937

Quantification of Collateral Blood Flow inCoarctation of the Aorta by Velocity Encoded

Cine Magnetic Resonance ImagingJohann C. Steffens, MD; Michael W. Bourne, MD; Hajime Sakuma, MD;

Margaret O'Sullivan, RN; Charles B. Higgins, MD

Background Knowledge about the volume of collateral flowprovides insight into the severity of coarctation of the aortaand may be critical in planning the operative approach. Thereis currently no method for the quantification of collateral flowin coarctation of the aorta. In this study, we applied velocityencoded cine magnetic resonance imaging (VENC-MR) toestablish the flow pattern and volume of collateral flow in thedescending thoracic aorta in normal subjects and patients withcoarctation, introducing a new possibility to quantify theseverity of the coarctation by determining the amount ofcollateral flow.Methods and Results VENC-MR was used to measure flow

in the proximal and distal descending thoracic aorta in 10normal subjects. In 23 patients with coarctation, flow wasmeasured near the coarctation site and above the diaphragm.Patients were divided into a group with moderate to severecoarctation and a group with mild coarctation on the basis ofclinical gradient between upper and lower extremities and theestimation of the gradient across the coarctation by Dopplerechocardiography. The gradient across the coarctation and thedegree of anatomic narrowing were also assessed by MR

imaging. In normal volunteers, VENC-MR showed a 7±6%decrease in total flow, from proximal to distal aorta. Theinterobserver reproducibility was 3.9% to 4.9% (mean, 4.4%).In patients with moderate to severe coarctation, VENC-MRdemonstrated an 83±50% increase in total flow from proximalto distal aorta, yielding a significant change compared withnormal subjects (P<.01). Patients with mild coarctationshowed a normal flow pattern and no significant change intotal flow. There was a significant relation between the amountof flow increase in the distal aorta and the reduction in luminaldiameter at the coarctation site (r=.94) as well as the clinicalgradient (r=.84).

Conclusions This study shows the normal flow pattern inthe descending thoracic aorta and its reversal in coarctationdue to collateral flow. Thus, VENC-MR can measure collat-eral flow in coarctation and serves as a unique method forproviding this important measurement of the severity ofcoarctation of the aorta. (Circulation. 1994;90:937-943.)Key Words * blood flow velocity * coarctation * heart

defects, congenital * magnetic resonance imaging

C oarctation of the aorta is the third most commoncongenital malformation of the cardiovascularsystem, occurring in 4.1/10 000 newborns.1 De-

pending on the degree of severity of the coarctation, avariable collateral supply develops to provide blood flowto the descending thoracic aorta via the subclavian andintercostal arteries.2 Accurate assessment of the siteand severity of the structural abnormality is essentialfor planning surgery.3 The adequacy of collateral flow isusually estimated before surgery by qualitative inspec-tion of the presence and size of collateral arteriesdisplayed by x-ray aortography. No method currentlyexists for the quantification of collateral blood flow incoarctation.

Magnetic resonance (MR) imaging has been shown toprovide both anatomic and functional informationabout the cardiovascular system. Anatomy is optimallydemonstrated on spin-echo images, in which high con-trast between moving blood, which is black, and tissueallows excellent depiction of myocardial or vascular

Received December 10, 1993; revision accepted April 6, 1994.From the Department of Radiology, University of California,

San Francisco.Correspondence to Charles B. Higgins, MD, Chief, MRI Sec-

tion, Department of Radiology, Box 0628, University of Califor-nia, San Francisco, 505 Parnassus Ave, Room L 308, San Fran-cisco, CA 94143-0628.© 1994 American Heart Association, Inc.

structures. MR spin-echo imaging has been shown to behighly effective for the evaluation of thoracic aorticdiseases, including thoracic aortic aneurysms, dissec-tions, and coarctations.4-10

Cine MR imaging has been applied to depict movinganatomy and provides information about cardiac andvalvular function.11-15 Recent studies have also evalu-ated the use of cine MR imaging to access the jet (flowvoid) across the coarctation16 and have inferred estima-tion of severity based on the length of the flow void.

