Combined Assessment of Aortic Root Anatomy and

Va s c u l a r a n d I n t e r ve n t i o n a l R a d i o l o g y • O r i g i n a l R e s e a r c h Blanke et al. DSCT of the Aortic Root and Aortoiliac Vasculature

FOCUS ON:

Downloaded from www.ajronline.org by 52.2.249.46 on 09/13/15 from IP address 52.2.249.46. Copyright ARRS. For personal use only; all rights reserved

Vascular and Interventional Radiology Original Research

Philipp Blanke1 Wulf Euringer 1 Tobias Baumann1 Jochen Reinöhl2 Christian Schlensak 3 Mathias Langer 1 Gregor Pache1 Blanke P, Euringer W, Baumann T, et al.

Keywords: aortic root dimensions, aortoiliac CT angiography, dual-source CT, transcatheter aortic valve implantation DOI:10.2214/AJR.10.4232 Received December 31, 2009; accepted after revision March 6, 2010. 1 Department of Diagnostic Radiology, University Hospital Freiburg, Hugstetter Strasse 55, 79106 Freiburg, Germany. Address correspondence to P. Blanke ([email protected]). 2 Department of Cardiology, University Hospital Freiburg, Freiburg, Germany. 3 Department of Cardiovascular Surgery, University Hospital Freiburg, Freiburg, Germany.

AJR 2010; 195:872–881 0361–803X/10/1954–872 © American Roentgen Ray Society

872

Combined Assessment of Aortic Root Anatomy and Aortoiliac Vasculature With Dual-Source CT as a Screening Tool in Patients Evaluated for Transcatheter Aortic Valve Implantation OBJECTIVE. The objective of our study was to investigate the feasibility, image quality, and clinical implications of a combined dual-source CT angiography (CTA) protocol to assess aortic root anatomy and aortoiliac vasculature in patients with severe aortic stenosis evaluated for transcatheter aortic valve implantation. SUBJECTS AND METHODS. Eighty consecutive patients (47 women and 33 men; mean age, 82.3 ± 7.8 [SD] years) with severe aortic stenosis evaluated for transcatheter aortic valve implantation underwent a combined single-dose contrast-enhanced dual-source CTA protocol (body weight < 70 kg, 110 mL of contrast medium; ≥ 70 kg, 130 mL) consisting of ECG-gated dual-source CTA of the chest with integrated cardiac CT and ungated CTA of the abdomen and pelvis. Two independent observers measured the dimensions of the aortic root and the aortoiliac vasculature and rated image quality semiquantitatively. Vessel attenuation was assessed. Amenability to transfemoral access was evaluated on the basis of vessel diameter (> 7 mm), anatomy, and the presence of vascular disease. RESULTS. Image quality of the aortic root was diagnostic in all 80 patients, and image quality of the aortoiliac vasculature was diagnostic in 79 patients. Vascular attenuation was greater than 200 HU at any vessel level. The mean diameter of the aortic annulus was 24.1 ± 2.9 (SD) mm. Inter- and intraobserver correlations for aortic root and aortoiliac measurements were high (r = 0.93–0.99). Aortic root dimensions were suitable for transcatheter aortic valve implantation in 65 patients (81%). Thirty-eight patients (48%) were deemed amenable to instant transfemoral access without another vasculature intervention. CONCLUSION. The dimensions of the aortic root and the aortoiliac vasculature can be assessed with a combined single-dose contrast-enhanced dual-source CTA protocol, thereby allowing determination of patient eligibility for transcatheter aortic valve implantation, prosthesis sizing, and evaluation of the access route in one examination.

A

ortic stenosis is the most frequent cardiac valve disease in Europe [1, 2]. Surgical valve replacement is the gold standard for treating severe aortic stenosis and is technically possible in patients of any age [3, 4]. However, as many as 30% of patients with aortic stenosis are not referred for surgery or are turned down because of comorbidities and expected perioperative mortality [5]. Transcatheter aortic valve implantation constitutes a new alternative for high-risk surgical candidates [6]. Before valve implantation, patients undergo an extensive workup to assess the anatomy of the aortic root and of the coronary and peripheral arteries, which is relevant for valve sizing and for positioning and access planning, respectively. Diagnostic techniques include echocardiography; cardiac catheteriza-

tion; and, in case of a transfemoral approach, aortoiliac angiography [6]. The results of these diagnostic studies will help to judge the procedure’s feasibility and to guide the decision about the access strategy, either transfemoral or transapical. CT can noninvasively provide relevant information about the anatomy of the aortic root [7] and peripheral vasculature [8]. The aortic annulus diameters and location of coronary ostia can be assessed reliably with cardiac CT [7]. Aortoiliac CT angiography (CTA) constitutes an accurate diagnostic technique for evaluating the size, tortuosity, and calcification of peripheral arteries and is routinely used before endovascular aortic repair procedures [8, 9]. With the introduction of dual-source CT (DSCT), it is now possible to integrate ECG-assisted CTA of the heart

AJR:195, October 2010


DSCT of the Aortic Root and Aortoiliac Vasculature Fig. 1—83-year-old man with severe aortic stenosis. A, Coronal scout view for planning CT data acquisitions. Upper and dashed rectangles indicate scan range of ECG-assisted CT angiography (CTA) of chest and integrated zone of increased tube current for cardiac CT, respectively. Lower rectangle indicates scan range for CTA of abdominal aorta and iliac vasculature. Overall scan range extended from proximal supraaortic vessels to level of femoral heads. ECG-assisted data acquisition was extended to diaphragm to cover all cardiac structures. B, Fused coronal maximum-intensity-projection image of thoracic and abdominal data sets.

