Eur Radiol (2013) 23:2482–2491 DOI 10.1007/s00330-013-2865-3
COMPUTED TOMOGRAPHY
Optimal table feed in run-off CT angiography in patients with abdominal aortic aneurysms T. Werncke & C. von Falck & M. Wittmann & T. Elgeti & F. K. Wacker & B. C. Meyer
Received: 4 January 2013 / Revised: 20 March 2013 / Accepted: 25 March 2013 / Published online: 19 May 2013 # European Society of Radiology 2013
Abstract Objectives To assess the influence of different table feeds (TFs) on vascular enhancement and image quality in patients with an abdominal aortic aneurysm (AAA) undergoing computed tomography (CT) angiography of the lower extremities (run-off CTA). Methods Seventy-nine patients (71±8 years) with an AAA (>30 mm) who underwent run-off CTA between January 2004 and August 2011 were included in this retrospective institutional review board-approved study. Run-off CTA was conducted using 16- and 64-row CT. The range of TFs was 30–86 mm/s and was categorised in quartiles TF1 (32.6± 1.9 mm/s), TF2 (38.9±0.9 mm/s), TF3 (43.9±3.1 mm/s) and TF4 (57.4±10.5 mm/s). Image quality was rated independently by two radiologists and vessel enhancement was assessed. Results Image quality was diagnostic at all aortic, pelvic and almost all thigh levels. Below the knee, the number of diagnostic levels was highest for TF1 and decreased to TF4.
T. Werncke (*) Klinik für Radiologie, Charité Universitätsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12200 Berlin, Germany e-mail:
[email protected] C. von Falck : F. K. Wacker : B. C. Meyer Institute of Diagnostic and Interventional Radiology, Medical School Hannover, Carl-Neuberg-Str. 1, 30625 Hannover, Germany C. von Falck e-mail:
[email protected] M. Wittmann Institut für diagnostische und interventionelle Radiologie, DRK Kliniken Berlin | Westend, Spandauer Damm 130, 14050 Berlin, Germany T. Elgeti Klinik und Hochschulambulanz für Radiologie, Charité Universitaetsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12200 Berlin, Germany
Arterial enhancement between the aorta and fibular trunk was not different in all TF groups, P>0.05. At the calf and foot strongest arterial enhancement was noted for TF1 and TF2 and decreased to TF4, P<0.01. Conclusion Results indicate that the highest image quality of run-off CTA in patients with an AAA may be obtained using table feeds measuring 30–35 mm/s. Key Points • CTA has become a key investigation for peripheral vascular disease. • Run-off CTA is more complex in patients with an abdominal aortic aneurysm. • Run-off CTA is feasible with a short bolus of intravenous contrast medium. • A constant 30–35 mm/s table feed provides the highest likelihood of diagnostic images. Keywords Multidetector computed tomography . Abdominal aortic aneurysm . Retrospective studies . Angiography . Lower extremity Abbreviations AAA abdominal aortic aneurysm TF table feed Run-off CTA CT angiography of the lower extremities
Introduction Multidetector computed tomography (MDCT) angiography of the lower extremities (run-off CTA) is a fast and reliable imaging method for the evaluation of peripheral arterial disease (PAD). Its high sensitivity and specificity for stenosis detection and quantification have been demonstrated in several studies [1]. Run-off CTA requires an optimised synchronisation of contrast material (CM) administration and table feed (TF) during image acquisition to minimise the risk of outpacing
Eur Radiol (2013) 23:2482–2491
the CM bolus on the one side and to minimise the venous enhancement on the other. This is already challenging in patients with PAD only, which is reflected by the numerous acquisition techniques published in the literature [2–7]. In patients with abdominal aortic aneurysm (AAA), the turbulent blood flow in the aneurysm sac changes the haemodynamic situation and reduces distal blood flow [8]. Thus, the bolus transit time is expected to increase and the use of a lower table feed seems reasonable in these patients to provide sufficient contrast enhancement of the abdominal aorta and the lower limb arteries. Several strategies have been proposed to ensure synchronisation of arterial enhancement and image acquisition for run-off CTA. Simple protocols use preselected table speeds with threshold-based triggering [3, 6]. More sophisticated and timeconsuming protocols utilise at least one test bolus to determine the transit time by monitoring the arterial enhancement at the aortic [9] or knee level [4, 7] in order to calculate a patientspecific table feed. However, the individual adaptation of the acquisition protocol is time-consuming and requires the presence of a physician or a specialised technician during the examination. With respect to a straightforward workflow in the daily clinical routine, a constant acquisition protocol is preferable. Therefore, the purpose of this retrospective study was twofold: firstly to evaluate the influence of different table feeds on the vascular enhancement and image quality, and secondly to demonstrate the applicability of a standardised table feed for run-off CTA to image the peripheral arterial vasculature of patients with AAA.