Further functional information can be provided byvelocity encoded cine MR imaging (VENC-MR).VENC-MR provides a measurement of blood flow atany site in the cardiovascular system. This technique hasbeen shown to be effective for measuring gradientsacross cardiac valves and vascular stenoses17-19 and flowthrough cardiac shunts in the ascending and descendingaorta and in peripheral vessels.20-22The hypothesis of the present study was that mea-

surement of flow in the proximal and distal thoracicaorta should show an increase from the proximal to thedistal aorta in hemodynamically significant coarctations.The rise in flow is due to retrograde flow in theintercostal arteries and other collateral channels. Thus,this approach might provide a measurement of thevolume of collateral flow in coarctation. To the best ofour knowledge, no other existing technique can providethis information.

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

938 Circulation Vol 90, No 2 August 1994

.(.: ..3~ .-

FIG 1 (THIS PAGE AND FACING PAGE). Oblique spin-echo image (A) shows a patient with severe coarctation of the aorta (arrow). Velocityencoded cine magnetic resonance image slices are prescribed perpendicular to the direction of blood flow at the site of the coarctation,below it, and above the diaphragm (B). On the transaxial magnitude image at the level of the diaphragm, a region of interest (ROI) wasprescribed around the aortic lumen (C). This ROI was then applied to the corresponding phase image (D). Aa indicates ascending aorta;Ad, descending aorta; and L, left subclavian artery.

Study PopulationNormal VolunteersTen normal subjects (3 women and 7 men) were

studied. Mean age was 32.6 years (range, 28 to 36 years).The subjects had no known cardiac or systemic diseasesand did not use any kind of medication. All subjectsgave informed consent to participate in this study.

PatientsTwenty-four patients were included in this study. All

patients were referred for evaluation of coarctation ofthe aorta. Mean age was 11.2 years (range, 8 months to48 years). One patient was excluded because of thepresence of a patent ductus arteriosus. The remaining23 patients were divided into two groups: Group A hadmoderate to severe coarctation, defined as a pressuregradient of >20 mm Hg across the coarctation by Dopp-ler echocardiography or a pressure gradient of >25mm Hg between the right arm and the right leg (n= 15patients). Group B had mild coarctation, defined as a

pressure gradient of <20 mm Hg across the coarctationby Doppler echocardiography and <25 mm Hg betweenthe right arm and the right leg (n=8 patients).

All but one patient in group A showed a clinicalgradient of >25 mm Hg. One patient showed a clinicalgradient of 20 mm Hg but was included in group Abecause the Doppler gradient was 30 mm Hg.

MethodsMagnetic Resonance Imaging TechniqueA sagittal spin-echo sequence was used to localize the

descending aorta and to prescribe the slices for the

VENC-MR sequences. In normal volunteers, three sliceswere located below the region of the left subclavian artery,at a level just above the diaphragm, and midway betweenthese two. In patients, the three imaging planes wereprescribed at the coarctation site, 10 mm below thecoarctation, and at a level just above the diaphragm (Fig1). All imaging planes were positioned perpendicular tothe direction of the blood flow, and the velocity encodingwas performed in the slice selection direction, which wasparallel to the direction of flow. For this purpose, anoblique section was prescribed that was perpendicular tothe long axis of the aorta at each site. Because of thecurvature of the descending aorta, various obliquities wereused at the three sampling sites (Fig 1).VENC-MR images were acquired by use of ECG

referencing, a repetition time (TR) of 24 milliseconds,an echo time of 7 milliseconds, and 10-mm obliqueslices at a rate of 16 time frames per RR interval. Thematrix size was 128 x 256, and the number of excitationswas two. Reconstruction of VENC-MR data providedboth a magnitude and a phase image.The phase of proton spins was modulated by use of a

bipolar gradient pulse for velocity encoding in the slicedirection. Each series consisted of two interleaved cineMR acquisitions using different flow-encoding gradientsalong the same direction. Subtraction of the data fromthe two acquisitions eliminates phase shifts introducedby chemical shift phenomena and magnetic field inho-mogeneities, resulting in cancellation of all phase shiftsexcept those due to flow along the velocity-encodeddirection according to the relation