CT Protocol

A

in an examination of the aortoiliac axis after a single injection of contrast medium [10, 11]. The purpose of this study was to investigate the feasibility, image quality, and clinical implications of a combined one-stop-shop DSCT protocol, consisting of ECG-assisted CTA of the chest with integrated cardiac CT followed by ungated CTA of the aortoiliac axis below the diaphragm, in evaluating patients with severe aortic stenosis for potential transcatheter aortic valve implantation. Subjects and Methods Study Population Between June 2008 and October 2009, 80 patients referred for diagnostic workup for transcatheter aortic valve implantation were consecutively included in this study. All had severe aortic stenosis (aortic valve area < 0.8 cm 2 by means of trans thoracic echocardiography) and were deemed not amenable to traditional aortic valve replacement based on clinical judgment and a score of > 20% according to the European Scoring System of Cardiac Operative Risk Evaluation or “EuroSCORE” [6, 11]. Contraindications for DSCT examination were severely impaired renal function (estimated glomerular filtration rate, < 40 mL/min/1.73 m 2) or previous severe adverse reaction (anaphylactic [i.e., profound hypotension, bronchiospasm, severe urticaria]) to an iodinated contrast agent. Pa-


B

tients with an estimated glomerular filtration rate of between 40 and 60 mL/min/1.73 m 2 underwent IV volume expansion with isotonic crystalloid (1.0–1.5 mL/kg/h) for 3–12 hours before the procedure and continued for 6–24 hours afterward. In addition, these patients received 1,200 mg of N-acetylcystein orally twice a day before and after the procedure. The study was approved by our local ethics committee, and all patients provided written informed consent.

All examinations were performed on a DSCT scanner (Somatom Definition, Siemens Healthcare). A scout view of the thorax and abdomen was obtained to plan data acquisitions (Fig. 1A). After a single contrast medium injection, combined ECG-assisted scanning of the thorax and non-ECG-assisted scanning of the abdomen were performed. The total amount of contrast agent (iomeprol, Imeron 350, Nycomed) and flow rate were adapted to body weight: Patients weighing less than 70 kg received 110 mL of contrast agent at 4 mL/s and those weighing 70 kg or more received 130 mL at 4.5 mL/s. The contrast agent was injected via an 18-gauge needle in an antecubital vein and was followed by a saline bolus chaser of 50 mL administered at an equal flow rate. Data acquisition was initiated 6 seconds after the attenuation of a region of interest (ROI) placed in the ascending aorta reached 120 HU. The reference tube current–time product was twice as high for the lower part of the chest as for the upper part, with the boundary approximately 2 cm below the carina (Table 1 and Fig. 1A). Attenuation-based tube current modulation (CareDOSE, Siemens Healthcare) and prospective ECG-triggered tube

TABLE 1: Scanning Parameters for CT Angiography (CTA) ECG-Assisted CTA Parameter

Upper Chest

Lower Chest (Cardiac)

Aortoiliac CTA Below Diaphragm

ECG assistance

Retrospective ECG gatinga

No ECG assistance

Collimation (mm)

2 × 32 × 0.6

32 × 0.6

Section acquisition (mm)

2 × 64 × 0.6b

64 × 0.6b

Tube voltage (kV) Tube current–time product (mAs per rotation) Rotation time (ms) Pitch Scan direction

120 320c

120 160c

330

240c 330

0.2–0.43d

0.7

Craniocaudal

Craniocaudal

aHelical mode.

bBy means of z-flying focal spot technique.

cReference tube current–time product using attenuation-based tube current modulation (CARE Dose 4D,

Siemens Healthcare).

dDepending on heart rate.

873

Image Reconstruction CT data sets of the heart were reconstructed using a relative percentage approach (70% of the cardiac cycle [i.e., middiastole]), and an absolute forward approach (300 milliseconds after the R peak [i.e., end-systole]) with a slice thickness of 0.6 mm (0.4-mm increment) using a B26f kernel. Axial slices of the chest (70% of the cardiac cycle) and abdomen were reconstructed with a slice thickness of 1 mm (0.7-mm increment) using a B20f kernel. Fused coronal maximum intensity projections (MIPs) were generated (Fig. 1B). Curved planar reformatted and volume-rendered images were reconstructed for assessment of the aortoiliac vasculature (Aquarius iNtuition, TeraRecon). Aortic root measurements were performed using 3D multiplanar reformations (MPRs) on a dedicated postprocessing workstation (Syngo Multimodalitiy Workplace, Siemens Healthcare).

600

Mean Attenuation (HU)

500 400 300 200 100

874

or ta bi fu rc m at on io n i l ia Ex c te ar rn te C al ry om ili ac m on ar te fe ry m or al ar te ry

ta

la

tic

om

or

C

A

A

bd

om

in a

la

or

rt a

Fig. 2—Mean attenuation values () and SDs (vertical bars) along aortoiliac axis indicate relatively homogeneous contrast distribution among various vascular segments. Leap between attenuation in abdominal aorta at level of diaphragm and at renal artery orifices is caused by scan delay of 4 seconds between data acquisition of chest and abdomen. Measurements of aortic arch were taken at level of left subclavian artery and for thoracic descending aorta, at level of left atrium. LVOT = left ventricular outflow tract.

tal or coronal images); 2, moderate visibility of the anatomic details of the aortic root, pronounced step or motion artifacts; 1, poor (nondiagnostic), delineation of the anatomic details of the aortic root not possible because of severe step artifacts or motion. In case of disagreement, a final decision was obtained in consensus. For aortoiliac CTA, the image quality of the axial sections and coronal MIP reconstructions was evaluated using a similar 4-point scale: 4, excellent; 3, good; 2, moderate (still diagnostic, moderate delineation due to image noise or poor contrast attenuation); and 1, poor (nondiagnostic, delineation of the vessel contour not possible). The presence of breathing artifacts was noted for both scans.

