Incremental value of biplane and multiplane transesophageal echocardiography for the assessment of active infective endocarditis

Incremental Value of Biplane and Multiplane lkansesophageal Echocardiography for the Assessment of Active Infective Endocarditis Frank P. Job, MD, Stefanie Franke, MD, Harald Lethen, MD, Frank A. Flachskampf, MD, and Peter Hanrath, MD In 41 patients with clinical evidence of active infective endocarditis, transesophageal echocardiography was performed in a stepwise manner, star-tin with evaluation of the mono lane views, followed 9,y the longitudinal plane, an B finall by the intermediate planes. Number, location, lengt it , area, density, extent, and mobility of vegetations and abscesses were assessed in the monoplane, biplane, and best intermediate planes to identify and quantify the incremental value of the longitudinal and intermediate lanes. Eighty-three vegetations and 6 abscesses were Pound. In 4 patients (10%) monoplane evaluation yielded false-negative results. There were no false-negative results using the biplane evaluation. However, when compared with multiplane evaluation, additional vegetations were missed in 23% of patients after monoplane and in 9% of patients after

biplane evaluation. Three abscesses were missed using the monoplane and 1 was missed using the biplane technique. The area was underestimated in 60% of all vegetations (mean underestimation, 37% f 23% [SD] of maximal area) and length in 49% of cases (mean underestimation, 38% 2 23% [SD] of maximal length) of all vegetations when biplane was compared with multiplane evaluation. Also, with monoplane and biplane evaluation, mobility and density were misinterpreted in 6% and 5% and 17% and 9% of all vegetations, respectively. Thus, multiplane transesopha eal echocardiography is more accurate than the monop Yane and biplane techniques in assessing patients with active infective endocarditis. (Am J Cardiol 1995;75: 1033-l 037)

he advent of the transesophagealtechnique has subT stantially improved the diagnostic accuracy of echocardiography,resulting in a sensitivity of 9O%.i-‘OHow-

into the study. All patients gave informed consent. No complications or adverseevents were encountered. Echocardiographic methods: Transesophagealechocardiography (TEE) was performed utilizing a multiplane 5 MHz phased-array probe (Hewlett-Packard) according to the following stepwise protocol: First the monoplane evaluation was simulated with the probe at 0” (transverse plane) and valve abnormalities (vegetations, abscesses)were assessed.Then the probe was rotated to 90” (longitudinal plane) to simulate biplane evaluation, and the searchfor valve abnormalities conducted. Finally, images were obtained at the intermediate planes in 10” stepsto complete a multiplane evaluation. At each stage, images had to be obtained with anteroposterior and lateral flexions of the probe.

ever, the reported values are based on experience with monoplane instruments. False-negative results are reported in postmortem studies and in 8% to 13% of patients with initially negative and subsequentlypositive monoplanetransesophagealechocardiographicresults.7,u Recently, biplane12and multiplane13,14transducershave become available. Their potential for improving the detection of endocarditic lesions has not yet been defined. Furthermore, the exact definition of the vegetation length, extent, density,and mobility can be important for an adequate risk stratification and for developing a therapeutic strategy.14J5 This study therefore addressesthe question: how often and to what extent false-negative transesophageal echocardiographic results or underestimation of endocarditic lesions may be expected by monoplane or biplane assessmentwhen comparedwith the multiplane technique?

Evaluation of valwlar or paravalvular abnormalities:

Videotape recordings were presented to 2 blinded observers.To ensure that the observerswere not biased by information from previously reviewed planes, the readings were subdivided into transverse,longitudinal, and intermediate image sequencesand stored on 3 different METHODS tapes.Furthermore, the sequenceof the patients was ranStudypatients: During an B-month period, all patients domly changed.Each reviewer was asked to frrst evalureferred to our echocardiography laboratory with clear ate the tape with the images acquired at 0”. The results clinical evidence of infective endocarditis (typical his- were documented and the procedure was repeated with tory, fever, murmur, positive blood culture) were entered the longitudinal images (90’ rotation of the transducer), and finally with all the intermediate planes. Four weeks after the first evaluation, the samereviewers were asked From the Medical Clinic I, Rheinisch Westfalische Technische Hochschule Aachen, Aachen, Germany. Manuscript received No to interpret the studies again, this time starting with the vember 29, 1994; revised manuscript received and accepted Feb multiplane images. Intra- and interobserver variances ruary 2 1, 1995. were calculated. Address for reprints: Frank P. Job, MD, Medical Clinic I, Rheinisch Characterization of vegetations: Vegetations were deWestfaltsche Technische Hochschule Aachen, Pauwelsstrasse 30, D 52057 Aachen, Germany. fined as discrete massesof echogenic material adherent VAlVULAR

