudies must be done to establish the correlationbetween these incomplete echocardiographic patternsand true aortic valve anatomy if two dimensional echo-cardiography is to have widespread application in boththe detection and the exclusion of the congenitalbicuspid aortic valve.
8 Smith DE, Matthews MB. Aortic valvular stenosis withcoarction of the aorta with special reference to thedevelopment of aortic stenosis upon congenital bicuspidvalves. BrHeartj 1955; 17: 198-206.
on March 23, 2020 by guest. P
rotected by copyright.http://heart.bm
j.com/
Br H
eart J: first published as 10.1136/hrt.48.5.428 on 1 Novem
9 Bacon APC, Matthews MB. Congenital bicuspid aorticvalves and the aetiology ofisolated aortic valvular stenosis.QJ' Med 1959; 28: 545-60.
10 Campbell M. Calcific aortic stenosis and congenitalbicuspid aortic valves. BrHeartj 1968; 30: 606-16.
11 Roberts WC. The structure of the aortic valve in clinicallyisolated aortic stenosis; an autopsy study of 162 patientsover 15 years of age. Circulation 1970; 42: 91-7.
20 Brandenburg RO Jr, Tajik AJ, Edwards WD, Reeder GS,Shub C, Seward JB. Accuracy of two-dimensional echo-cardiographic diagnosis of bicuspid aortic valve: echo-cardiographic-anatomic correlative study in 115 patients(abstract). Am J Cardiol 1982; 49: 1040.
21 Gondi B, Nanda NC. Two-dimensional echocardiographyin bicuspid aortic valves: usefulness and limitations(abstract). Clin Res 1981; 29 (2): 197A.
22 Richards DW. Pulmonary changes due to aging. In: FennWO, Rahn H, eds. Handbook of physiology. A criticalcomprehensive presentation of physiologic knowledge andconcepts. Section 3: Respiration Vol. 2. Washington, DC:American Physiologic Society, 1965: 1525-9.
23 Ingram RH Jr. Chronic bronchitis, emphysema andchronic airways obstruction. In: Isselbacher KJ, AdamsRD, Braunwald E, Petersdorf RG, Wilson JD, eds.Hamsons princples of internal medicine. 9th ed. New York:McGraw-Hill, 1980: 1239.
Requests for reprints to Dr Michael J Zema, Divisionof Cardiology, Department of Medicine, BrookhavenMemorial Hospital Medical Center, 101 HospitalRoad, Patchogue, New York 11772, USA.
433
on March 23, 2020 by guest. P
rotected by copyright.http://heart.bm
j.com/
Br H
eart J: first published as 10.1136/hrt.48.5.428 on 1 Novem