Assessment of Image Quality Parameters Intraluminal attenuation (i.e., in Hounsfield units) and image noise were assessed as objective image quality parameters. The magnitude and uniformity of arterial enhancement were measured using circular ROIs at different levels along the aortoiliac axis, as outlined in Figure 2. Image noise was determined as the SD of the attenuation in the ROIs [15]. The diagnostic quality of both cardiac phases was independently graded by two observers using the following semiquantitative 4-point scale: 4, excellent visibility and differentiation of the anatomic details of the aortic root, image virtually free of image degradation; 3, good visibility of the anatomic details, only minor motion (blurring or virtually thickened aortic wall) or step artifacts (discontinuity in the aortic wall on sagit-

in a om

bd A

es

ce

nd

in g

ao

ar ch

rt a

tic

ao

or D

sc A

A

lv a en

of s Si nu

di ng

Va ls a

LV O T

tr iu m

0

ft a

current modulation [12] were used for radiation dose reduction, the latter with a pulsing window between 30% and 80% of the R-R cycle and tube current lowered to 20% of the maximum outside the pulsing window. The scanner was operated in the single-source mode for the abdomen (Table 1). A preceding delay of 4 seconds was necessary to change the scanning mode. Patients were instructed to sustain their breath-hold if possible or to continue shallow breathing. No additional β-blockade was administered. The scanning length and dose–length product (DLP) were noted from the scanner console. Radiation exposure was estimated by multiplying the DLP with a conversion factor k (0.017 mSv/mGy–1cm−1) as described previously [13, 14], and we recorded the heart rate, heart rate variability (SD of the heart rate), heart rhythm, and scan time.

Le


Blanke et al.

Aortic Root Measurements Coronal and sagittal oblique views through the aortic valve were reconstructed independently by each observer with the resulting intersection of both planes representing the centerline axis of inlet from the left ventricular outflow tract (LVOT) into the aortic root (Figs. 3A–3C), as reported by others [7]. Using the coronal and sagittal oblique views, we measured the diameter of the sinus of Valsalva and sinotubular junction (Fig. 3D). The distances between the aortic valve annulus and coronary ostia were assessed on reconstructed planes perpendicular to the aforementioned axis (Fig. 3E). The dimensions of the aortic annulus were assessed at the level of the basal attachments of all three aortic valve cusps, representing the inlet from the LVOT into the aortic root [16]. Using the cor-

onal and sagittal oblique views, the corresponding double oblique transverse view was adjusted to transect through the most caudal attachments of all three native cusps, defining the orientation and position of the annulus. To assess the cross-sectional area (CSA), the luminal contours were tracked on the double oblique transverse plane using automatic vessel analysis with manual correction (Fig. 3E). Using the equation for the area of a disk (πr 2), where r is its radius, the average diameter of the encircled area was calculated as follows: 2 × √(CSA / π). The maximum diameter of the mid ascending aorta was assessed using an MPR oriented perpendicular to the vessel axis. To assess intraobserver variability, measurements were repeated by one observer 2 weeks after inclusion of the last patient.

Assessment of the Aortoiliac Vasculature Minimal diameters of the common and external iliac arteries and the common femoral artery were independently assessed by two observers perpendicular to the centerline of the constructed curved planar reformatted images using electronic calipers. Window center and level settings were adjusted to reduce the effect of calcium blooming on vessel measurements. Measurements were repeated by one observer 2 weeks after inclusion of the last patient. The underlying causes for lumen narrowing at the minimal diameters, either vessel anatomy or vascular abnormalities (e.g., calcifications), were noted. Calcifications of vessel segments were independently scored by the two observers; they inter-



DSCT of the Aortic Root and Aortoiliac Vasculature

A

B

C

D

E

F

Fig. 3—80-year-old woman with severe aortic stenosis. A–C, Assessment of aortic root anatomy by reconstructing coronal oblique view (A), sagittal oblique view (B), and resulting transverse image (C). Color lines depict orientation of other views: red line = sagittal oblique, green line = coronal oblique, blue line = transverse. D, Diameters of sinus of Valsalva and sinotubular junction (arrows) were assessed using both coronal (D) and sagittal (not illustrated) views. E, Image shows measurement of distance between aortic annulus and left coronary sinus (arrow). F, Annulus diameters were assessed at level of most basal attachment points of all three cusps by means of planimetry. Red circle = cross-sectional area of aortic annulus for average diameter calculation, purple and orange lines = maximum and minimum diameters (not reported in this study).

preted longitudinal extension on MIP images and curved planar reformatted images and circumferential extension on axial studies using a 4-point scale: 0, vessel sections without calcified plaque; 1, mildly calcified with solitary hard punctiform plaques; 2, moderate calcifications (i.e., coherent plaques that covered < 50% of the vessel’s length); and 3, severe calcification (coherent plaque over > 50% of the vessel’s length, patchy plaque along the entire vessel length, or circumferential wall calcification). The presence of other vascular abnormalities, such as aneurysms, dissections, ulcers, or severe tortuosity, was also noted. Severe atheroma, defined as the presence of at least one calcium deposit or one clearly visualized area of hypoattenuation at least 4 mm thick adjacent to the aortic wall, were also noted.

Amenability to Transfemoral Vascular Access Amenability to transfemoral access was rated by two interventional radiologists with 7 and 3


years of experience in endovascular aortic repair in consensus. Based on information about vessel diameter, the presence of aortoiliac abnormalities, and the subjective assessment of vessel tortuosity, patients were classified regarding potential transfemoral access as amenable, potentially amenable after an additional vascular intervention, or not amenable. The relevant minimal diameter required was set at 7 mm.

Statistical Analysis All statistical analyses were performed using SPSS software (version 17.0, SPSS). Continuous variables are reported as means ± 1 SD. Categoric data are reported as frequencies and percentages. Interobserver agreement on grading the diagnostic quality and detecting atheroma of the aortic arch was calculated using kappa statistics. Intra- and interobserver agreements in continuous measurements were assessed by Bland-Altman analysis and Pearson’s analysis. Differences in aortic root

dimensions in middiastole and end-systole were evaluated with the paired Student’s t test. Subjective image quality scores for the different cardiac phases were compared using the Wilcoxon’s test for paired data. A p value of less than 0.05 was considered statistically significant.