HEART DISEASE/ASSESSMENT

OF ENDOCARDITIS

WITH

MULTIPLANE

TEE

1033

to the surface of the leaflet and distinct in characterfrom the remainder of the leaflet. Maximal length was assessed by measuring the vegetation from the point of adherence to the leaflet surface or to paravalvular structures. If the vegetations were totally adherent to the leaflet, the maximal diameter of the vegetation was measured. Becauseit is recognized that a vegetation of >lO mm bearsan increasedrisk of seriouscomplications (congestive heartfailure, emergencyvalvereplacement,death),15J6 all vegetations >lO mm detected by multiplane examinations were correlated to the lengths assessedin the monoplane and biplane images to identify the number of underestimated critical vegetations. The frequency of vegetations >lO mm by multiplane assessmentand underestimated ~10 mm by monoplane or biplane assessment, or both, was thus determined. The area of each vegetation was measuredby planimetry. Mobility, extent, and consistency were each graded on a scale of 0 to IV The mobility score was: grade 0, no vegetation detectable; grade I, fixed vegetation with no independent motion; grade II, fixed base but with mobile free edge; grade III, pedunculated vegetation; and grade IV, prolapsing vegetation, crossing the coaptation point of the leallet at somepoint during the cardiac cycle. The extent of a vegetation was scored as: grade 0, no vegetation detectable; grade I, single vegetation; grade II, multiple

RESULTS Within the 18-month study period, 41 patients (mean age 53 + 13 years) were included in the study. Six patients, all with vegetations, also had abscesses.Fortymultiplane

multiplane

0

vegetation, limited to a single valve leaflet; grade III, involvement of multiple leaflets; and grade IV, involvement of extravalvular structures.The scorefor the density of a vegetation was: grade 0, no vegetation detectable; grade I, completely calcified vegetation; grade II, partially calcified vegetation; grade III, denserthan myocardial structures, not calcified; and grade IV, consistency equal to or below myocardial echoes. Characterization of abscesses:An abscesswas defined as an inhomogeneousecholucent spacein the neighborhood of a valve different from the surrounding normal cardiac tissues. Area and the maximal diameter of the cavity were assessedin the different planes. Statistics:Results were analyzed for statistical significance using the chi-square test. Linear regression analysis was performed to correlate vegetation size and area assessedby monoplane, biplane, and multiplane evaluation. Intraobserver variances for observer I and II for length-area measurementswere 6 + 1% (area 5 +: 1) and 5 f 2% (area 5 rf: 2); interobserver variance was 10 it 2% (area 10 r 3).

1

I 2

3

monoplane

I

I

’

[cm] ’

multiplane

0

1

2

biplane a

’

[cm]’

multiplane 9.6

km4 3

1

biplane”

2 [cm2T

FIGURE 1. Correlation of vegetation length between monoplane, biplane, and multiplane evaluation. A, monoplane versus multiplane: Y = 1.20x + 0.2; r = 0.82 (SDR 0.33); p ~0.0001. B, biplane versus multiplane: Y = 0.90x + 0.35; r = 0.90 (SDR 0.35); p
1034 THE AMERICAN JOURNAL OF CARDIOLOGY@