Materials and methods
2483
Four TF groups (TF1, TF2, TF3, TF4) were defined by quartiles of the overall range of table feeds used in this study and all patients were assigned to one of the four TF groups depending on the TF applied (Table 1). Image acquisition Run-off CTA was performed using either a 16-slice MDCT (n = 24) (Somatom Sensation; Siemens Healthcare; Forchheim, Germany) or 64-slice MDCT (n=55) (Somatom Definition, Siemens Healthcare; Forchheim, Germany) (Table 2). Initial run-off CTA studies were performed using a standard TF of 40 mm/s. However, as part of an institutional protocol optimisation process, different TFs, ranging between 30 mm/s and 86 mm/s independently on the presence of an AAA, were applied. All patients were placed supine with feet first in the gantry. Tourniquets were tied round each thigh in order to reduce the venous return [12]. The examination covered the range between the costodiaphragmatic recess and the forefoot. Run-off CTA was performed after intravenous injection of 100 ml Iomeprol (400 mg I/ml, Imeron 400; Bracco, Milan, Italy) followed by a 60-ml saline flush (NaCl 0.9 %) at a flow rate of 4.0 ml/s using a dual power injector (Stellant, Medrad, Volkach, Germany). Image acquisition was started 5 s after a threshold attenuation of 250 HU was reached in the suprarenal aorta. The enhancement time curve of the monitoring images in the aorta was observed by either a trained technician or a radiologist according to our institutional quality standards to allow for a manual acquisition start if the predefined threshold was not reached. Images were reconstructed using a soft kernel (B25f). The dose-length product (DLP) and the volume computed tomography dose index (CTDIVol) were recorded (Table 2).
Study design Image reconstruction This retrospective study was approved by the institutional review board with a waiver of consent granted (EA4/116/11). A total of 1,011 patients who underwent a run-off CTA between January 2004 and August 2011 were evaluated for study inclusion. All patients with known or incidentally detected AAA with a minimum outer diameter of 30 mm were included [10, 11]. The maximum outer and luminal diameter of the AAA, the presence of cardiovascular risk factors and PAD, the body weight and patient size were recorded, and the body mass index (BMI) was calculated. Patients with additional aneurysms in the lower extremities were excluded. Eventually, 79 patients (69 men, mean age 71±8 years, range 46–87; 10 women, 79±6 years, range 69–85 years) met the inclusion criterion. Cardiovascular risk factors in the study population were arterial hypertension in 64 patients (81 %), a smoking history in 61 patients (77 %), hypercholesterolaemia in 41 patients (52 %) and diabetes mellitus in 9 patients (12 %).
Post-processing and image analysis were performed on a Thin Client System (Visage 7; Visage Imaging, Berlin, Germany). Semi-automatic bone removal was applied and rotating maximum intensity projections (MIPs) were created separately at three overlapping regions (region 1: from the suprarenal aorta to the femoral bifurcation; region 2: from the common femoral artery to the origin of the tibiofibular trunk; region 3: from the P3 segment of the popliteal artery to the forefoot arteries) over a viewing range of 180° in 15° intervals. If vessel calcifications interfered with interpretation of the MIP images, additional curved planar reformats (CPR) were generated. Qualitative image assessment Qualitative assessment of the bolus timing was performed independently by two radiologists with 6 years’ experience
2484
Eur Radiol (2013) 23:2482–2491
Table 1 Patient demographics Characteristic
Table feed group TF1
TF2
TF3
TF4
Mean table feed [mm/s] a Mean range (minimum to maximum) [mm/s] Mean acquisition time [s] a Number of patients Overall
32.6±1.9 29-35 39±4
38.9±0.9 35-40 34±2
43.9±3.1 40-48 30±3
57.4±10.5 48-86 23±4
20
21
20
18
Male/female Mean body weight [kg] a Mean body height [cm] a Mean BMI [kg/m2] a AAA Mean maximum diameter [mm] a Mean maximum luminal diameter [mm] a Fontaine stage [number of patients] Mild (0-IIb) 0 1 2a 2b Severe (III and IV) 3 4
16/4 77±15 173±10 26±6
19/2 79±16 175±7 26±4
19/1 79±14 175±7 26±4
15/3 71±13 172±7 24±3
54±7 38±10
54±15 35±9
51±15 32±8
56±14 37±13
18 7 0 2 9 2 0 2
17 0 9 1 7 4 2 2
18 8 1 3 6 2 1 1
P value
0.23 0.44 0.22
b
0.45 0.38 0.70
b
b b
b c
17 12 0 3 2 1 1 0
AAA abdominal aortic aneurysm, TF table feed group a
Data are the means (± standard deviation)
b
P value of the Kruskal-Wallis test
c
P value of the chi-squared test
in reading run-off CTA. A maximum of nine arterial level regions were evaluated per patient, comprising the aortic level (from the suprarenal aorta to the aortic bifurcation), the left and right pelvic level region (from the aortic bifurcation to the femoral bifurcation), the left and right thigh level region (from the femoral bifurcation to the popliteal artery), the left and right calf level
region (from below the popliteal artery to the ankle) and the left and right foot level region (from the ankle to the forefoot). Synchronisation of image acquisition and contrast medium bolus was rated for each arterial level region independently by both readers using MIP as well as the axial slices using a previously published rating scale [3] as described in detail in