Results CT examinations of all 80 patients were successfully performed and were well tolerated by all 80 patients. No patient underwent repeated enhanced scanning. The baseline characteristics of the study population are outlined in Table 2. The mean scan length was 24.0 ± 3.6 cm (range, 17.5–31.2 cm) for the chest and 29.3 ± 13.1 cm (range, 21.2–43.2 cm) for the abdomen. The mean scan length of the integrated cardiac scan was 13.8 ± 3.2 cm (range, 10.1–22.2 cm). The mean scan time was 16.5 ± 4.6 seconds (range, 10.0– 25.6 seconds) for the chest and 8.2 ± 1.4 sec-

875

Blanke et al. TABLE 2: Clinical and Echocardiographic Characteristics of the Study Population Characteristic Age (y), mean ± SD

Value 82.3 ± 7.8


Sex Male–female ratio (no. of patients)

33:47

Body mass index (kg/m2), mean ± SD

25.5 ± 5.4

EuroSCOREa (%), mean ± SD

35.1 ± 19.9

No. (%) of patients with previous CABG Aortic valve areab (cm2), mean ± SD Ejection fractionc (%), mean ± SD

18 (23) 0.67 ± 0.17 43 ± 13

Note—EuroSCORE = European Scoring System of Cardiac Operative Risk Evaluation CABG = coronary artery bypass graft. a A score of > 20% indicates not amenable to traditional aortic valve replacement [6, 11]. bPer continuity equation by transthoracic echocardiography. cBy means of the Simpson method.

onds (range, 5.6–11.4 seconds) for the abdomen. The mean effective radiation dose was 16.1 ± 4.1 mSv and 7.5 ± 2.7 mSv, respectively. The mean effective radiation dose of the whole examination was 23.7 ± 4.5 mSv. For ECG-gated CTA of the chest, the mean heart rate during data acquisition was 71.3 ± 13.0 beats per minute (bpm) (range, 51.6– 106.2 bpm), and the mean heart rate variability was 7.9 ± 6.5 bpm (0.5–23.1 bpm). ECG tracings obtained during data acquisition showed sinus rhythm in 57 patients (71%) and atrial fibrillation in 23 patients (29%). Heart rate variability was significantly higher in patients with atrial fibrillation than in those without atrial fibrillation (13.4 ± 5.1 bpm vs 5.1 ± 5.8 bpm, respectively; p < 0.001). Image Quality The mean image noise was 29.6 ± 8.0 HU and 20.6 ± 5.0 HU in the ascending and abdominal aorta above the bifurcation, respectively. The mean attenuation values are reported in Figure 2. The mean subjective image quality for the aortic root was not significantly different for middiastolic and end-systolic reconstructions in patients with sinus rhythm (n = 57; mean subjective score, 4.0 vs 3.9, respectively; p = 0.083), with most studies ranked as grade 4 (motion or step artifacts present in one of 57 vs four of 57 patients). In patients with atrial fibrillation (n = Fig. 4—84-year-old man with atrial fibrillation (mean heart rate, 89 beats per minute [bpm]; range, 42–115 bpm). A, Multiplanar reconstructions show motion and step artifacts in middiastolic reconstructions, evident by blurring of aortic wall and discontinuity of semilunar cusps and anterior mitral valve leaflet. B, Multiplanar reconstructions in end-systole are virtually artifact-free.

876

oral arteries was 3.9 ± 0.2 (κ = 0.89), 3.8 ± 0.4 (κ = 0.86), and 3.7 ± 0.5 (κ = 0.84), respectively. For the common and external iliac arteries, image quality was diagnostic in all 80 patients, whereas for the common femoral arteries, image quality was nondiagnostic in one patient on one side because a unilateral total hip endoprosthesis caused severe beam hardening. Minor breathing artifacts during abdominal scanning were noted in 13 patients but did not impair image interpretation.

23), motion and step artifacts were present in most of the middiastolic image reconstructions (20/23), whereas end-systolic image reconstructions revealed minor step artifacts in only two of the 23 patients (mean subjective score, 2.7 vs 3.9; p < 0.001) (Fig. 4). Overall agreement between both observers was good (κ = 0.86). Four patients had breathing artifacts at the lower end of the thoracic scan range that did not compromise diagnostic quality. The quality of the images of the thoracic and abdominal aorta was diagnostic in all 80 patients, with an average score of 4.0 ± 0.0 (κ = 1.00) at all levels. The mean grade of image quality for the common iliac arteries, external iliac arteries, and common fem-

Anatomic Measurements of the Aortic Root and Valve Aortic root dimensions were assessable in all CT examinations at end-systole. All aortic valves were tricuspid and were heavily calcified. Results of the CT measurements of the dimensions of the aortic root and the maximum diameter of the ascending aorta and the results of the Bland-Altman analysis and Pearson’s analysis are listed in Table 3. Interobserver correlation coefficients ranged from 0.93 to 0.99 (p < 0.001), and intraobserver correlation coefficients ranged from 0.95 to 0.99 (p < 0.001). The cross-sectional area of the annulus and the mean calculated average annulus diameter were not significantly different for end-systolic and middiastolic reconstructions (24.1 ± 2.9 mm vs 24.0 ± 2.9 mm, p = 0.191; 462.7 ± 114.6 mm2, 457.3 ± 115.0 mm2, p = 0.184; n = 57). Middiastolic measurements were not performed in patients with atrial fibrillation because of impaired image quality and the expected con-

A

B


DSCT of the Aortic Root and Aortoiliac Vasculature TABLE 3: Results of CT Measurements of the Aortic Root and Ascending Aorta

CT Measurement

Mean for Both Observers

Mean Interobserver Difference (mm)a

95% Limitsb

Pearson’s Correlation Coefficient

−29.0, 45.2

0.99

SEE

Mean Intraobserver Difference (mm)a

95% Limitsb


SEE

−30.0, 32.0

0.99

15.87


Aortic annulus diameter Cross-sectional area (mm2)

