Multicenter Evaluation of SonoVue forImproved Endocardial Border DelineationNavin C. Nanda, M.D.,* Daniel C. Wistran, M.D.,† Ronald P. Karlsberg, M.D.,‡Terrence C. Hack, M.D.,§ William B. Smith, M.D., David A. Foley, M.D.,#
Michael H. Picard, M.D.,** and Bruno Cotter, M.D.††
*University of Alabama, Alabama, USA, †Cardiology Physicians Inc., Salem, Massachusetts;‡Cardiovascular Research Institute of Southern California, Beverly Hills, California; §PrimaryCare Cardiology Research, Ayre, Massachusetts; New Orleans Center for Clinical Research,New Orleans, Louisiana; #Mayo Clinic, Rochester, Minnesota; **Massachusetts GeneralHospital, Boston, Massachusetts; ††University of California at San Diego, California
Objectives: Two multicenter studies were conducted to evaluate the safety and ef�cacy of SonoVue asa contrast agent for enhanced left ventricular endocardial border delineation (LVEBD), and tocompare the ef�cacy of SonoVue and Albunex in adult patients with a suboptimal, nonenhancedechocardiogram. Background: The use of contrast to enhance echocardiographic assessment ofLVEBD is well-established. SonoVue is a new microbubble contrast agent that contains sulfurhexa�uoride. Methods: Patients were randomized to receive four injections of SonoVue (0.5, 1, 2, and4 ml), or two injections of Albunex and two injections of hand-agitated saline (0.08 and 0.22 ml/kg).Echocardiographic images were evaluated at the study centers and by four blinded, offsite reviewersfor degree of left ventricle opaci�cation (LVO), duration of contrast enhancement, and LVEBD.Results: LVO scores were signi�cantly higher for all doses of SonoVue. Patients with complete LVOranged from 34%– 87% for SonoVue and from 0%–16% for Albunex. The mean duration of usefulcontrast effect ranged from 0.8–4.1 minutes for SonoVue and 15 seconds for Albunex. Meanincreases in LVEBD scores ranged from 3.8 –18.2 for SonoVue and 0.1–4.3 for Albunex. SonoVue(cumulative 7.5 ml dose) was well-tolerated, with a safety pro�le similar to that observed in thecontrol group. Conclusions: SonoVue is superior to Albunex for improving visualization of endocar-dial borders in patients with suboptimal noncontrast echocardiograms. Optimal increases inLVEBD, LVO, and duration of useful contrast effect were observed at the 2.0 ml dose of SonoVue.(ECHOCARDIOGRAPHY, Volume 19, January 2002)
Delineation of cardiac borders during echo-cardiographic examination is of paramount im-portance in the determination of global andregional left ventricular (LV) systolic perfor-mance and, consequently, in the evaluation,treatment, and follow-up of patients with cor-onary artery disease. In many patients, how-ever, poor anatomic visualization either pre-cludes or reduces con�dence in the evaluationof LV function.1-3 Echocardiographic detection
of the endocardial border in the multiple imag-ing planes required for accurate assessment ofLV function using Simpson’s rule may be sub-optimal in up to 50% of patients.4,5 A number ofintravenous contrast agents that contain stabi-lized gas microbubbles have been developed foruse during echocardiography. Gas-�lled micro-bubbles are effective echocardiographic con-trast agents due to their acoustic propertiesand their ability to transit the pulmonary cir-culation. Use of these agents produces opaci�-cation of the left ventricle and improves borderdelineation during echocardiography at restand during stress.6-11
SonoVue (Bracco Diagnostics Inc., Princeton,NJ, USA) is a new microbubble contrast agentthat consists of sulfur hexa�uoride (SF6)-�lledmicrobubbles in a phospholipid shell.12 Sulfur
This work was supported by Bracco Diagnostics Inc.,Princeton, New Jersey.
Address for correspondence and reprint requests: NavinNanda, M.D., University of Alabama at Birmingham,Heart Station SW-S102, 619 South 19th Street, Birming-ham, AL 35249. Fax: 205-934-6747; E-mail: [email protected]
Reprinted with permission fromECHOCARDIOGRAPHY, Volume 19, No. 1, January 2002
27ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech.Vol. 19, No. 1, 2002
hexa�uoride is a poorly soluble, totally innoc-uous gas that does not undergo metabolismand is eliminated in expired air. SonoVue hasbeen evaluated for cardiac and other vascularultrasound applications.13-16 We report the re-sults of two multicenter clinical trials thatevaluated the safety and ef�cacy of SonoVuerelative to that of Albunex (Mallinckrodt Med-ical, Inc., St. Louis, MO, USA) in patients withsuspected cardiac disease and suboptimal en-docardial border delineation (EBD) on unen-hanced echocardiography at rest. The optimaldose of SonoVue for delineation of endocardialborders also was investigated.