Aphase- aMv

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

Steffens et al Collateral Flow and MRI 939

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

940 Circulation Vol 90, No 2 August 1994

where d is the gyromagnetic ratio, M is the difference ingradient moment between two acquisitions, and v is thevelocity. If the gradients of the two experiments arechosen to keep the phase shift < 180° to avoid aliasing, thevelocity component along the velocity-encoded directioncan be determined directly from the phase difference andM. The peak velocity, which is the predicted highestvelocity to be encoded, was selected over the rangebetween 4 and -4 milliseconds in normal volunteers andbetween 7 and -7 milliseconds in patients with coarcta-tion. If the actual maximal velocity exceeds the selectedpeak velocity, aliasing occurs, with resultant underestima-tion of the true velocity. Therefore, the peak velocity wasset substantially higher than expected to be found in thedescending thoracic aorta. It was set higher in the patientsso as to avoid any possibility of aliasing by very high jetvelocity, which might be predicted at the two sites in theproximal descending aorta.

Imaging TimeIn normal volunteers, the imaging time was 5 minutes

for the locators and 12 minutes for the VENC-MRsequences. In patients, VENC-MR sequences wereperformed as part of a routine clinical evaluation, andno additional locators were necessary. Additional imag-ing time was 12 minutes.

Image AnalysisImages were analyzed independently by two observ-

ers. An observer-defined region of interest (ROI) wasdrawn around the aorta on the magnitude image (cross-sectional area in square centimeters) and then appliedto the corresponding phase image (centimeters persecond) for all cardiac phases at each level to givemeasurements for the calculation of flow (cm2 xcm/scm3/s). Results were reported for each time frameduring the cardiac cycle, resulting in a curve of flowversus time. This curve was then integrated to calculatestroke volume in the aorta. Collateral flow was ex-pressed as the percentage of flow increase from theproximal to the distal descending thoracic aorta.To calculate the gradient, the VENC-MR image slice

at the coarctation site was chosen. An ROI (threepixels) was placed at the area of the highest signalintensity in the aortic lumen. The gradient was calcu-lated by use of the modified Bernoulli equation (p=4v2,where p is pressure gradient and v is peak velocity).The percentage reduction in luminal diameter at the

site of the coarctation was determined by measuring thediameter of the descending aorta 1 cm above the dia-phragm, midway in the ascending aorta, and at the coarc-tation site. The percent stenosis at the coarctation site wasexpressed according to the equation

% stenosis= diameter of coarctation/[(diameter AsAo+diameter DescAo)/2]x 100%

where AsAo is the ascending aorta and DescAo is thedescending aorta.

StatisticsLinear regression analysis was used to determine the

relation among (1) MR-estimated collateral flow, (2)pressure gradient calculated by Doppler echocardiogra-phy, (3) pressure gradient calculated by VENC-MR, (4)pressure difference between right arm and right leg

160

140

_ 120

Q 1X(DC)100 "

E 80t

0 t-L 40

- proximaldesc. aorta

mid desc.aorta

n distaldesc. aorta

20

0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Cardiac phasesFIG 2. Curve of flow vs time at three sites in the descending(desc.) thoracic aorta in a normal volunteer. Notice the slightdecrease in flow from the proximal to the distal aorta.

obtained at clinical examination, and (5) the degree ofnarrowing at the coarctation.

Interobserver variability for collateral flow and MRgradient measurements was expressed as the percentageof variability and SD. Percentage of variability wasdetermined as the absolute value of the differencebetween the two measurements over the mean of thetwo measurements.