462.7 ± 114.6 (261.9–795.2)

8.1 ± 18.9

19.2

1.0 ± 15.8

24.1 ± 2.9 (18.3–31.8)

0.20 ± 0.52

−0.8, 1.2

0.99

0.53

0.03 ± 0.41

−0.8, 0.8

0.99

0.411

Coronal oblique view (mm)

32.3 ± 3.5 (27.4–39.2)

0.23 ± 0.84

−1.4, 1.9

0.97

0.084

−0.21 ± 0.69

−1.6, 1.2

0.98

0.071

Sagittal oblique view (mm)

30.6 ± 3.9 (24.2–39.2)

0.25 ± 0.98

−1.7, 2.2

0.97

0.099

0.03 ± 0.78

−1.5, 1.6

0.98

0.078

Coronal oblique view (mm)

27.3 ± 3.7 (21.9–35.1)

0.11 ± 0.96

−1.8, 2.0

0.97

0.098

−0.09 ± 0.65

−1.4, 1.2

0.99

0.066

Sagittal oblique view (mm)

26.2 ± 3.4 (20.1–35.7)

−0.09 ± 0.89

−1.8, 1.7

0.97

0.088

−0.04 ± 0.78

−1.6, 1.5

0.98

0.079

Left coronary artery ostium (mm)

14.0 ± 3.0 (9.3–22.7)

−0.08 ± 1.00

−2.0, 1.9

0.93

0.102

0.22 ± 0.81

−1.4, 1.8

0.96

0.078

Right coronary artery ostium (mm)

14.2 ± 3.0 (8.5–20.3)

−0.02 ± 0.94

−1.9, 1.8

0.94

0.095

−0.17 ± 1.00

−2.1, 1.7

0.95

0.096

Average diameter (mm) Sinus of Valsalva

Sinotubular junction

Distance (mm) between aortic annulus and

Note—Data reported are for image reconstructions at end-systole. Ranges are presented in parentheses. SEE = standard error of the estimate. aBland-Altman analysis. bCalculated as mean difference ± 1.96 × SD, data reported for diastolic reconstructions. Correlation was significant for all inter- and intraobserver analyses (p < 0.001).

comitant inaccuracy of measurements due to step artifacts. Fifteen patients (19%) showed annulus diameters of more than 27 mm. Aorta and Iliac Vasculature Exophytic atheroma or bulky calcifications of the aortic arch larger than 4 mm were present in eight patients (10%, κ = 1.00). An ascending aortic aneurysm (> 45 mm) was seen in eight patients, a thoracoabdominal aortic aneurysm in one patient, and an infrarenal abdominal aortic aneurysm in four patients. Penetrating ulcers were diagnosed in three patients, and focal dissections of the abdominal aorta or iliac vessels were detected in four patients. False aneurysms of the common femoral or superficial femoral arteries were observed in three patients. The average minimum diameter for the iliofemoral axis and inter- and intraobserver variabilities according to the Bland-Altman analysis are reported in Table 4. Interobserver correlation coefficients ranged from 0.93 to 0.98 (p < 0.001), and intraobserver correlation coefficients ranged from 0.95 to 0.98 (p < 0.001). Minimal diameters of 7 mm or less were found in 40 patients (50%): on


one side of the aortoiliac axis in 11 patients (14%) and on both sides in 29 patients (36%). Vessel anatomy (i.e., small size without additional abnormalities) was identified as the underlying cause of stenosis in 13 patients, vessel calcification in 22 patients, and soft atheroma in five patients. Vessel calcifications of varying degrees and distributions were present in all patients. Calcifications were present in the ascending aorta in 28 patients (35%; mean grade, 1.3 ± 0.6; κ = 0.82), aortic arch in 74 patients (93%; 1.7 ± 0.6; κ = 0.75), descending thoracic aorta in 58 patients (73%; 1.4 ± 0.6; κ = 0.77), and abdominal aorta in 78 patients (98%; 2.4 ± 0.7; κ = 0.79). Calcifications of the common iliac arteries, external iliac arteries, and common femoral arteries were present in 76, 46, and 59 patients on the right (95%, 58%, 74%; mean grade, 1.9 ± 0.9, 0.8 ± 0.9, 1.2 ± 1.0; κ = 0.82, 0.84, 0.79, respectively) and in 76, 48, and 66 patients on the left (95%, 60%, 83%; 1.9 ± 0.9, 0.9 ± 0.9, 1.0 ± 1.0; κ = 0.79, 0.84, 0.82, respectively). Amenability to Transfemoral Vascular Access Of the 160 iliofemoral axes evaluated, 89 (56%) were deemed amenable to transfemo-

ral access, 38 (24%) were deemed nonamenable, and 33 (21%) were deemed amenable only after a preceding vascular intervention because of focal narrowing. The reasons for nonamenability (n = 38) were as follows: small vessel diameter (10 axes), borderline iliofemoral diameter associated with severe calcification (eight axes [Fig. 5]), and severe calcifications with significant long-segment narrowing of the iliofemoral arteries or narrowing not amenable to interventional repair (16 axes). Severe tortuosity was found in the four remaining patients, affecting six axes (Fig. 6). Of those six axes, two were deemed nonamenable solely because of tortuosity, whereas concomitant vessel narrowing was noted in the remaining four axes. On a per-patient basis, 38 of the 80 patients (48%) were classified as amenable to instant transfemoral access on both sides, and instant transfemoral access was considered possible for just one side in 13 patients (16%). Of the remaining 29 patients (36%), 11 patients (14%) were considered amenable to transfemoral access on at least one side after a preceding intervention, whereas 18 patients (23%) were judged to be nonamenable to transfemoral access.