Table 2 Acquisition protocols
CTDI computed tomography dose index, DLP dose-length product a
Data are the means (± standard deviation)
16-slice MDCT
64-slice MDCT
MDCT Collimation Tube voltage Tube current Tube current modulation Rotation time
Somatom Sensation 16×1.5 mm 120 kVp 120 mAs 0.5 s
Somatom Definition 2×32×0.6 mm 120 kVp 120 mAs CareDose4D 0.33 s
Slice thickness Reconstruction increment CTDIVOLa DLP a
2 mm 1 mm 7.4±0.3 mGy 1015±48 mGycm
1 mm 0.7 mm 5.8±1.5 mGy 789±213 mGycm
Eur Radiol (2013) 23:2482–2491
Table 3. (Timing of image acquisition in relation to CM bolus: too early with restricted [ART2] or unrestricted but suboptimal [ART1] image quality owing to poor arterial enhancement; too late with restricted [VEN2] or unrestricted but suboptimal [VEN1] image quality owing to venous enhancement; optimal [OPT]; weak [AOC1] or missing [AOC2] arterial enhancement owing to acute arterial occlusion). A sudden truncation of enhancement with a lack of distal arterial opacification was interpreted as an acute occlusion, while gradual peripheral fading of enhancement was rated as overriding of the contrast bolus. A level region was considered diagnostic if the quality assessment was ART1, OPT or VEN1 and non-diagnostic otherwise. Severe vessel wall calcifications considered to compromise the diagnostic assessment of the level region were documented. For the purpose of the bolus timing assessment, all level regions rated AOC1, AOC2 or with compromising vessel wall calcifications were excluded from further analysis.
2485
Quantitative image assessment Quantitative image analysis was performed by one reader with 6 years’ experience in reading run-off CTA. The arterial enhancement of the abdominal and lower limb arteries was measured in 13 different anatomical positions in one leg per patient using a round region of interest (ROI) within the vessel lumen. The measurement was performed in the leg with the superior enhancement. The placement of each ROI along the z-axis was performed according to anatomical landmarks as defined in Table 4 and shown in Fig. 1. The size of each ROI was carefully adapted to cover the entire vessel lumen without including the arterial wall or plaques. If native arteries were occluded, measurements were performed within collateral vessels of >1 mm diameter, if present. The venous enhancement of the lower leg was measured at two anatomical positions per patient in the calf and foot as
Table 3 Classification of the synchronisation of the table feed and the contrast medium bolus [3] Category
Properties
ART2
Image acquisition in relation to CM bolus too early (“bolus overtaken”) Image properties: Peripheral fading of arterial enhancement with absent contrast enhancement in some parts of one artery. No venous enhancement.
ART1
OPT
VEN1
VEN2
AOC1
AOC2
Diagnostic confidence: Diagnosis of the arterial state and of potential findings is restricted for at least one artery owing to insufficient arterial enhancement. Image acquisition in relation to CM bolus early (“weak enhancement”) Image properties: Peripheral fading of arterial enhancement, but preserved albeit weak contrast enhancement of all arterial segments. No venous enhancement. Diagnostic confidence: Diagnosis of the arterial state and of potential findings is unrestricted but suboptimal owing to poor arterial enhancement in at least one artery. Optimal synchronisation of image acquisition in relation to CM bolus Image properties: Strong homogeneous enhancement throughout all arterial segments. No venous enhancement. Diagnostic confidence: Diagnosis of the arterial state and of potential findings is unrestricted in all segmental arteries. Image acquisition in relation to CM bolus late (“venous enhancement”) Image properties: Weak venous enhancement of superficial or deep veins. Differentiation of veins, arteries and background tissue is still possible. Homogeneous enhancement throughout the arteries. Diagnostic confidence: Diagnosis of the arterial state and of potential findings is unrestricted but suboptimal owing to the venous enhancement. Image acquisition in relation to CM bolus too late (“venous contamination”) Image properties: Strong venous enhancement of superficial or deep veins. Differentiation of veins, arteries and background tissue is restricted by venous contamination and poor arterial enhancement. Diagnostic confidence: Diagnosis of the arterial state and of potential findings is restricted or insufficient for at least one artery owing to venous contamination. Weak arterial enhancement due to acute arterial occlusion Image properties: Arteries distal to an acute occlusion show poor arterial enhancement. Diagnostic confidence: Diagnosis of the arterial state and of potential findings is restricted but some distal arteries remain assessable. Missing arterial enhancement due to acute arterial occlusion Image properties: Arteries distal to an acute occlusion show no arterial enhancement. Diagnostic confidence: Diagnosis of the arterial state and of potential findings is impossible owing to an acute arterial occlusion.