ResultsNormal Volunteers

Fig 2 shows the flow pattern observed at the threelevels in the descending aorta. There was a 7±6%decrease in total flow from the proximal to the distalpart of the descending thoracic aorta. At the level abovethe diaphragm, 65 + 10% of the total flow was contrib-uted in systole, and 35±9% occurred in diastole.

PatientsFlow Pattern

All patients with moderate to severe coarctation(group A) showed an increase in flow from the proximalto the distal descending thoracic aorta. The meanincrease was 83±50%. The relation between systolicand diastolic contribution to total flow was 48±32% insystole and 52±38% in diastole. Fig 3 shows a typicalflow curve in a patient with severe coarctation. Therelation of systolic and diastolic flow in these patientswas significantly different compared with normal sub-jects (P<.01) and with patients with mild coarctation(P<.01).Seven of 8 patients with mild coarctation showed no

increase in flow from the proximal to the distal descend-ing thoracic aorta. One patient showed a 6% increase inflow. Fifty-nine percent of total flow was contributed insystole and 41% of flow in diastole. Fig 4 summarizesthe flow changes from the proximal to the distal de-scending thoracic aorta in all three groups.

Pressure GradientsClinical gradients (pressure difference between right

arm and right leg) were available in all 23 patients. Thepressure gradient across the coarctation by Doppler

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

Steffens et al Collateral Flow and MRI 941

0

* /

0

0

0

. 0

0 50 100 150 200 250Flow change (%)

20 -YS<C i,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Cardiac phasesFIG 3. Curve of flow vs time at two sites in the descending(desc.) aorta in a patient with severe coarctation of the aorta.This patient showed a 47% increase in flow in the distal aortacompared with the proximal aorta.

echocardiography was obtained in 17 patients. Six pa-

tients had no Doppler data because of difficulty invisualization of the region of the coarctation. Gradientsmeasured by VENC-MR were available in 18 patients.In 5 patients, no gradient could be measured byVENC-MR because of partial volume effects in patientswith very severe coarctation.

There was a close correlation between gradientsobtained with VENC-MR and Doppler echocardiogra-phy (r=.95). Both methods, however, showed poorercorrelation with the clinical gradient (r=.63 for Dopplerechocardiography, r=.54 for VENC-MR).The correlation coefficient for the relation between

the increase in flow from the proximal to the distaldescending thoracic aorta and the clinical pressuregradient across the coarctation was .76 for gradientsobtained with Doppler echocardiography and .80 forgradients measured by VENC-MR.The amount of flow increase correlated significantly

with the clinical gradient (r=.84, P<.05) (Fig 5) and thedegree of anatomic narrowing at the site of the coarc-

tation (r=.94, P<.05) (Fig 6).

Interobserver VariabilityThe interobserver variability and SD for flow change

in normal volunteers was 3.9±2.8%. For flow change inpatients, the result was 4.2±3.1%, and the interob-server variability for the calculated MR gradients inpatients was 4.9±3.5%.

C')

5

5

-

E

140

120

100

80

60

40

20

0

-20

FIG 5. Plot showing percentage of flow change on the x axisand clinical gradient on the y axis.

DiscussionThis study has shown the ability of VENC-MR to

measure flow in different parts of the descending tho-racic aorta in normal subjects and in patients withcoarctation of the aorta. MR imaging was able tocalculate the amount of flow change from the proximalto the distal part of the descending aorta and therebyquantify the volume of collateral flow into the distaldescending aorta.The findings in normal volunteers show that there is a