877

Blanke et al. TABLE 4: Minimal Diameters of the Iliofemoral Vasculature Diametera

Mean Interobserver Difference (mm)b

95% Limits (mm)b


SEE

Mean Intraobserver Difference (mm)b

95% Limits (mm)b


SEE

Mean (mm)

Range (mm)

Right

8.8 ± 2.1

4.6–13.3

0.05 ± 0.46

−0.84, 0.95

0.98

0.463

0.04 ± 0.45

−0.83, 0.92

0.98

0.452

Left

8.8 ± 2.0

4.6–13.0

0.07 ± 0.46

−0.84, 0.98

0.98

0.463

0.05 ± 0.41

−0.76, 0.86

0.98

0.415

Right

7.3 ± 1.1

4.2–9.3

−0.08 ± 0.42

−0.90, 0.74

0.93

0.419

−0.10 ± 0.35

−0.79, 0.59

0.95

0.357

Left

7.4 ± 1.4

3.2–9.6

0.01 ± 0.47

−0.91, 0.93

0.95

0.476

0.04 ± 0.47

−0.88, 0.95

0.95

0.476

Right

7.0 ± 1.4

3.6–9.6

−0.10 ± 0.47

−1.02, 0.88

0.96

0.399

−0.06 ± 0.40

−0.85, 0.73

0.95

0.472

Left

7.2 ± 1.5

3.7–10.4

−0.90, 0.91

0.95

0.420

−0.05 ± 0.44

−0.91, 0.81

0.96

0.373

Artery


Common iliac artery

External iliac artery

Common femoral artery 0.00 ± 0.46

Note—Correlation was significant for all inter- and intraobservers analyses (p < 0.001). SEE = standard error of the estimate. aPerpendicular to a single path curved planar reformatted. bBland-Altman analysis.

Discussion With more transcatheter aortic valve implantation procedures being performed, MDCT is increasingly used to plan the procedure [17]. In this study, we investigated a combined one-stop-shop DSCT protocol for the evaluation of patients with severe aortic stenosis for potential transcatheter aortic valve implantation. Assessment of the individual aortic root dimensions is crucial for prosthesis sizing because improper valve sizing may result in paravalvular leakage and prosthesis migration [18], and this assessment is essential for placement because covering the coronary ostia by the upper part of the prosthesis and even occlusion of the left coronary artery can occur [18, 19]. Currently, there are two prosthesis types available: the Edwards SAPIEN transcatheter heart valve (Edwards Lifesciences) and the CoreValve ReValving System (Medtronic). Prosthesis diameters range between 23 and 29 mm, which are suitable for annulus sizes of 18–27 mm [20]. The use of ECG-assisted MDCT to assess the dimensions of the aortic annulus and the location of the coronary ostia has been reported in several studies [7, 21]. These measurements, as shown in this study, are reliable and reproducible with low inter- and intraobserver differences and high inter- and intraobserver correlations. Importantly, diagnostic image quality was achieved in all patients regardless of heart rate or rhythm. A considerable percentage of patients (29%) in our study group had atrial fibrillation during data acquisition. In these patients, good image quality was ob-

878

tainable almost exclusively during end-systole, whereas diastolic reconstructions were hampered by step and motion artifacts. Similar to Tops et al. [7], we found no significant difference in aortic root measurements between end-systole and middiastole in patients with sinus rhythm; thus, we can assume that these measurements

can be used interchangeably in case of impaired image quality during diastole, just as in patients with atrial fibrillation. Furthermore, this study shows that there is no need for additional frequency control such as β-blockade for DSCT of the aortic root given the consistent diagnostic image quality of end-systolic reconstructions.

A

B

Fig. 5—Coronal maximum intensity projections and transverse views of curved planar reconstructions of aortoiliac vasculature. Avg. = average, Min. = minimum, Max. = maximum. A, 82-year-old woman with hardly any calcification of iliac arteries (red) but small vessel anatomy. B, 84-year-old woman with extensive calcifications of iliac arteries (blue). Both patients were deemed nonamenable to transfemoral transcatheter aortic valve implantation.



DSCT of the Aortic Root and Aortoiliac Vasculature

A

B

C

Fig. 6—84-year-old man with severe aortic stenosis. A, Volume-rendered image of aortoiliac vasculature (30° left anterior oblique projection) depicts severe bilateral tortuosity of external iliac artery. B and C, Corresponding curved multiplanar reformations along centerline of right and left iliac arteries (B and C, respectively). This patient was deemed not amenable to transfemoral transcatheter aortic valve implantation.

Besides needing information about the anatomy of the aortic root, understanding the arterial anatomy before the procedure is important because there is a spectrum of access strategies, ranging from a percutaneous transfemoral approach to a transapical approach [6]. For the transfemoral approach, information about the anatomy and about the presence of aortoiliac axis disease is of paramount importance. Compared with iliofemoral angiography, which is routinely used in the preoperative assessment of transcatheter aortic valve implantation patients, evaluation of aortoiliofemoral anatomy by MDCT provides high-resolution isotropic volume data that allow more definitive assessment of the 3D course of the iliofemoral arteries and of the distribution and extent of calcification and plaque formation. Furthermore, MDCT evaluation is not limited to the iliofemoral arteries because the aorta can be assessed in its entire length, including the aortic arch.