2486
Eur Radiol (2013) 23:2482–2491
Table 4 Definition of vessel segments and anatomical position for quantitative assessment of intravascular contrast density values (adopted from [3])
Arteries
Level
Vessel segment
Abdomen
Proximal abdominal aorta Upper abdominal aorta Distal aorta Common iliac artery External iliac artery Superficial femoral artery 1 3 Superficial femoral artery 2 3 Superficial femoral artery 3 3
AA1 AA2 AA3 CIA EIA SFA1
~2 cm below the diaphragm Renal arteries Aortic bifurcation Iliac bifurcation Crossing of the inguinal ligament Level of the ischial tuberosities
SFA2
At the middle of the superficial femoral artery
SFA3
At the level of the end of the adductor channel
Popliteal artery
PA CA1
At the level of the distal intercondylar region of the femur At the level below the tibiofibular trunk a
2
CA2
In the middle of the calf
3
CA3
At the level of the ankle jointa
FA CV3
At the metatarsal level At the calf level a
a
FV
At the metatarsal level
a
Pelvis Thigh
Calf
Calf artery artery) Calf artery Calf artery
Veins
Foot Calf
1 3
(anterior tibial, posterior tibial or fibular
3 3
Dorsalis pedis artery or Plantar artery Greater saphenous vein or any other visible superficial vein Superficial or deep foot vein
Foot 1
Anatomical position
a
upper third, 2 3 middle third, 3 3 lower third of the vessel a Measurement was performed in the vessel with the strongest enhancement 3
defined in Table 4. A mean vessel enhancement value was calculated for each position in each TF group. Statistical analysis Among the TF groups, the mean body weight, patient size, body mass index (BMI), maximum outer and luminal diameter of the AAA were compared using the Kruskal-Wallis test. The frequency of severe peripheral artery disease (defined as Fontaine Stage III and IV) among the TF groups was compared using the chi-squared test. The inter-observer variability of the two readers was calculated using Cohen’s kappa statistic. The proportion of patients with different ratings of the corresponding left and right vascular level region was calculated for each reader and analysed using the chi-squared test for trend to assess the frequency of relevant differences in transit times between the left and right leg. To test for a potential difference in the arterial and venous enhancement dependent on the TF group and the anatomical position, the Kruskal-Wallis test was applied. To test for a potential linear correlation between the size of the aneurysm and the arterial enhancement for each TF group for the distal calf and foot arteries, a regression analysis was performed. A P value<0.05 was considered statistically significant.
Results Among the four TF groups, the distribution of body weight, patient size, BMI, maximum outer and inner diameter and the frequency of severe PAD revealed no statistically significant differences (Table 1). In all patients, the preselected threshold was reached and triggered the image acquisition. All aneurysms were juxtarenal or infrarenal. No aneurysm extended above the first suprarenal measurement. Qualitative assessment One patient had an above the knee amputation. Thus, a total of 708 level regions were available for analysis. In two patients, an acute occlusion (Fig. 1) was observed. Therefore, six level regions localised distal to the occlusion showed no or only weak arterial enhancement (AOC1, five level regions; AOC2, one level region) and were excluded from further analysis. In one patient, severe vessel calcifications led to a non-diagnostic image quality of the calf and foot arteries. Thus, bolus timing was evaluated in 698 level regions (Fig. 2). The inter-observer agreement of the two readers was very good (κ=0.82). The timing at the aortic level region was
Eur Radiol (2013) 23:2482–2491
considered optimal (OPT) in all patients by both readers. At the pelvic and thigh level, only a small proportion of level regions showed suboptimal bolus timing. Weak arterial enhancement (ART1) was noted in 5 % of the pelvic level regions in the TF3 (2/40, both readers) and TF4 (2/40, both readers) groups. At the thigh level, best timing was observed in the TF1 and in the TF2 group (TF 1, 38/39=97 %; TF 2, 40/42= 95 %; both readers). The largest proportion of arterial level regions with weak arterial enhancement were observed in the TF4 group (reader 1, 9/35=26 %; reader 2, 6/35=17 %). In one patient of the TF3 group, missing arterial enhancement of both thigh level regions (ART2) was present (2/40= 5 % level regions, both readers). At the calf and foot level, the number of level regions with optimal bolus timing decreased and was lowest at the foot level. At the calf level, the highest portion of level regions with non-optimal bolus timing was noted for the TF4 group. In this group, all of the level regions showed arterial fading (ART1: reader 1, 6/35=17 %; reader 2, 4/35=11 %; ART2: reader 1, 14/35=36 %; reader 2, 11/35=31 %). The lowest portion of level region non-optimal bolus timing (ART1: reader 1, 2/39=5 %; reader 2, 1/39=3 %; ART2: reader 1, 1/39=3 %; reader 