Methods
Study Population
A total of 264 adult patients were enrolled intwo multicenter, active-controlled, parallel-group trials conducted using identical protocols(Table I). One hundred thirty-eight patientswere randomized to receive SonoVue and 126were randomized to receive control (Albunexand saline). The study protocol was approvedby the human subjects review board at eachparticipating institution. Written informedconsent was obtained from all patients prior toenrollment.
Patient characteristics are shown for thecombined study populations in Table II. Pa-tients were eligible for study if they were atleast 18 years old, with highly suspected car-diac disease and suboptimal endocardial bor-der detection in the LV (two or more segmentspoorly visualized) on unenhanced echocardio-gram performed within 7 days prior to random-ization in the study. Patients were excluded ifthey did not meet inclusion criteria or if theywere pregnant or lactating, had severe conges-tive heart failure, unstable angina, recent myo-cardial infarction (within 5 days and not stabi-lized), severe arrhythmia, severe pulmonary
hypertension, or known or suspected hypersen-sitivity to blood products.
Study Drugs and Administration
Patients were randomized to receive eitherfour bolus injections of sulfur hexa�uoride mi-crobubbles (SonoVue) at doses of 0.5, 1, 2, and4 ml, or two injections of sonicated human se-rum albumin (Albunex) and two injections ofhand-agitated saline at doses of 0.08 and 0.22ml/kg. Within each study arm, dose order wasrandomized. SonoVue is a sterile, pyrogen-free,lyophilized powder. A white, milky suspensionof sulfur hexa�uoride microbubbles is obtainedby adding 5 ml of physiological saline (0.9%sodium chloride) to the powder, using standardclinical aseptic techniques followed by handagitation. After reconstitution, bubble concen-trations are in the range of 1–5 x 108 micro-bubbles per ml. SonoVue microbubbles rangein diameter from 1–10 mm (mean diameter, 2.5mm; 90% 8 mm).12
Sonicated human serum albumin (Albunex)was chosen as the comparator agent because atthe time the study (1997), it was the only Foodand Drug Administration (FDA)-approvedagent for enhanced echocardiography. The ad-ministered Albunex doses were those recom-mended in the package insert for left heartechocardiography. Due to differences in volumeand color between the two study agents anddifferences in volume between the doses ad-ministered within each study arm, the studydesign was single blind.
Two-Dimensional Echocardiography
At each site, commercially available echocar-diographic equipment and transducers were
TABLE II
Patient Characteristics
SonoVue(n 138)
Control*(n 126)
Mean age (yr) 56.8 13.4 58.7 12.0Male 89 (65) 90 (71)Previous MI 42 (30) 47 (37)Previous heart failure 33 (24) 41 (33)Arrhythmia 38 (28) 49 (39)Angina 51 (37) 37 (29)Hypertension 90 (65) 76 (60)Cardiac surgery 36 (26) 38 (30)
Values are presented as mean SD or number of patients(%). MI myocardial infarction.*Control Albunex and saline.
TABLE I
Study Description
Protocol Number of Sites
Number of TreatedPatients
Total SonoVue Control*
Study A 11 centers 143 76 67Study B 10 centers 121 62 59
*Control Albunex and saline.
NANDA, ET AL.
28 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 19, No. 1, 2002
used for all patients. All patients were imagedin continuous, fundamental B-mode. Patientswere examined in either the supine or lateraldecubitus position. The transmit power was setas low as possible, and the gain setting wasadjusted at the beginning of each procedure toobtain optimal endocardial visualization. Bothsettings were kept constant throughout eachpatient’s evaluation. Every effort was made tokeep the optimal transducer position un-changed during each patient’s entire echocar-diographic study. Images were recorded onsVHS videotape.
A recording of two-dimensional (2-D) trans-thoracic echocardiography was made from 30seconds prior to injection of each dose of studyagent and continued for at least 15 minutesafter injection or until the end of the contrasteffect (return to baseline), which ever waslonger. For both pre- and postinjection images,the apical four-chamber view was recorded�rst, followed by the apical two-chamber view.The exploration of each view was obtained in15–30 second intervals. At least 2 minuteswere allowed to elapse between the disappear-ance of the contrast effect and injection of thenext dose of study agent. The �rst preinjectionechocardiogram had to con�rm suboptimalEBD in the left ventricle.