MDCT can identify atheroma or bulky calcifications in the aortic arch [22], which might cause stroke when dislodged by mechanic manipulation during an intervention [18]. Furthermore, there is broad experience using MDCT in endovascular aortic aneurysm repair, where it provides sufficient anatomic information for evaluating the feasibility of endograft repair, choosing the appropriate endograft, and optimizing procedural outcome [9], thus making prior angiography dispensable [23]. This experience can be applied to transcatheter aortic valve implantation. In transcatheter aortic valve implantation, potential contraindications of the transfemoral approach are severe calcification, tortuosity, or small diameter (< 6–9 mm depending on the device used) of the iliac arteries [24]. Concentric or circumferential calcifications are considered a relative contraindication, particularly when borderline vessel diameters are present because calcifications may

limit the arterial expandability to accommodate a large-profile delivery sheath. At our institution, because we use 18French sheaths for the transfemoral implantation of the self-expanding CoreValve ReValving System, the minimal vessel diameter required is 7 mm. As shown in this study, the aortoiliac axis can easily be assessed using a combined examination protocol with a single application of contrast medium; diagnostic image quality of all vascular segments was achieved in all patients except one. Inter- and intraobserver correlations were excellent and inter- and intraobserver differences, low. Based on information acquired by dualsource CTA, instant transfemoral access was deemed not suitable in 36% of the patient cohort, most often because of small vessel size or vessel narrowing due to calcification in the setting of advanced peripheral artery disease (Fig. 5). Severe tortuosity of the iliac arteries restricted iliofemoral access in four patients

879


Blanke et al. (Fig. 6). These findings concur with those of Kurra et al. [25] who reported in their recent study of 100 patients that one third of patients with severe aortic stenosis evaluated for trans catheter aortic valve implantation had unfavorable atherosclerotic iliofemoral disease by means of MDCT. Patients being evaluated for potential transcatheter aortic valve implantation undergo an extensive workup. Aortic stenosis is often diagnosed via transthoracic echocardiography and subsequent transesophageal echocardiography. Cardiac catheterization is required for the preoperative evaluation of coronary artery disease; in addition, aortoiliac angiography is required in case of a potential transfemoral approach [24]. As shown in this study, one major advantage of DSCT in planning the procedure lies in its capacity to provide the most essential information for planning transcatheter aortic valve implantation in a single examination. Importantly, besides identifying patients with limited transfemoral access, we found 15 patients with annulus diameters of more than 27 mm; thus, they were not amenable to transcatheter aortic valve implantation with the currently commercially available devices. Furthermore, severe atheroma or bulky calcifications were present in five patients in this study, constituting a relative contraindication for the transfemoral approach [24]. Although not evaluated in this study, MDCT can provide additional information, such as aortic valve area quantification or cardiac function assessment, as reported in the literature [26, 27]. Instead of two examinations— namely, cardiac CT and aortoiliac CTA on separate occasions—a combination of the two can simplify logistic needs and reduce the total amount of contrast medium used, thereby serving as a one-stop-shop screening examination. However, one must keep in mind that this CT protocol provides a vast amount of information, of which a good portion must be acquired by postprocessing of CT data. Postprocessing can be time consuming and still takes up to 15 minutes per examination even for an experienced reader. The combined DSCT examination—in particular, the scanning time of this protocol—was tolerated by the most patients undergoing workup for trans catheter aortic valve implantation despite the study group’s relatively advanced age. Study Limitations This study is limited by the fact that we did not analyze the impact of preprocedural

880

CT findings or assessed anatomic dimensions on decision making and procedural outcome. These factors deserve attention in future studies. Furthermore, we did not compare aortic root dimensions on DSCT with measurements using other imaging techniques, such as echocardiography, because this comparison was not the primary aim of this study. However, the accuracy of CT for aortic annulus assessment has been the focus of recent studies [7, 16, 28]. Also, we did not compare aortoiliac CTA with conventional aortoiliac angiography, but there are ample studies in the literature showing the accuracy of CTA in assessing the aortoiliac axis [29]. Although performing precise measurements of luminal dimensions with MDCT is limited by partial volume averaging of calcium, which is associated with overestimation of the calcified plaque area and luminal stenosis, CTA is not limited to selected angiographic planes, thereby allowing cross-sectional measurements of the vessels at any point along the vessel centerline. Furthermore, the cutoff diameter for transfemoral access was arbitrarily set to 7 mm on the basis of the 18-French sheaths used at our institution. In clinical routine and depending on the experience of the interventionalist, transfemoral access may be attempted in patients with even lower vessel diameters if the stenosis is localized, assuming arterial distensibility if no circumferential calcification is present. Our CT protocol is associated with considerable radiation exposure to the patient. However, compared with other CT scanner systems, DSCT allows varying mAs settings for different ECG-assisted data acquisition levels [30], thus reducing the need for high mAs settings throughout the entire chest. Moreover, the use of a single examination protocol avoids redundant coverage of body parts. The radiation dose may be reduced by using lower tube voltage; shorter pulsing windows; or prospective ECG-triggered sequential data acquisition, as reported in the literature on cardiac CT [31]. Although not addressed in this study, these technical features should be investigated for future use. In conclusion, the dimensions of the aortic root and the aortoiliac vasculature can be assessed with a combined single-dose contrastenhanced DSCT protocol, thereby allowing determination of patient eligibility for trans catheter aortic valve implantation, prosthesis sizing, and evaluation of the access route in one examination. With the anticipated rise in the number of transcatheter aortic valve im-

plantation procedures in the future, CT has the potential to serve as a screening tool to evaluate the feasibility of the transcatheter aortic valve implantation procedure before invasive examinations are considered. References 1. Iung B, Baron G, Butchart EG, et al. A prospective survey of patients with valvular heart disease in Europe: the Euro Heart Survey on Valvular Heart Disease. Eur Heart J 2003; 24:1231–1243 2. Carabello BA, Crawford FA Jr. Valvular heart disease. N Engl J Med 1997; 337:32–41 3. Bonow RO, Carabello BA, Kanu C, et al.; American College of Cardiology/American Heart Association Task Force on Practice Guidelines; Society of Cardiovascular Anesthesiologists; Society for Cardiovascular Angiography and Interventions; Society of Thoracic Surgeons. ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease) developed in collaboration with the Society of Cardiovascular Anesthesiologists—endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons. Circulation 2006; 114:e84– e231 4. Vahanian A, Baumgartner H, Bax J, et al.; Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology; ESC Committee for Practice Guidelines. Guidelines on the management of valvular heart disease: the Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology. Eur Heart J 2007; 28:230–268 5. Iung B, Cachier A, Baron G, et al. Decision-making in elderly patients with severe aortic stenosis: why are so many denied surgery? Eur Heart J 2005; 26:2714–2720 6. Vahanian A, Alfieri O, Al-Attar N, et al.; European Association of Cardio-Thoracic Surgery; European Society of Cardiology; European Association of Percutaneous Cardiovascular Interventions. Transcatheter valve implantation for patients with aortic stenosis: a position statement from the European Association of Cardio-Thoracic Surgery (EACTS) and the European Society of Cardiology (ESC), in collaboration with the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur Heart J 2008; 29: 1463–1470 7. Tops LF, Wood DA, Delgado V, et al. Noninvasive evaluation of the aortic root with multislice computed tomography: implications for transcatheter