2, 1/39=3 %) was observed in the TF1 group. Conversely, venous enhancement and venous contamination were not observed in the TF4 group and the proportion of segments with venous enhancement and venous contamination was highest in the TF1 group (VEN1: reader 1, 7/39=18 %; reader 2, 4/39=10 %; VEN2: 0 %, both readers). At the foot level, the main cause of non-diagnostic image quality was overriding of the contrast bolus in all TF groups. The portion of level regions with non-optimal bolus timing and overriding of the contrast bolus was highest in the TF4 group (ART1: reader 1, 7/35=20 %; reader 2, 8/35=23 %; ART2: reader 1, 18/35=51 %; reader 2, 15/35=43 %) and lowest in the TF1 group (ART1: 0 %; ART2: 3/39=8 %, both readers). The number of level regions at the foot level with compromising venous enhancement (VEN2) was comparable for the TF1 (2/39=5 %, both readers), TF2 (2/41= 5 %, both readers) and TF3 (2/39= 5 %, both readers) groups. The qualitative assessment of differences in the rating of the bolus timing between the right and left leg revealed the lowest differences at the pelvic (0/79, both readers) and thigh (reader 1, 3/78=4 %; reader 2, 4/78=5 %) levels. These intra-individual differences significantly increased from the pelvis to the calf and foot level region (reader 1: calf, 14/77=18 %; foot, 26/77=34 %, P<0.001; reader 2: calf, 17/77=22 %; foot, 16/77=21 %, P<0.001). In terms of diagnostic confidence, all level regions at the aortic and pelvic level were diagnostic. At the thigh level region, in only one patient in the TF3 group presenting with
2487
a huge aneurysm (outer diameter 89 mm) was no arterial enhancement observed in both legs. All other level regions were diagnostic in the thigh level region. Below the knee, the biggest portion of diagnostic level regions at the calf and foot level was observed in the TF1 group (calf: 38/39=97 % foot: 34/39=87 %, both readers) and the lowest number of diagnostic level regions were observed in the TF4 group (calf: reader 1, 21/35=60 %; reader 2:24/35=69 %; foot: reader 1, 17/35=49 %; reader 2, 20/35=57 %), while the number of diagnostic level regions in the TF2 (calf: reader 1, 35/41=85 %; reader 2, 36/41= 88 %; foot: 31/41=76 %, both reader) and TF3 (calf: reader 1, 33/39=85 %; reader 2, 36/39=92 %; foot: reader 1, 30/39 =77 %; reader 2, 31/39=79 %) groups was similar. Quantitative assessment Overall, a total of 1,012 arterial and 158 venous enhancement measurements were obtained. In one patient, measurements of the calf and foot arteries were not assessable owing to severe vessel wall calcifications, and in a second patient an acute aortic occlusion was present in both legs leading to non-assessable thigh, calf and foot arteries. From the proximal aorta (AA1) to the external iliac artery (EIA), a nearly equivalent mean arterial enhancement was observed in the TF1–TF3 groups (Fig. 3). In the TF4 group, the mean arterial enhancement slightly decreased in the infrarenal abdominal aortic aneurysm between the renal arteries (AA2, 371±57 HU) and the common iliac artery (CIA, 324±82 HU). The arterial enhancement below the abdominal aorta (AA3) of the TF4 group was inferior for each anatomical position compared with the TF1–TF3 groups. Between the external iliac artery and the distal superficial femoral artery (SFA3), the arterial enhancement increased and revealed a maximum enhancement at the SFA3 for all TF groups. Comparing the maximum arterial enhancement between the different TF groups, the strongest enhancement was obtained in the TF2 (437±75 HU) and TF3 (422±132 HU) groups followed by the TF1 (405±96 HU) and TF4 (361±114 HU) groups. In the further run-off, the arterial enhancement decreased in every TF group. At the distal calf (CA3), the strongest arterial enhancement was observed in the TF2 group, while the enhancement was significantly lower for the TF1, TF3 and TF4 groups (CA3, P<0.05). At the foot arteries (FA), the strongest arterial enhancement was observed in the TF1 group (228±83 HU) and decreased in the order TF2 (219±122 HU), TF3 (179±81 HU) to TF4 (126±90 HU, P<0.01). There was no linear correlation between the size of the aneurysms and the arterial enhancement observed in the subgroups for the calf and foot arteries (P>0.1; R2 <0.2). The venous enhancement at the foot (FV, P=0.78) and the ankle joint (CV3, P=0.84) was weak in all four TF
2488
Eur Radiol (2013) 23:2482–2491
Fig. 1 Peripheral enhancement patterns of the lower limb. Frontal maximum intensity projections (MIP) of the lower limb after bone subtraction of four different patients who have undergone run-off CTA studies. a A 76-year-old man with known AAA (max. outer diameter 57 mm). Strong homogeneous enhancement throughout all arterial segments from aorta and the pelvic arteries down to the toes without venous enhancement (TF1 group); white bars indicate the anatomical positions of the arterial and venous enhancement measurements (see Table 4). b A 61-year-old man with known abdominal aortic aneurysm (AAA; maximum outer diameter 41 mm, *) and artefacts from a knee joint replacement (**) on the right side with compromising venous enhancement at the