VOL. 7.5

MAY 15, 1995

three valves were affected, with a total number of 83 tional 15% and 19% of mitral and aortic vegetations, vegetations (Table I). All patients had fever >38”C and respectively, to be identified as >10 mm. The remaining 95% also had murmur of the heart. In 30 patients, pos- 10% of mitral vegetations >lO mm were identified coritive blood cultures were obtained (35% streptococcus, rectly only when the views from intermediateplaneswere 30% staphylococcus, 30% enterococcus subtypes; 5% available (Figure 2). When all the vegetations assessedby the biplane Haemophilus injluenzae). Nine of the 41 infected valves (22%) were prosthetic valves. The embolic event rate in technique were compared with those assessedby the the study collective was 23%, and the rate of valve re- monoplane technique, 6%, 12%, and 15% were graded placement within 6 weeks after the inclusion into this differently in extent, mobility, and density, respectively, study was 51%. The mitral valve was slightly more affected than the aortic valve. Mitral vegetations were preTABLE I Affected Valves and Number and Side of Attachment dominantly attached to the atria1 side, whereas 36% of of Vegetations the aortic vegetations were located upstream at the aorNo. (“A) tic side of the valve. Affected valves 43 (100) In 4 of the 41 patients, vegetations could not be deMitral 22 (51) tected by the monoplane examination, but could be deAortic 18 (42) tected by the biplane or multiplane examinations. One Tricuspid 3 (7) aortic vegetation (length 0.45 cm, area 0.05 cm2) shadNumber‘of vegetations 83 (100) owed by a mitral prosthesis, 2 mitral vegetations (length Mitral valve 47 (57) Aortic valve 30 (36j 0.37 cm, area 0.05 cm2; length 0.6 cm, area 0.08 cm2) Tricuspid valve 6 (7) on the atria1 side of native valves, and 1 aortic vegetaSide of attachment tion (length 1.05 cm, area 0.42 cm2) on the aortic side Mitral valve (total) 47 (100) of the valve were missed on the monoplane examinaAtrial side 41 (87) Ventricular side tion. Thus the overall sensitivity for the detection of veg4 (9) Both sides 1 (2) etations by the monoplane versus the multiplane techAortic valve (total) 30 (100) nique is 90.25%. Ventricular side 16 (53) The total number of vegetations assessedin this study 11 (37) Aortic side by the monoplane, biplane, and multiplane techniques Both sides 3 (101 Tricuspid valve was 5868, and 82, respectively. Thus, in someinstances Atrial side 6 (100) additional vegetations were missed with the monoplane and biplane techniques, although there were no false-negative results in the TABLE II Localization, Diameter, and Area of Abscesses biplane examinations: 29% of the vegMonoplane Biplane Multiplane etations seenwith the multiplane technique were not detectedwith the monoAbscess Valve Diam. Area Diam. Area Diam. Area plane technique. Suchvegetationswere Number Position WI (4 (cm21 (4 b-4 (cm? mostly small and appearedas 1 unique 1.08 1.97 1.08 2.25 1.16 1 Aortic 1.97 vegetation in the monoplane views, 1.15 1.46 1.15 1.46 1.15 2 Mitral 1.46 whereas the intermediate planes re3.22 3.74 3.22 3.74 3.61 3 Aortic prosth. 3.74 1.09 0.21 1.09 0.21 vealed that the structure consisted of 4 Aortic NS 0.96 0.18 0.73 0.30 5 Aortic NS >l vegetation. The frequency of such NS 0.69 0.24 6 Mitral NS “missed additional vegetations” was Diom. = diameter; NS = not seen; prosth. = prosthesis. 17% when the biplane was compared with the multiplane views. Both area and length were systematically underestimated by monoplane and biplane aortic valve mitral valve assessment.Monoplane and biplane 0 deyee lengths were smaller than multiplane 44/o 0 degree ---, lengths by 30 + 21.0%and 25 +: 17.4%, respectively; monoplane and biplane areas were smaller than multiplane 0 + 90 deg 19% areasby 47 r 21.2% and 32 f 15.5%, respectively. There was a good corre90 degree lation between the monoplane and 15% multiplane biplane, the monoplane and multi37% 10% plane, and the biplane and multiplane measurements(Figure 1A and 1B). Twenty-five percent of mitral and FIGURE 2. Number and distribution of v etations with a true length ~10 mm after multiplane evaluation, but underestimate 7 as lO mm by evaluation. 6/a& section, vegetation length z-10 mm assessed by monoplane; wide multiplane TEE were estimatedas < 10 stripes, vegetation length 210 mm assessed by biplane, but not by monoplane, mm by the monoplane technique. The transeso ha eal echocardiography; narrow stripes, vegetation length ~10 mm biplane technique allowed an addi- assessecf on Py by multiplane transesophageal echocardiography. VALVULAR

HEART DISEASE/ASSESSMENT

OF ENDOCARDITIS

WITH

MULTIPLANE

TEE

1035

using the ,monoplane technique. When assessmentsby the multiplane technique were compared with those by the monoplane technique, ll%, 15%, and 17% of vegetations were scored differently in extent, mobility, and density, respectively, using the monoplane technique. The advantage of multiplane over biplane examination was 5% for the extent, 9% for mobility, and 6% for density in all vegetations. Abscesses:Table II lists localization, diameter, and area of the abscesses.When compared with multiplane examination, 3 abscesseswere missed after monoplane and 1 after biplane examination; this led to a sensitivity

of 50% for the monoplane and 83% for the biplane technique when comparedwith the multiplane technique. All of the patients with abscessesalso had vegetations that were detectable with the monoplane technique.