DSCT of the Aortic Root and Aortoiliac Vasculature aortic valve replacement. JACC Cardiovasc Imaging 2008; 1:321–330 8. Murray D, Ghosh J, Khwaja N, Murphy MO, Baguneid MS, Walker MG. Access for endovascular aneurysm repair. J Endovasc Ther 2006; 13:754–761 9. Prokop M. Multislice CT angiography. Eur J Radiol 2000; 36:86–96 10. Theisen D, von Tengg-Kobligk H, Michaely H, Nikolaou K, Reiser MF, Wintersperger BJ. CT angiography of the aorta [in German]. Radiologe 2007; 47:982–992 11. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999; 16:9–13 12. Jakobs TF, Becker CR, Ohnesorge B, et al. Multislice helical CT of the heart with retrospective ECG gating: reduction of radiation exposure by ECG-controlled tube current modulation. Eur Radiol 2002; 12:1081–1086 13. Macari M, Chandarana H, Schmidt B, Lee J, Lamparello P, Babb J. Abdominal aortic aneurysm: can the arterial phase at CT evaluation after endovascular repair be eliminated to reduce radiation dose? Radiology 2006; 241:908–914 14. Menzel H, Schibilla H, Teunen D, eds. European guidelines on quality criteria for computed tomography. Luxemburg, Luxembourg: European Commission, 2000: Publication no. EUR 16262 EN 15. Stolzmann P, Leschka S, Scheffel H, et al. Dualsource CT in step-and-shoot mode: noninvasive coronary angiography with low radiation dose. Radiology 2008; 249:71–80 16. Schultz CJ, Moelker A, Piazza N, et al. Three dimensional evaluation of the aortic annulus using multislice computer tomography: are manufac-


turer’s guidelines for sizing for percutaneous aortic valve replacement helpful? Eur Heart J 2010; 31:849–856 17. Leipsic J, Wood D, Manders D, et al. The evolving role of MDCT in transcatheter aortic valve replacement: a radiologists’ perspective. AJR 2009; 193:823; [web]:W214–W219 18. Webb JG, Chandavimol M, Thompson CR, et al. Percutaneous aortic valve implantation retrograde from the femoral artery. Circulation 2006; 113: 842–850 19. Grube E, Laborde JC, Gerckens U, et al. Percutaneous implantation of the CoreValve self-expanding valve prosthesis in high-risk patients with aortic valve disease: the Siegburg first-in-man study. Circulation 2006; 114:1616–1624 20. Piazza N, De Jaegere P, Schultz C, Becker A, Serruys PW, Anderson RH. Anatomy of the aortic valvar complex and its implications for transcatheter implantation of the aortic valve. Circ Cardiovasc Interv 2008; 1:74–81 21. Stolzmann P, Knight J, Desbiolles L, et al. Remodelling of the aortic root in severe tricuspid aortic stenosis: implications for transcatheter aortic valve implantation. Eur Radiol 2009; 19:1316–1323 22. Thenappan T, Ali Raza J, Movahed A. Aortic atheromas: current concepts and controversies—a review of the literature. Echocardiography 2008; 25:198–207 23. Wyers MC, Fillinger MF, Schermerhorn ML, et al. Endovascular repair of abdominal aortic aneurysm without preoperative arteriography. J Vasc Surg 2003; 38:730–738 24. Vahanian A, Alfieri OR, Al-Attar N, et al. Transcatheter valve implantation for patients with aortic stenosis: a position statement from the European

Association of Cardio-Thoracic Surgery (EACTS) and the European Society of Cardiology (ESC), in collaboration with the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur J Cardiothorac Surg 2008; 34:1–8 25. Kurra V, Schoenhagen P, Roselli EE, et al. Prevalence of significant peripheral artery disease in patients evaluated for percutaneous aortic valve insertion: preprocedural assessment with multidetector computed tomography. J Thorac Cardiovasc Surg 2009; 137:1258–1264 26. Pouleur AC, le Polain de Waroux JB, Pasquet A, Vanoverschelde JL, Gerber BL. Aortic valve area assessment: multidetector CT compared with cine MR imaging and transthoracic and transesophageal echocardiography. Radiology 2007; 244:745–754 27. Stolzmann P, Scheffel H, Trindade PT, et al. Left ventricular and left atrial dimensions and volumes: comparison between dual-source CT and echocardiography. Invest Radiol 2008; 43:284–289 28. Willmann JK, Weishaupt D, Lachat M, et al. Electrocardiographically gated multi-detector row CT for assessment of valvular morphology and calcification in aortic stenosis. Radiology 2002; 225:120–128 29. Tillich M, Bell RE, Paik DS, et al. Iliac arterial injuries after endovascular repair of abdominal aortic aneurysms: correlation with iliac curvature and diameter. Radiology 2001; 219:129–136 30. Johnson TR, Nikolaou K, Becker A, et al. Dualsource CT for chest pain assessment. Eur Radiol 2008; 18:773–780 31. Alkadhi H, Stolzmann P, Scheffel H, et al. Radiation dose of cardiac dual-source CT: the effect of tailoring the protocol to patient-specific parameters. Eur J Radiol 2008; 68:385–391

881