slight decrease in the flow volume from the proximal tothe distal part of the descending thoracic aorta, proba-bly as a result of blood leaving the aorta into theintercostal arteries. In patients with coarctation, thisnormal flow pattern is reversed to a variable degree.Patients with moderate to severe coarctation show anincrease in flow from the proximal to the distal aorta.This increase is caused by the collateral flow enteringthe descending aorta via retrograde flow from theintercostal arteries.2,23 Therefore, determining theamount of flow increase from the proximal to the distaldescending thoracic aorta provides a direct measure-ment of the volume of collateral blood supply to thelower part of the body. Because collateral blood flowdevelops only if the stenosis in the original vessel issignificant enough to cause a lack of blood supply in thedependent part,24 the amount of collateral flow andtherefore the amount of flow increase in the descendingaorta should be an indicator of the hemodynamic sever-ity of the coarctation. The strong correlation betweenthis functional indicator of the severity and the degreeof anatomic narrowing at the site of the coarctationsupports this notion.The present study also disclosed an alteration in the

shape of the flow curve between normal volunteers andpatients with severe coarctation. Whereas normal sub-jects and patients with mild coarctation distribute mostof the blood flow during systole, patients with moderate

FIG 4. Bar graph showing summarization of flowchanges from the proximal to the distal aorta forall three groups. Although there is no significantdifference between group B and the normal vol-unteers, the flow change in group A is signifi-cantly different compared with the other two(P<.01). Group A had moderate to severe coarc-tation; Group B, mild coarctation.

Group B Normals

120

100

80

60

40

0

0

U-

60_

E 50E

40-, 300

ca 20

U -

-0 _-50

*_A--w

Group A

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

942 Circulation Vol 90, No 2 August 1994

100 _

80 -

60 -

._n

° 40 -

)20 h

00-s o s 1 oo 1so 20 2s-50 0 50 100 150 200 250

Flow change (%)

FIG 6. Plot showing percentage of flow change on the x axisand percentage of anatomic narrowing on the y axis.

to severe coarctation had a temporal shift of the flowcurve to the diastolic phase of the cardiac cycle. Thereare two possible explanations for this phenomenon.This shift in flow curve could be a result of the increasedtime required for the blood provided by the collateralsto course through the intercostal arteries into thedescending aorta. Conversely, this flow pattern in pa-tients with moderate to severe coarctation appearssimilar to flow patterns obtained in stenotic peripheralvessels22 and could therefore be caused primarily bydelayed transmission of flow through the stenosis in theproximal descending aorta itself. In our study, the flowcurve below the coarctation showed the same shape asthe one above the diaphragm in terms of diastoliccontribution to total blood flow, although no collateralinflow is present at the higher location. This observationsupports the latter mechanism as at least contributing tothe temporal shift of blood flow to the diastolic part ofthe cardiac cycle.

It should be recognized that the average age of thevolunteers was two decades older than the average ageof the patients with coarctation. The flow pattern de-scribed as normal in the present study may not beexactly characteristic of the normal aortic flow in youngchildren as a consequence of alteration in aortic com-

pliance and other factors that are known to occur duringaging. Previous studies192' have shown the accuracy ofVENC-MR techniques to measure flow in the aorta.This technique has been applied to measure flowchanges in the ascending aorta and in the descendingaorta in patients with coarctation.2' Additionally,VENC-MR has been used for the measurement of thevolume of left-to-right shunts.22 Gradients across coarc-tation of the aorta,25 across Rastelli conduits,26 andacross valvular stenoses18'9 have been measured byVENC-MR to define the peak velocity in the flow jet.

Although the gradients across the coarctation measuredby VENC-MR and by Doppler echocardiography corre-lated clearly in the present study, both showed poorcorrelation with the clinical blood pressure gradient be-tween the right arm and the right leg. This supportsfindings of previous studies that the gradient across thecoarctation can be misleading in terms of severity of thedisease.27 28This may be a problem in severe coarctationbecause the large amount of collateral blood flow enteringthe descending aorta raises pressure beyond the coarcta-tion and may minimize the effect of the severe coarctationon distal aortic pressure. To support this theory, studiescomparing Doppler gradients with morphological data inpatients with coarctation have found poor correlationbetween the degree of anatomic obstruction and thepressure gradient, especially in patients with a large col-lateral supply.28,29

In addition, both Doppler echocardiography andVENC-MR were unreliable in measuring the gradientacross severe coarctations in the present study. WhereasDoppler echocardiography failed to create adequatesignal across the coarctation to visualize the exactlocation of the narrowing, MR imaging did not givereliable information about the velocity across the coarc-tation because of partial volume effect and higherorders of turbulent motion in very tight coarctations.