DISCUSSION

Echocardiographyhasimproved the clinical detection of endocarditis by directly visualizing valve abnormalities, and the development of TEE in-its various modalities allows a detailed, morphologic description of the infection. The prognosisof the diseasedependson an early diagnosisof valvular abnormalitiesandlocal comulications suchasparavalvularleakageor abscesses. In this study, imaging of endocarditic lesionsby monoplane and biplane TEE was compared with multiplane imaging, which incorporates transverse and longitudinal planes and thus can serve as a reference standard. From a clinical viewpoint it is intriguing to know (1) how often the diagnosis of endocarditis was missed altogether by monoplane and biplane imaging, implying that no vegetation or abscess was seenin a patient who had 11 lesion by multiplane TEE, and (2) how well the number and the morphologic characteristics of the lesions were assessedby monoplane and FIGURE 3. Left: Aortic v etation shadowed by a bioprosthesis (BIO) in the aortic position. biplane imaging compared Right: Vegetation (arrow4 can be detected in 116” multiplane view. A0 = aorta; LA = left with multiplane TEE. atrium; LV = left ventricle; MV q mitral valve; RA = right atrium.

FIGURE 4. Tricuspid valve endocarditis. and right, 90” (vegetations on anterior

1036

I

Ve et&ions shown at: left 0” (1 vegetation); middle, 73” (vegetations on all 3 leaflets]; an cl lateral leaflet). RV = right ventricle; other abbreviations as in Figure 3.

THE AMERICAN JOURNAL OF CARDIOLOGY”

VOL. 75

MAY 15, 1995

For the lirst time, this study comparesthe 3 modalities to analyze the incremental value of additional planes for assessingendocarditic lesions in patients with clinically suspectedendocarditis. Our findings show that at least 1 endocarditic lesion was seenwith biplane imaging in all patients with a multiplane diagnosis of endocarditic lesions, but in only 90.25% of patients with monoplane imaging. Thus, about 10% of these patients would have been diagnosed as free of endocarditic lesions by monoplane TEE. The number of individual vegetations detectedby monoplane and biplane imaging was considerably lower: only 71% and 83% of the vegetations seen by multiplane imaging were detected by monoplane and biplane imaging, respectively, thus underestimating the severity of the disease.Both monoplane and biplane imaging also led to a significant underestimation of lesion length, area, mobility, and extent. All missed vegetations measuredcl.1 cm in length. Figure 3 is an example of a small aortic vegetation that was shadowed by a mitral bioprosthesis in the transverse plane. It could also not be detected in the longitudinal plane, but was easily detectable in the 116” plane after multiplane assessment.In a larger number of patients, a secondor third vegetation was missed without the intermediate planes.Twenty-nine percent of such vegetations were not detectable after monoplane examination and 1% was not detectable after biplane examination. Figure 4 depicts a typical exarnljle of the incremental value of the longitudinal and intermediate planes for assessing tricuspid valve endocarditis. Some of these multiple vegetations were misinterpreted as 1 single vegetation. Although many of these vegetations were mobile, their density was equal or below myocardial density, which might account for their diminished visibility. Further, monoplane and biplane evaluation in 25% to 39% of all vegetations led to an underestimation at a range of 37% and 40% of their true length. Although vegetations of a critical length (>lO mm) were usually identified by all 3 modalities, 20% of those attachedto the aortic valve and 4% attached to the mitral valve were underestimated. Extent, mobility, and density were underestimated in 11% to 17% of all vegetations when the measurements by the monoplane technique were comparedwith those by the multiplane technique. Underestimation was only 5% to 9% of all vegetations using the biplane technique when compared with the multiplane technique. Although the total number of abscessesas serious complications of infective endocarditis is low, 3 were not seenafter monoplane evaluation. Both reviewers discovered only 2 of the 3 abscesseswhen the longitudinal planes were available. All of them were confirmed subsequently,2 intraoperatively and 1 by autopsy. Study limitations: A limitation of the study is the fact that data were acquired with 1 multiplane probe at 1 examination by 1 examiner. Becauseit was ethically not acceptableto perform multiple examinations in 1 patient, we tried to compensatefor this bias by 2 experienced independent reviewers who were blinded to the rest of the examination while reading the monoplane or biplane images and who did not know the results of the other reviewer. Intra- and interobserver varianceswere accept-

ably low. Nevertheless,this design may have resulted in some decreased accuracy for the biplane technique. Since independent pathologic data are not available in most study patients, multiplane TEE had to serve as the reference standard with a certain risk of a false-positive result despite the typical clinical course of the disease. Conclusions: Monoplane TEE has improved the diagnostic sensitivity for detecting endocarditic vegetations to about 90%, which is already very high. Our findings of a sensitivity of 90.25% correspond well to previous published studies.3-8Data from studies with the biplane technique in that field are very rare. Our finding is that biplane imaging primarily underestimatesthe extent of the disease. However, the diagnosis of “endocarditis” itself is usually not missed and clinical decision making in such caseswill always consider aspectssuch as coexisting valve incompetence, clinical status of the patient, and responseto antibiotic treatment. This is also true if we take into account that 1 small, surgically contirmed abscesswas overlooked by the biplane technique. In that particular case,5 vegetations with a moderate to severe valve incompetencewere also found, indicating an extent of the diseasethat made surgery inevitable.