Ef�cacy Assessment
Echocardiographic images were evaluated atthe individual study centers and by fourblinded, offsite reviewers (two for Study A andtwo for Study B). The offsite readers used inStudy A were investigators in Study B and,conversely, the two offsite readers used inStudy B were investigators in Study A. Theoffsite readers were blinded to patient identity,history, study agent, and dose associated witheach image. Parameters assessed included LVopaci�cation (LVO), duration of useful contrastenhancement, left ventricular endocardial bor-der delineation (LVEBD), and diagnostic con�-dence.
The degree of LVO following each injectionwas graded according to the following four-point rating scale: 0 none (no visible contrastwithin the LV cavity); 1 faint (weak or traceeffect of contrast within the LV cavity); 2moderate (some areas of the LV cavity fullyopaci�ed but without a time when the wholecavity was �lled with contrast to the same highdensity); or 3 complete (homogeneous andhigh intensity effect).
The duration of useful contrast effect wasde�ned as the time, after disappearance ofshadowing effect, from the �rst useful appear-ance of contrast within the LV (i.e., LVO ratingof 2 or 3) until the useful effect ended. For thepurposes of this study, shadowing was consid-ered to be a contrast effect of the injectedagent.
A six-segment model was used to evaluateLVEBD on pre- and postcontrast apical two-chamber and four-chamber views. For eachsegment, LVEBD was graded as follows: 0inadequate (border not visible); 1 suf�cient(border barely visible); or 2 good (borderclearly visible). A total delineation score (0–24)was obtained by adding the scores from the sixindividual segments in each of the two views.
Pre- and postcontrast images were rated bythe offsite readers as to whether they wereconsidered diagnostic, and con�dence in theprimary diagnosis was rated on a scale from 0to 5, where 0 value assigned to nondiagnosticimages, 1 low, and 5 high.
Safety Assessment
Physical examination, Mini-Mental StateExamination (MMSE), and standard clinicallaboratory tests were obtained prior to studyagent administration and were repeatedwithin 1–24 hours after the last injection ofstudy agent. Vital signs and pulse oximetrywere obtained prior to injection and at 5, 10,and 15 minutes after each injection. Twelve-lead electrocardiogram (ECG) was obtained atscreening and at 5 and 60 minutes after thelast injection of study agent. Patients weremonitored for adverse events up to 72 hoursafter the last injection of study agent. In asubset of patients (20 SonoVue, 21 Albunex/saline), the antigenic potential of study agentswas assessed by measuring blood histamine,tryptase, IgE, C3, C4, and CH50 prior to studyagent administration and 60 minutes after thelast injection of study agent.
Statistical Analysis
Ef�cacy data were analyzed separately foreach of the four offsite readers. Statistical sig-ni�cance was de�ned as P 0.05 (two-tailed).The difference in LVO scores for each dose ofSonoVue compared to the maximum opaci�ca-tion scores obtained in the control group (max-imum score between all four doses for eachpatient) was tested using a Wilcoxon rank sumtest. For duration of useful contrast effect, each
SONOVUE FOR IMPROVED EBD
29ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech.Vol. 19, No. 1, 2002
dose of SonoVue was compared with the max-imum response in the control group (maximumduration of useful contrast among all fourdoses for each patient) using separate analysisof variance (ANOVA) models with terms forinvestigational site and study agent group. Thechange from baseline in total LVEBD scorewas compared for the two treatment groupsusing an analysis of covariance (ANCOVA)model with terms for investigational site, studyagent group, and baseline score as covariate.Each dose of SonoVue and the maximum re-sponse for SonoVue (maximum change frombaseline between all four doses for each pa-tient) was compared with the maximum re-sponse in the control group using separateANCOVA models. For diagnostic con�dencescore, the change from baseline was comparedfor each dose of SonoVue versus the maxi-mum response in the control group using anANCOVA model with terms for investigationalsite, study agent group, and baseline score ascovariate. For duration of useful contrast andchange from baseline in LVEBD, data werecombined for the four offsite readers and sum-marized descriptively.
Results
LVO
For all four offsite readers, opaci�cationscores were signi�cantly higher for each of thefour doses of SonoVue as compared to the max-imum opaci�cation scores obtained in the con-trol group (all P 0.001; Wilcoxon rank sumtest). Across all readers and doses, moderate orcomplete LVO (i.e., score of 2 or 3) was ob-served for 73%–93% of patients following ad-ministration of SonoVue, compared to 14%–38% of patients following administration of Al-bunex (Fig. 1) (Table III). Opaci�cation wasnegligible after agitated saline injection. Withrespect to the percentage of patients with com-plete opaci�cation, some reader differenceswere apparent. For two of the four offsite read-ers (Study A, Reader 1 and Study B, Reader 2),the percentage of SonoVue patients with com-plete opaci�cation ranged from 34%–57%,while for the other two readers, the percentagewith complete opaci�cation ranged from 61%–87%. For all readers, a dose response was evi-dent in the SonoVue group, and in the controlgroup for the two Albunex doses.