In mild to moderate coarctation, however, Dopplerhas been found to accurately estimate the gradient.Consequently, the Doppler gradient was consideredreliable for dividing patients into groups with gradientsabove and below 20 mm Hg.These limitations of gradient-based assessments of

the severity of coarctation further suggest that thequantification of the amount of collateral flow might bemore useful for the determination of the severity ofcoarctation of the aorta. Conversely, there are certainpatients with severe coarctation who, for unknownreason, have an inappropriately low volume of collateralflow. The present technique may serve as a method toidentify these patients with a mismatch between themorphological severity and the volume of collateral flowto the distal descending aorta. Recognizing such amismatch is important in planning surgical managementof these patients.

Clinical ImplicationsMagnetic resonance spin-echo imaging has been

shown to be effective as a noninvasive method forevaluating the morphology of coarctation of the aorta.30It accurately depicts the location and the degree ofnarrowing. However, this morphological informationdoes not necessarily indicate the hemodynamic conse-quences of the narrowing. Additional VENC-MR infor-mation can provide evaluation of both anatomic andhemodynamic severity of the coarctation.

Information about the amount of collateral flow isimportant for planning surgery. Depending on the quan-tity of collateral flow, surgery can be performed withcross-clamping alone (high amount of collateral flow), oradditional techniques such as left heart bypass, internalshunt, or a jump graft may be applied to ensure properblood flow to the lower part of the body.31 In addition,evaluating the flow in the descending aorta could be usedto monitor patient outcome after corrective surgery.

Clinical application of this technique might be partic-ularly useful in two clinical situations. First, the pres-ence of collateral circulation might be used to verify theclinical impression that a coarctation is sufficientlysevere to require surgery or angioplasty. Moreover, theadequacy of an angioplasty might be documented byreversion of the flow pattern to normal. Second, theabsence of adequate collateral circulation defined be-fore surgery would be useful in planning the surgicalapproach to minimize jeopardy for spinal cord ischemia.

LimitationsThis type of evaluation of patients with coarctation

could be compromised by partial volume errors causedby a very small size of the ROI in infants with tightcoarctations, yielding unreliable measurements of flowat the area below the coarctation. Other limitationsinclude patients with additional cardiovascular malfor-mations, such as a patent ductus arteriosus, which

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

Steffens et al Collateral Flow and MRI 943

would make it impossible to distinguish between left toright shunting and collateral flow into the distal de-scending thoracic aorta.

Collateral flow entering below the level of the dia-phragm cannot be measured by this method because thephysiological outflow of blood through the celiac arterywould prevent accurate measurements at a lower level.This leads to underestimation of the total volume ofcollateral flow. However, measuring the blood flow atthe same location in all patients still provides sufficientdata for defining characteristics indicative of hemody-namically significant coarctations.

ConclusionsThis study demonstrates the effectiveness of VENC-

MR for quantifying flow in the descending thoracicaorta in normal subjects and in patients with coarctationof the aorta. In patients with hemodynamically signifi-cant coarctation of the aorta, the normal flow pattern isreversed. The increasing volume of flow in the distalthoracic aorta in patients with moderate to severecoarctation represents the contribution to total flowmade by the collateral circulation via the intercostalarteries, and this measurement may serve as an addi-tional parameter for predicting the severity of theanatomic and physiological obstruction. It may also beused to detect a mismatch between the severity of theanatomic obstruction and collateral flow, which can be acentral factor in planning the surgical procedure.