calf and foot due to too slow acquisition compared with the contrast enhancement (***). Note: the quantitative assessment was performed in the left leg. c An 80-year-old man with an AAA (maximum outer diameter 76 mm, empty white arrowhead) and insufficient peripheral arterial enhancement due to overriding of the bolus (empty white arrow). d An 85-year-old woman with previously unknown AAA (maximum outer diameter 59 mm, white arrowhead) and multiple subacute arterial thrombo-embolic occlusions (white arrow) in the superficial femoral artery and the distal popliteal artery. The arterial filling of the left lower limb is consecutively delayed. Note: the left leg was excluded from qualitative assessment
groups (mean: 74±67 HU, range 38–314 HU). However, the venous enhancement at the ankle joint (CV3) was slightly higher for the three TF groups TF1, TF2 and TF3 compared with TF4 (Fig. 3).
that a peripheral run-off CTA can be readily added to the abdominal CTA using a robust constant acquisition protocol with lower TFs as suggested for patients with peripheral artery disease only (40–48 mm/s) [3, 14]. The total amount of contrast material remains unchanged and the additional radiation exposure is small due to low conversion factors and low tube currents when using dose modulation. Therefore, the depiction of the peripheral arteries represents a reasonable extension of the examination according to the guideline recommendations in these patients [5, 13]. Due to the concept of effective tube current as used in this study, no dose variations are to be expected by changing the TF compared with alternative approaches [15].
Discussion In patients with AAA undergoing pretherapeutic CTA [13], it is highly desirable to depict not only the abdominal aorta but also to evaluate the whole vascular tree down to the feet to detect possible thrombo-embolic or atherosclerotic peripheral disease that may have a decisive impact on the therapeutic approach. In this manuscript, we have shown
Eur Radiol (2013) 23:2482–2491
Fig. 2 Rating of the synchronisation of image acquisition in relation to the contrast material bolus for both readers: too early with restricted (ART2) or suboptimal (ART1) image quality due to poor arterial enhancement, too late with restricted (VEN2) or suboptimal (VEN1)
2489
image quality due to venous enhancement, optimal (OPT). TF table feed group, TF1 (30–35 mm/s); TF2 (35–40 mm/s); TF3 (40–48 mm/ s); TF4 (48–86 mm/s)
Fig. 3 Arterial (a) and venous (b) vessel enhancement (mean±95 % confidence interval) dependent on the anatomical level defined in Table 4 and the table feed group (TF). * ≤0.05, ** <0.01
2490
Our patient collective is representative for patients with abdominal aortic aneurysms as shown by the male: female ratio of 6.9:1 and the distribution of the PAD stages, with mainly mild PAD [11, 16]. Peripheral artery disease with stenosis might decrease the peripheral arterial blood flow. Some patients had ulcers which might increase the venous flow at the foot as given by PAD stage 4. However, these patients were equally distributed between the four groups. The acquisition protocol used in this study was based on the following principles: a high iodine flux (approximately 1.6 g I/ml/s) in combination with automatic bolus tracking using a high threshold (Δ250 HU) in the aorta, followed by a short imaging delay of 5 s, was chosen to maximise arterial enhancement. Compared with a triggering at the popliteal level [4], this ensures a strong opacification of the whole vascular tree and allows a relatively low total CM volume of 100 ml to be used [1]. The whole run-off CTA dataset is acquired during the arterial first pass and venous contamination, and recirculation of the CM is largely avoided. Therefore, biphasic or even more complex injection protocols generating a prolonged bolus are unnecessary using our approach. Furthermore, if the high threshold in the aorta is not reached (which was not the case in this study) image acquisition can still be started manually. The initial delay and the bolus duration of 25 s were sufficient to compensate for differences between the TF and bolus transit time. Our results demonstrate that a high number of diagnostic examinations can be achieved in patients with AAA undergoing run-off CTA using a simple and robust standardised acquisition protocol with no individual adaptation of the TF. A cranio-caudal acquisition from the suprarenal aorta down to the feet, combined with a fixed TF between 30 and 35 mm/s, results in the best arterial enhancement profile with the lowest number of non-diagnostic segments owing to insufficient arterial enhancement or compromising venous filling. Reliable imaging of the aortic, pelvic and thigh arteries was obtained in all patients using this TF protocol. However, the visualisation of the calf and foot arteries remains the most challenging region for run-off CTA, as 3 % of the calf and 13 % of the foot artery levels were nondiagnostic. These results for patients with AAA are very similar to the published results for patients with PAD only. Siriapisith et al. [6] observed adequate arterial opacification in 97 % of the calf arteries and 11 % of the foot segments in patients with PAD using a constant acquisition protocol. Meyer et al. [3] obtained diagnostic visualisation of the foot arteries in 90 % of PAD patients. As the proportion of non-diagnostic arterial levels due to compromising venous enhancement or insufficient arterial contrast was well balanced in the foot, this can probably be attributed to the individual variation in bolus transit time and does not reflect a systematic shortcoming of the acquisition