Duration of Enhancement
For all four offsite readers, the duration ofuseful contrast effect (i.e., moderate or com-plete opaci�cation) was signi�cantly greaterfor each of the four doses of SonoVue comparedwith the maximum response in the controlgroup (all P 0.001; treatment comparisonfrom ANOVA). Across readers and doses, themean duration of useful contrast effect rangedfrom 0.8–4.1 minutes for SonoVue and lessthan 15 seconds for Albunex, with no usefulcontrast effect noted after saline injection(Figs. 2 and 3). The duration of useful contrastincreased with increasing SonoVue dose. Whendata for the four offsite readers were combined,mean ( SD) duration of useful contrast was1.1 ( 1.27), 1.6 ( 1.63), 2.2 ( 1.71), and 2.7( 1.94) minutes for the 0.5, 1, 2, and 4 mlSonoVue doses, respectively.
EBD
The mean change in LVEBD score after con-trast agent administration is shown for thecombined offsite readers in Figure 2. Across allreaders and doses, mean increases in LVEBDscore ranged from 3.8–18.2 after SonoVue ad-ministration, and from 0.1– 4.3 after Albunex.For all four offsite readers, the maximum re-sponse in the SonoVue group was signi�cantlygreater than the maximum response for thecontrol group (all P 0.001; treatment com-parison from ANCOVA). When data for thefour offsite readers were combined, mean (SD) increases in LVEBD score were 7.5 (4.95), 8.2 ( 4.68), 9.6 ( 4.35), and 10.1 (4.39) for the 0.5, 1, 2, and 4 ml SonoVue doses,respectively, and 1.3 ( 2.74) and 2.5 ( 3.90)for the 0.08 and 0.22 ml/kg Albunex doses,respectively.
The mean baseline scores for LVEBD in theSonoVue group differed between the four off-site blinded readers, as did the mean changefrom baseline scores across the four SonoVuedose levels. Thus, the reader with the lowestbaseline scores (Study B, Reader 2) showed thelargest increase with SonoVue administration,while the reader with the highest baselinescores (Study B, Reader 1) showed the smallestincrease. However, the results for each readerdemonstrated a dose response for SonoVue.Comparison of the individual SonoVue doselevels with control indicated signi�cantlygreater increases in LVEBD score at the 2 mland 4 ml SonoVue doses for all four offsite
NANDA, ET AL.
30 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 19, No. 1, 2002
readers (all P 0.001; treatment comparisonfrom ANCOVA).
Percentage of Diagnostic Images
The administration of SonoVue resulted in agreater increase in the percentage of patients
with diagnostic images as well as in overallcon�dence in the primary diagnosis as com-pared with administration of Albunex. For thefour offsite readers, the percentages of patientswith nondiagnostic images preinjection thatwere converted to diagnostic images with con-
Figure 1. Apical two- (top) and four-chamber views (bottom) showing opaci�cation of the left ventricular cavity followingintravenous injection of SonoVue.
TABLE III
Percent (%) of Patients with Moderate or Complete Left Ventricular Opaci�cation After Contrast Agent Administration
Data presented as % with complete opaci�cation (score 3) / % with moderate or complete opaci�cation (score 2 or 3).
SONOVUE FOR IMPROVED EBD
31ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech.Vol. 19, No. 1, 2002
trast ranged from 59%–97% for the 2 mlSonoVue dose, 70%–91% for the 4 ml SonoVuedose, and 18%–47% for the 0.22 ml/kg Albunexdose (Table IV).
For the two offsite readers in Study A, themean change from baseline in overall con�-dence score (rated on a scale from 0 to 5, where0 value assigned to nondiagnostic images,1 low, and 5 high) ranged from 1.3–2.1 inthe SonoVue group and from 0.0–0.7 in thecontrol group. For the two readers in Study B,the mean change from baseline in overall con-�dence score ranged from 0.8–3.9 in theSonoVue group and from 0.0–1.0 in the controlgroup. For three of the four offsite readers,statistically signi�cant differences were foundfor the three highest dose levels of SonoVuecompared to the maximum response in the con-trol group (P 0.05; treatment comparisonfrom ANCOVA), while for the remainingreader, the difference from control was signi�-cant for the 1 ml and 2 ml SonoVue doses only.