References1. Silver EN. Heart Disease. 2nd ed. New York, NY: Macmillan

Publishing Co; 1987;765.2. Kirklin JW, Boyes BH. Cardiac Surgery. 2nd ed. New York, NY:

Churchill Livingstone; 1993;1267.3. Braunwald E. Heart Disease: A Textbook of Cardiovascular

Medicine. 3rd ed. Philadelphia, Pa: WB Saunders Co; 1988:994-997.

4. Kersting-Sommerhoff BA, Sechtem U, Schiller NB, Lipton MJ,Higgins CB. MR imaging of the thoracic aorta in Marfan patients.J Comput Assist Tomogr. 1987;11:633-639.

5. Dinsmore RE, Liberthson RR, Wismer GL, Miller SW, Liu P,Thompson R, McCloud TC, Marshall J, Saini S, Stratemier EJ,Okada RD, Brady T-J. Magnetic resonance imaging of thoracicaortic aneurysms: comparison with other imaging methods. AJRAm J Roentgenol. 1986;146:309-314.

6. Boxer RA, LaCorte MA, Singh S, Davis J, Goldman M, Stein HL.Evaluation of the aorta in the Marfan syndrome by magneticresonance imaging. Am Heart J. 1986;111:1001-1007.

7. Kersting-Sommerhoff BA, Higgins CB, White RD, SommerhoffCP, Lipton MJ. Aortic dissection: sensitivity and specificity of MRimaging. Radiology. 1988;3:651-656.

8. Nienaber CA, Spielmann RP, von Kodolitsch Y, Sieglow W,Piepho A, Jaup T, Nicolas V, Weber P, Triebel HJ, Bleifeld W.Diagnosis of thoracic aortic dissection: magnetic resonanceimaging versus transesophageal echocardiography. Circulation.1992;85:434-447.

9. Amparo EG, Higgins CB, Hricak H, Sollitto R. Aortic dissection:magnetic resonance imaging. Radiology. 1985;155:399-406.

10. Rees S, Somerville J, Ward C, Martinez J, Mohiaddin RH,Underwood R, Longmore DB. Coarctation of the aorta: MRimaging in late postoperative assessment. Radiology. 1989;173:499-502.

11. Suzuki JI, Chang JM, Caputo GR, Higgins CB. Evaluation of rightventricular early diastolic filling by cine nuclear magnetic res-

onance imaging in patients with hypertrophic cardiomyopathy.JAm Coll Cardiol. 1991;18:120-126.

12. Semelka RC, Tomei E, Wagner S, Mayo J, Caputo GR, O'SullivanM, Parmley WW, Chatterjee K, Wolfe C, Higgins CB. Interstudyreproducibility of dimensional and functional measurementsbetween cine magnetic resonance studies in the abnormal leftventricle. Am Heart J. 1990;119:1367-1373.

13. Sechtem U, Pflugfelder PW, White RD, Gould RG, Holt W,Higgins CB. Cine MR imaging: potential for the evaluation ofcardiovascular function. AJRAm J Roentgenol. 1987;148:239-245.

14. Wagner S, Auffermann W, Buser P, Lim TH, Kirchner J,Pflugfelder P, Higgins CB. Diagnostic accuracy and estimation ofthe severity of valvular regurgitation from the signal void on cineMR imaging. Am Heart J. 1989;118:760-767.

15. Higgins CB, Wagner S, Kondo C, Suzuki JI, Caputo GR. Eval-uation of valvular heart disease with cine gradient echo magneticresonance imaging. Circulation. 1991;84:198-206.

16. Didier D, Ratib 0, Friedli B, Oberhaensli I, Chatelain P, FaiduttiB, Rutishauser W, Terrier F. Cine gradient echo MR imaging inthe evaluation of cardiovascular diseases. Radiographics. 1993;13:561-573.