1. Erbel R, Rohmann S, Drexler M, Mobr KahaIy S, Gerharz CD, Iversen S, Oehlert H, Meyer J. Improved diagnostic value of echocardiography in patients with infective endocarditis by tramesophageal approach. A prospective study. Eur Heart .I 1988;53:43-53. 2. Miigge A, Daniel WG, Frank G, Lichtlen PR. Echocardiography in infective endocarditis: reassessment of prognostic implications of vegetation size determined by the transtboracic and the transesophageal approach. JAm Coil Cardiol 1989;14: 631-638. 3. Daniel WG, ScbrSder E, Nonnast-Daniel B, Lichtlen PR. Conventional and transesophageal echocardiography in the diagnosis of infective endocarditis. Eur Heart J 1987;8:287-292. 4. Daniel WG, Schreder E, Miigge A, Lichtlen PR. Tramesophageal echocardiography in infective endocarditis. Am J Card Zmq 1988;2:78-85. 5. Sbively BK, Gurule FT, Roldan CA, Leggett JH, Scbiller NB. Diagnostic value of tramesophageal compared with transthoracic echocardiography in infective endocarditis. .I Am Coil Cardiol 1991;18:391-397. 6. Birmingham GD, Rahko PS, Ballaniyne F III. Improved detection of infective endocarditis with transesophageal echocardiography. Am Heart J 1992,3:774-781. 7. Daniel WG, Miigge A, Martin RP, Lindert 0, Hausmann D, Nonnast-Daniel B, Lass J, Lichtlen RP. Improvement in the diagnosis of abscesses with endocarditis by transesophageal echocardiography. N EngZ .I Med 1991;32&795-800. 8. Pedersen WR, Walker M, Olson JD, Gobel F, Lange HW, Daniel JA, Rogers J, Lange T, Kane M, Mooney MR, Goldenberg IF. Value of transesophageal echocardiography as an adjunct to transthoracic echocardiography in evaluation of native and prosthetic valve endocarditis. Chest 1991;100:351-356. 9. Buda AJ, Zotz RJ, Lemire MS. Prognostic significance of vegetations detected by two-dimensional echocardiography in infective endocarditis. Am Hearz J 1986; 112:1291-1298. 10. PI&n JF. The evolving role of echocardiography in management of bacterial endocarditis [editorial]. Chest 1988;96:904-908. 11. Sochowski R, Chan K-L. Implication of negative results on a monoplane transesophageal echocardiographic study in patients with suspected infective endocarditis. JAm Coil Cardiol 1993;21:216-221. 12. Seward JB, Khrmderia BK, Edwards WD, Oh JK, Freemann WK, Tajik AJ. Biulanar transesoohageal echocxdiozraohv: anatomic correlations. imane orientatidn, and clinical ippications. Mayobk &c 1990;65:1193-1213. 13. Flachskampf FA, Hoffmann R, Verlande M, Ameling W, Hanrath P. Initial experience with a multiplane transesophageal echotransdu&r: assessment of diagnostic potential. Eur Heart J 1992;13:1201-1206. 14. Roelandt JRTC, Thompson IR, Vlettter WB, Brommersmar P, Bomn N, Linker DT. Multiplane transesophageal echocardiography: latest evolution in an imaging revolution. J Am Sot Echocardiogr 1992;5:361-367. 15. Sanfilippo A& Picard MH, Newell JB, Rosas E, Davidoff R, Thomas JD, Weymamt A. Echocardiographic assessment of patients with infectious endocarditis: prediction of risk for complication. J Am Cull Cardiol 1991;18: 1191-l 199. 16. Lutas EM, Roberts RB, Devereux RB. Relation between the presence of endocarditic vegetations and the complication rate in infective endocarditis. Am Hearf J 1986;112:107-113.

VAtVUtAR HEART DISEASE/ASSESSMENT OF ENDOCARDITIS WITH MULTIPLANE TEE

1037