Dose Selection
The results for all four offsite readers dem-onstrated a pronounced dose-response effectfor SonoVue, with doses of 2 ml and 4 ml pro-viding the best results for LVO, duration ofuseful contrast, and improvement in border de-lineation (Table III) (Figs. 2 and 3). There wasa trend towards decreasing duration of usefulcontrast with increasing body weight. Thistrend was apparent at all SonoVue dose levels.There were no other apparent trends related todemographic or baseline characteristics, in-cluding age, gender, race group, or New YorkHeart Association Class. For patients withbody weight greater than 100 kg, a 4 ml dose of
SonoVue may be needed to ensure at least 2minutes of useful contrast effect.
Investigator Assessments
In both studies, the nonblinded investigatorassessments of LVO, duration of useful con-trast, and LVEBD scores were consistent withthose of the offsite readers.
Safety and Tolerability
Adverse events were reported for 22 (15.9%)of 138 patients who received SonoVue and 33(26.2%) of 126 patients receiving Albunex andsaline. The majority of adverse events in bothstudy agent groups were mild in intensity andresolved without sequelae. Adverse events con-sidered by investigators to be related to studyagent are shown in Table V. There were nosigni�cant differences between the SonoVueand control groups.
Mean changes in vital signs, oxygen satura-tion, and clinical laboratory values afterSonoVue administration were clinically insig-ni�cant and were comparable to those seen inthe control group. Individual abnormalitieswere generally attributable to underlying con-ditions and/or concomitant medications. ECGchanges were generally transient, including STsegment changes, sinus bradycardia, and iso-lated atrial and ventricular premature depo-larizations, and occurred with the same fre-quency in both study agent groups. There wereno signi�cant changes in MMSE scores. Therewere no clinically signi�cant changes in mark-ers of anaphylaxis and no evidence of antibodyproduction or complement activation.
Figure 2. Mean duration of useful contrast effect by studyagent (offsite image evaluation).
Figure 3. Endocardial border delineation of the left ven-tricle. Mean change from baseline in total apical view scoreby study agent (offsite image evaluation).
NANDA, ET AL.
32 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 19, No. 1, 2002
Discussion
Albunex was the �rst FDA-approved con-trast agent for echocardiography. In clinicaltrials, Albunex produced LVO and improvedEBD in a majority of patients, with inadequatedelineation on noncontrast echocardiogram.11
In clinical practice, however, Albunex wasfound to be of limited usefulness because of itsshort duration of contrast effect. Air-�lled mi-crospheres, such as constitute Albunex, aresusceptible to decreases in size due to diffusionand destruction by ventricular systolic pres-sure and acoustic pressure during fundamen-
tal ultrasound imaging.17,18 Signi�cantlygreater duration of contrast has been achievedwith per�uoropropane-�lled albumin micro-spheres (Optison),6 as well as with other micro-bubble preparations that contain �uorocarbongases, which have low solubility and diffusivityand are more pressure resistant.19,20 Studieshave shown that such agents are more effectivethan Albunex for improving delineation of en-docardial border.6,7
Improved delineation of endocardial borderis associated with more accurate assessmentsof LV systolic function. The use of contrastenhancement improves echocardiographic as-sessment of wall motion at rest and duringdobutamine stress,8,21 and increases the accu-racy of LV volume and ejection fraction mea-surements.22 With adequate contrast enhance-ment, echocardiographic assessment of wallmotion is as accurate as cine magnetic reso-nance imaging.22
The ef�cacy of SonoVue as a contrast agentfor echocardiography was assessed in two sep-arate multicenter studies. The results of fourindependent, blinded readers demonstratedthe superiority of SonoVue to Albunex in pro-ducing LVO, increasing the duration of usefulcontrast effect, improving delineation of car-diac borders, and increasing diagnostic con�-dence in patients undergoing echocardiogra-phy at rest.
LVO
At the two highest SonoVue doses tested (2ml and 4 ml), moderate or complete LVO wasachieved in more than 85% of patients com-pared with less than 40% for Albunex. Thestability of sulfur hexa�uoride microbubbles isdemonstrated by the prolonged duration of con-trast relative to Albunex. For both studies com-bined, the mean duration of useful contrast
TABLE IV
Changes from Nondiagnostic Preinjection to Diagnostic Postinjection
SonoVue Dose (ml)Albunex Dose
(ml/kg)Saline Dose
(ml/kg)
0.5 1.0 2.0 4.0 0.08 0.22 0.08 0.22
Study A Reader 1 61% 67% 79% 80% 21% 32% 12% 14%Study A Reader 2 75% 84% 86% 78% 37% 47% 6% 30%Study B Reader 1 33% 57% 59% 70% 5% 18% 5% 0%Study B Reader 2 74% 87% 97% 91% 19% 38% 5% 12%
33ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech.Vol. 19, No. 1, 2002
(intermediate or full opaci�cation of the LVcavity) was greater than 2 minutes at the twohighest SonoVue dose levels, while the dura-tion of useful contrast was less than 15 secondsfor Albunex.