Eur Radiol (2013) 23:2482–2491
protocol. A slower TF below 30 mm/s would increase the risk of strong venous enhancement of calf and foot arteries, as demonstrated in Fig. 2. On the contrary, a faster TF above 35 mm/s considerably increased the number of patients in whom an outpacing of the contrast material bolus in the distal arteries was observed. As suggested in the literature, a more reliable vessel enhancement of the calf and foot arteries might be achieved using test bolus techniques and individual TFs [3, 7, 9]. Interestingly, the results for the qualitative assessment of the calf and foot arteries in two published studies using a test bolus technique reveal a diagnostic depiction of the arteries without any compromising venous enhancement in all cases [4, 9]. Taking into account that in our study differences in bolus transit times between the two legs were observed in up to 22 % of the calf and 33 % of the foot level, this reported perfect rate of diagnostic segments was not achieved in our study, though the flow rate and the total volume of contrast material were equal to or lower than in our study. However, a direct comparison of the results is impossible, as the effective TF used in these studies is not given and a comprehensive bolus timing assessment was not carried out in either study. In cases with differences in bolus transit times between the two legs, an additional delayed CT acquisition of the calf and foot arteries might be considered to improve the vessel delineation [17, 18]. This additional CT acquisition should be planned in advance, so that it can immediately be initiated at the console in case of visually impaired vessel opacification. Some authors even suggest timeresolved imaging techniques to improve the visualisation of severely calcified arterial segments [18–20]. Other approaches that aim at increasing the overall acquisition time (e.g. measurement of enhancement in the aneurysm sac or increasing the delay after reaching peak enhancement in the aorta) do not take into account the overall reduced flow caused by the aneurysm sac as demonstrated in this study. In extreme cases the aortic enhancement above the aneurysm sac might even decrease using these approaches. While the high accuracy of run-off CTA compared with the “gold standard” of digital subtraction angiography (DSA) with regard to the detection and quantification of steno-occlusive disease has been proven [1, 21], there are only a few published studies investigating the acquisition technique and the technical reliability of the applied imaging protocols [3, 4, 6, 7, 9]. In preceding studies by Albrecht et al. [14] and Meyer et al. [3], a high reliability of a constant acquisition protocol using a standardised TF of 40–48 mm/s in a larger population of patients with PAD could be proven. In the study by Meyer et al., the trend of outpacing the contrast bolus in a small subgroup of 13 patients with AAA was observed if a high TF was used [3]. This initial result was confirmed in the present study in a larger patient population.
Eur Radiol (2013) 23:2482–2491
A number of limitations of this study need to be acknowledged. Firstly, the study is limited by its retrospective nature and the different TFs used. However, to the best of our knowledge, this is the first study to evaluate a larger, representative patient collective [16] with AAA with respect to the influence of the TF on the quality of run-off CTA. Secondly, there was no comparison with the DSA gold standard, as evaluation of diagnostic accuracy was not the purpose of the study. In order to overcome the limitation of a subjective qualitative assessment of the arterial enhancement, additional quantitative measurements were carried out. Finally, our results apply only to the short monophasic contrast medium bolus used in this study; other (e. g. biphasic) protocols may alter the results. In conclusion, a constant TF ranging between 30 and 35 mm/s was demonstrated to be the best trade-off between the risk of outrunning the bolus of contrast agent and compromising venous enhancement for run-off CTA in patients with an AAA. Acknowledgements This study was funded by the German Federal Ministry for Education and Research (Bundesministerium für Bildung und Forschung) through the grant 02NUK008D (www.helmholtzmuenchen.de/kvsf/).