17. Kilner PJ, Firmin DN, Rees RS, Martinez J, Pennell DJ,Mohiaddin RH, Underwood SR, Longmore DB. Valve and greatvessel stenosis: assessment with MR jet velocity mapping.Radiology. 1991;178:229-235.

18. Eichenberger AC, Jenni R, von Schultheiss GK. Aortic valvepressure gradients in patients with aortic valve stenosis: quantifi-cation with velocity encoded cine MR imaging. AJR Am JRoentgenol. 1993;160:971-977.

19. Rees RSO, Firmin DN, Mohiaddin RH, Underwood SF, Long-more DB. Application of flow measurements by magnetic res-onance velocity mapping to congenital heart disease.Am J Cardiol.1989;64:953-956.

20. Brenner LD, Caputo GR, Mostbeck G, Steinmann D, Dulce M,Cheitlin MD, O'Sullivan M, Higgins CB. Quantification of left toright atrial shunts with velocity encoded cine nuclear magneticresonance imaging. JAm Coll Cardiol. 1992;20:1246-1250.

21. Bogren HG, Klipstein RH, Firmin DN, Mohiaddin RH, Under-wood SR, Rees RS, Longmore DB. Quantification of antegradeand retrograde blood flow in the human aorta by magnetic res-onance velocity mapping. Am Heart J. 1989;117:1214-1222.

22. Caputo GR, Higgins CB. Magnetic resonance angiography andmeasurement of blood flow in the peripheral vessels. Invest Radiol.1992;27:97-102.

23. Krieger KH, Spencer FC. Is paraplegia after repair of coarctationof the aorta due principally to distal hypotension during aorticcross-clamping? Surgery. 1985;97:2-7.

24. Lopez JA, Armstrong ML, Harrison DG, Piegors DJ, Heistad DD.Responsiveness of iliac collateral vessels to constrictor stimuli inatherosclerotic primates. Circ Res. 1988;63:1020-1028.

25. Mohiaddin RH, Kilner PJ, Rees S, Longmore DB. Magnetic res-onance volume flow and jet velocity mapping in aortic coarctation.JAm Coll CardioL 1993;22:1515-1521.

26. Martinez JE, Mohiaddin RH, Kilner PJ, Khaw K, Rees RS, Som-erville J, Longinore DB. Obstruction in extracardiac ventriculo-pulmonary conduits: value of nuclear magnetic resonance imagingwith velocity mapping and Doppler echocardiography. JAm CollCardioL 1992;20:338-344.

27. Michaelson M, Sunnegardh J, Hallberg M, Aberg T. Cuff pressuresand Doppler gradients after coarctectomy: a long term follow-up.Acta Paediatr Scand. 1986;329:94-97.

28. Wendel H, Teien D, Human DG, Nanton MA. Doppler echocar-diography and morphologic evaluation of patients followingoperative repair of aortic coarctation. Acta Paediatr Scand. 1992;81:247-252.

29. Hanson E, Eriksson BO, Soerensen SE. Intra-arterial bloodpressures at rest and during exercise after surgery for coarctationof the aorta. EurJ Cardiol. 1980;11:245-257.

30. Higgins CB, Caputo GR. Role of MR imaging in acquired andcongenital cardiovascular disease. AJR Am J Roentgenol. 1993;161:13-22.

31. Lousto R, Kyllonen KE, Merikallio E. Surgical treatment of coarc-tation of the aorta with minimal collateral circulation. Scand JThorac Cardiovasc Surg. 1980;14:217-220.

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from

J C Steffens, M W Bourne, H Sakuma, M O'Sullivan and C B Higginsmagnetic resonance imaging.

Quantification of collateral blood flow in coarctation of the aorta by velocity encoded cine

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1994 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.90.2.937

1994;90:937-943Circulation. 

http://circ.ahajournals.org/content/90/2/937World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on June 13, 2018http://circ.ahajournals.org/

Dow

nloaded from