The results for LVO with Albunex are con-sistent with those reported in other compara-tive studies,6,7 while the results obtained withSonoVue are comparable to those reported forper�uorocarbon-containing microbubble con-trast agents. Full LVO in up to 87% of patientswas reported with Optison,6 while 69%–78% ofpatients showed full or intermediate opaci�ca-tion with EchoGen.7 In the present study, thefour offsite readers were consistent in theirassessment of the percentage of patients withmoderate or complete opaci�cation (score of 2or 3) following each dose of SonoVue; however,there were some clear reader differences withrespect to the percentage of patients assigned ascore of 3 (complete LVO). This may have beendue to reader differences in interpretation ofthe distinction between moderate and completeopaci�cation based on the de�nitions provided.
EBD
The administration of SonoVue resulted insigni�cantly greater increases in LVEBDscores relative to unenhanced (baseline) im-ages than did administration of Albunex orsaline. Across all doses of SonoVue, mean in-creases in LVEBD scores (7.5–10.1) were 3–4times greater than those observed for Albunex0.22 ml/kg (2.5). A similar improvement in bor-der delineation score was observed for Echo-Gen as compared to Albunex.7
Among the four offsite blinded readers, therewas considerable variability in the range ofincreases in LVEBD score following SonoVueadministration. For three of the four readers,there was no overlap in mean change frombaseline scores across dose levels. These differ-ences may be related in part to differences inbaseline (noncontrast) LVEBD scores. The tworeaders with the lowest baseline scores showedthe greatest improvement with SonoVue ad-ministration, while the two with the highestbaseline scores showed the least improvement.Such differences in scoring are not unexpected,as each blinded reader brings his own interpre-tive preferences and thresholds to each readingsession. Since the offsite readers for each studywere investigators on the complimentarystudy, no training sessions were performedprior to the blinded reads and no attempt was
made to standardize the scoring approach ofthe four readers. The observed differencesamong readers in the grading of images withrespect to border delineation may re�ect differ-ences in diagnostic threshold, with a givenreader adopting a more conservative de�nitionof “visualization” than another reader.
The mean increase in LVEBD score of ap-proximately 10 observed at the two higherSonoVue doses would correspond to visualiza-tion of �ve additional segments (from inade-quate 0 to good 2) compared to noncontrastechocardiogram, or improvement in visualiza-tion for 10 segments in the two views (frominadequate 0 to adequate 1, or inadequateto good). This extent of improvement in EBDwould be expected to have signi�cant implica-tions for diagnostic con�dence. This was re-�ected in the increase in percentage of patientimages considered diagnostic and increase inoverall con�dence in the primary diagnosis.For all four offsite readers, the rate of conver-sion from nondiagnostic to diagnostic imageswas at least 70% at the two highest SonoVuedose levels (Table IV).
Dose Selection
Dose response effects of SonoVue were ap-parent for LVO, duration of useful contrasteffect, and increases from baseline in LVEBDscore. In all cases, the magnitude of responsewas greater for the two higher dose groups (2ml and 4 ml) compared with the lower dosegroups (0.5 ml and 1 ml). However, the differ-ences between the 2 ml and 4 ml doses wereless marked. With respect to conversion fromnondiagnostic to diagnostic images, there didnot appear to be a clear advantage for the 4 mldose as compared with the 2 ml dose (Table IV).These data suggest that 2 ml is the minimumeffective SonoVue dose for endocardial borderenhancement during echocardiography.
Signi�cant reductions in LVO following con-trast have been observed for obese versus nor-mal-weight patients.23 In the present study,duration of contrast with SonoVue diminishedwith increasing body weight. For patients withbody weight greater than 100 kg, a 4 ml dose ofSonoVue may be required to ensure at least 2minutes of useful contrast effect.
Safety
SonoVue was well tolerated in patients un-dergoing diagnostic echocardiography. The in-cidence of adverse events was similar to that
NANDA, ET AL.
34 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 19, No. 1, 2002
for Albunex/saline. There were no clinicallymeaningful trends in vital signs or clinical lab-oratory tests, and no clinically signi�cantchanges in ECG. Individual abnormalities invital sign and clinical laboratory values weregenerally attributable to concomitant illnessesand medications. There were no clinically sig-ni�cant changes in oxygen saturation and nochanges in MMSE scores. Because SonoVuemicrobubbles are constituted in a nonprotein-based carrier, the antigenic potential is mini-mal. In a subset of patients who underwentlaboratory tests to assess the immunologic po-tential of SonoVue, there was no evidence ofantibody production or complement activation.