References 1. Met R, Bipat S, Legemate DA, Reekers JA, Koelemay MJW (2009) Diagnostic performance of computed tomography angiography in peripheral arterial disease: a systematic review and metaanalysis. JAMA 301:415–424 2. Fleischmann D, Rubin GD (2005) Quantification of intravenously administered contrast medium transit through the peripheral arteries: implications for CT angiography. Radiology 236:1076–1082 3. Meyer BC, Oldenburg A, Frericks BB et al (2008) Quantitative and qualitative evaluation of the influence of different table feeds on visualization of peripheral arteries in CT angiography of aortoiliac and lower extremity arteries. Eur Radiol 18:1546–1555 4. Nakaya Y, Kim T, Hori M et al (2009) 64-slice multidetector row computed tomographic angiography of aortoiliac and lower extremity arteries: efficacy of test injection using a monitoring scan at knee level. J Comput Assist Tomogr 33:20–25 5. Shimizu K, Utsunomiya D, Nakaura T et al (2010) Uniform vascular enhancement of lower-extremity artery on CT angiography using test-injection monitoring at the central level of the scan range: a simulation flow phantom study with clinical correlation. Acad Radiol 17:1153–1157
2491 6. Siriapisith T, Wasinrat J, Mutirangura P, Ruangsetakit C, Wongwanit C (2010) Optimization of the table speed of lower extremity CT angiography protocols in different patient age groups. J Cardiovasc Comput Tomogr 4:173–183 7. Nakaya Y, Kim T, Hori M et al (2011) Correlations between aortopopliteal bolus transit speed and aortic and popliteal bolus transit time during CT angiography of aortoiliac and lower extremity arteries. Eur J Radiol 79:272–276 8. Biasetti J, Gasser T, Auer M, Hedin U, Labruto F (2010) Hemodynamics of the normal aorta compared to fusiform and saccular abdominal aortic aneurysms with emphasis on a potential thrombus formation mechanism. Ann Biomed Eng 38:380–390 9. Laswed T, Rizzo E, Guntern D et al (2008) Assessment of occlusive arterial disease of abdominal aorta and lower extremities arteries: value of multidetector CT angiography using an adaptive acquisition method. Eur Radiol 18:263–272 10. Suh G-Y, Les A, Tenforde A et al (2011) Quantification of particle residence time in abdominal aortic aneurysms using magnetic resonance imaging and computational fluid dynamics. Ann Biomed Eng 39:864–883 11. Metcalfe D, Holt PJE, Thompson MM (2011) The management of abdominal aortic aneurysms. BMJ 342:d1384 12. Albrecht T, Foert E, Holtkamp R et al (2007) 16-MDCT angiography of aortoiliac and lower extremity arteries: comparison with digital subtraction angiography. AJR Am J Roentgenol 189:702– 711 13. Pua U, Tan K (2012) Radiology of advanced EVAR techniques in complex abdominal aortic aneurysms. Eur Radiol 22:387–397 14. Albrecht T, Meyer B (2007) MDCT angiography of peripheral arteries: technical considerations and impact on patient management. Eur Radiol Suppl 17:5–15 15. McCollough CH, Bruesewitz MR, Kofler JM (2006) CT dose reduction and dose management tools: overview of available options. Radiographics 26:503–512 16. Eckstein HH, Bockler D, Flessenkamper I, Schmitz-Rixen T, Debus S, Lang W (2009) Ultrasonographic screening for the detection of abdominal aortic aneurysms. Dtsch Arztebl Int 106: 657–663 17. Fotiadis N, Kyriakides C, Bent C, Vorvolakos T, Matson M (2011) 64-section CT angiography in patients with critical limb ischaemia and severe claudication: Comparison with digital subtractive angiography. Clin Radiol 66:945–952 18. Sommer W, Helck A, Bamberg F et al (2010) Diagnostic value of time-resolved CT angiography for the lower leg. Eur Radiol 20: 2876–2881 19. Sommer WH, Becker CR, Haack M et al (2012) Time-resolved CT angiography for the detection and classification of endoleaks. Radiology 263:917–926 20. Iezzi R, Santoro M, Marano R et al (2012) Low-dose multidetector CT angiography in the evaluation of infrarenal aorta and peripheral arterial occlusive disease. Radiology 263:287–298 21. Napoli A, Anzidei M, Zaccagna F et al (2011) Peripheral arterial occlusive disease: diagnostic performance and effect on therapeutic management of 64-section CT angiography. Radiology 261: 976–986