Study Limitations
Continuous 2-D fundamental imaging wasused in the present study because it was themost widely available and generally applicablemode of echocardiography in 1997. The use ofmore advanced imaging techniques that areavailable currently may have affected the re-sults. For example, the use of second harmonicimaging has been shown to improve EBD bothwith and without contrast enhancement.24,25
Conclusions
In two multicenter, randomized, parallel-group trials, SonoVue was shown to be superiorto Albunex for visualization of cardiac cham-bers and delineation of endocardial bordersduring echocardiography at rest. The safetypro�les of the two agents were similar. ASonoVue dose of 2 ml appears to be optimal forincreasing diagnostic ef�cacy in patients withsuboptimal transthoracic echocardiograms.
Appendix: Principal Investigators for theSonoVue Multicenter Trials
Martha A. Carr, M.D., New Orleans Centerfor Clinical Research, New Orleans, LA; Greg-ory Clarke, M.D., Lindner Center for ClinicalCardiovascular Research, Cincinnati, OH;Mary C. Coretti, M.D., University of MarylandMedical Center, Baltimore, MD; Anthony N.DeMaria, M.D.,* University of California SanDiego Medical Center, San Diego, CA; David A.Foley, M.D.,† Mayo Clinic, Rochester, MN;Lloyd Goodman, M.D., Memorial Medical Cen-ter, Inc., Savannah, GA; Terrence C. Hack,M.D., Deaconess-Nashoba Hospital, Ayer, MA;Jack J. Hall, M.D., Memorial Hospital, India-napolis, IN; William Herndon, M.D., Charlotte,
NC; Ronald P. Karlsberg, M.D., CardiovascularResearch Institute of Southern California, Bev-erly Hills, CA; Arthur J. Labovitz, M.D., St.Louis University Health Sciences Center, St.Louis, MO; Emile R. Mohler, III, M.D., Presby-terian Medical Center, Philadelphia, PA;Navin C. Nanda, M.D.,† University of Ala-bama, Birmingham, AL; Julio E. Perez, M.D.,Washington University School of Medicine, St.Louis, MO; Michael H. Picard, M.D.,* Massa-chusetts General Hospital, Boston, MA;Thomas Ryan, M.D., Duke University MedicalCenter, Durham, NC; Mikel Smith, M.D., Uni-versity of Kentucky, Lexington, KY; William B.Smith, M.D., Louisiana Cardiovascular Re-search Center, New Orleans, LA; Jeffrey S.Soble, M.D., Rush Presbyterian-St. Luke’sMedical Center, Chicago, IL; James D.Thomas, M.D., Cleveland Clinic Foundation,Cleveland, OH; Daniel C. Wistran, M.D., Car-diology Physicians Inc., Salem, MA. *Offsitereaders for Study A; †Offsite readers forStudy B.
References
1. Visser CA, Kan G, David GK, et al: Echocardiographiccineangiography correlation in detecting left ventric-ular aneurysm: A prospective study of 422 patients.Am J Cardiol 1982;50:337–341.
2. Erbel R, Schweizer P, Meyer J, et al: Sensitivity ofcross-sectional echocardiography in detection of im-paired global and regional left ventricular function:Prospective study. Int J Cardiol 1985;7:375–389.
3. Freeman AP, Giles RW, Walsh NF, et al: Regional leftventricular wall-motion assessment: Comparison oftwo-dimensional echocardiography and radionuclideangiography with contrast angiography in healedmyocardial infarction. J Am Coll Cardiol 1985;56:8–12.
4. Schiller NB, Shah PM, Crawford M, et al: Recommen-dations for quantitation of the left ventricle by two-dimensional echocardiography. J Am Soc Echocar-diogr 1989;2:358–367.
5. Stamm RB, Carabello BA, Mayers DL, et al: Two-dimensional echocardiographic measurement of leftventricular ejection fraction: Prospective analysis ofwhat constitutes an adequate determination. AmHeart J 1982;104:136–144.
6. Cohen JL, Cheirif J, Segar DS, et al: Improved leftventricular endocardial border delineation and opaci-�cation with Optison (FS069), a new echocardio-graphic contrast agent: Results of a phase III multi-center trial. J Am Coll Cardiol 1998;32:746–752.
7. Grayburn PA, Weiss JL, Hack TC, et al: Phase IIImulticenter trial comparing the ef�cacy of 2% dode-ca�uoropentane emulsion (EchoGen) and sonicated5% human albumin (Albunex) as ultrasound contrastagents in patients with suboptimal echocardiograms.J Am Coll Cardiol 1998;32:230–236.
8. Ikonomidis I, Holmes E, Narbuvold H, et al: Assess-ment of left ventricular wall motion and delineation ofthe endocardial border after intravenous injection of
SONOVUE FOR IMPROVED EBD
35ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech.Vol. 19, No. 1, 2002
Infoson during dobutamine stress echocardiography.Coron Artery Dis 1998;9:567–576.
9. Firschke C, Koberl B, von Bibra H, et al: Combineduse of contrast-enhanced two-dimensional and colorDoppler echocardiography for improved left ventricu-lar endocardial border delineation using Levovist, anew venous echocardiographic contrast agent. IntJ Card Imaging 1997;13:137–144.
10. Falcone RA, Marcovitz PA, Perez JE, et al: Intrave-nous Albunex during dobutamine stress echocardiog-raphy: Enhanced localization of left ventricular endo-cardial borders. Am Heart J 1995;130:254–258.
11. Crouse LJ, Cheirif J, Hanly DE, et al: Opaci�cationand border delineation improvement in patients withsuboptimal endocardial border de�nition in routineechocardiography: Results of the phase III Albunexmulticenter trial. J Am Coll Cardiol 1993;22:1494–1500.
12. Schneider M, Arditi M, Barrau M-B, et al: BR1: A newultrasonographic contrast agent based on sulfurhexa�uoride-�lled microbubbles. Invest Radiol 1995;30:451–457.
13. Senior R, Andersson O, Caidahl K, et al: Enhancedleft ventricular endocardial border delineation withan intravenous injection of SonoVue, a new echocar-diographic contrast agent: A European multicenterstudy. Echocardiogr 2000;17:705–711.
14. Lindner JR, Villanueva FS, Dent JM, et al: Assess-ment of resting perfusion with myocardial contrastechocardiography: Theoretical and practical consider-ations. Am Heart J 2000;139:231–240.
15. Kaps M, Legemate DA, Ries F, et al: SonoVue intranscranial Doppler investigations of the cerebralarteries. J Neuroimaging 2001;11:261–267.
16. Madjar H, Prompeler HJ, Del Favero C, et al: A newDoppler signal enhancing agent for �ow assessmentin breast lesions. Eur J Ultrasound 2000;12:123–130.
17. Gandhok NK, Block R, Ostoic T, et al: Reduced for-ward output states affect the left ventricular opaci�-
cation of intravenously administered Albunex. J AmSoc Echocardiogr 1997;10:25–30.
18. Wei K, Skyba DM, Firschke C, et al: Interactionsbetween microbubbles and ultrasound: In vitro and invivo observations. J Am Coll Cardiol 1997;29:1081–1088.
19. Grayburn P. Per�enapent emulsion (EchoGen): Anew long-acting phase-shift agent for contrast echo-cardiography. Clin Cardiol 1997;20:I12–I18.
20. Porter TR, Xie F, Kricsfeld A, et al: Reduction in leftventricular cavity attenuation and improvement inposterior myocardial contrast with higher molecularweight intravenous per�uorocarbon-exposed soni-cated dextrose albumin microbubbles.J Am Soc Echo-cardiogr 1996;9:437– 441.
21. Hundley WG, Kizilbash AM, Afridi I, et al: Effect ofcontrast enhancement on transthoracic echocardio-graphic assessment of left ventricular regional wallmotion. Am J Cardiol 1999;84:1365–1368.
22. Hundley WG, Kizilbash AM, Afridi I, et al: Adminis-tration of an intravenous per�uorocarbon contrastagent improves echocardiographic determination ofleft ventricular volumes and ejection fraction: Com-parison with cine magnetic resonance imaging. J AmColl Cardiol 1998;32:1426–1432.
23. Hausnerova E, Gottdiener JS, Gould JJ, et al: Differ-ing susceptibility of echocardiographic contrastagents to adverse effects of biologic factors: Multi-center, videodensitometric comparison of octa�uoro-propane-�lled microspheres with air-�lled micro-spheres for left ventricular opaci�cation. J Am SocEchocardiogr 1999;12:957–966.
24. Lindner JR, Dent JM, Moos SP, et al: Enhancement ofleft ventricular cavity opaci�cation by harmonic im-aging after venous injection of Albunex. Am J Cardiol1997;79:1657–1662.
25. Kasprzak JD, Paelinck B, Ten Cate FJ, et al: Com-parison of native and contrast-enhanced harmonicechocardiography for visualization of left ventricularendocardial border. Am J Cardiol 1999;83:211–217.
NANDA, ET AL.
36 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 19, No. 1, 2002
