Pediatric Radiology

, Volume 37, Issue 10, pp 998–1006 | Cite as

Visualization of coronary arteries in patients after childhood Kawasaki syndrome: value of multidetector CT and MR imaging in comparison to conventional coronary catheterization

  • Raoul Arnold
  • Sebastian Ley
  • Julia Ley-Zaporozhan
  • Joachim Eichhorn
  • Jens-Peter Schenk
  • Herbert Ulmer
  • Hans-Ulrich Kauczor
Original Article

Abstract

Background

After childhood Kawasaki syndrome (KS) the coronary arteries undergo a lifelong dynamic pathological change, and follow-up coronary artery imaging is essential. At present, conventional coronary catheterization (CCC) and angiography is still regarded as the gold standard. Less-invasive methods such as multidetector CT angiography (MDCT-A) and MRI have been used sporadically.

Objective

To compare the diagnostic quality of MDCT-A and MRI with that of CCC for coronary imaging in a group of patients with coronary artery pathology after childhood KS.

Materials and methods

A total of 16 patients (aged 5–27 years) underwent CCC and 16-row MDCT-A and 14 patients MRI (1.5 T).

Results

There was 100% agreement between MDCT-A and CCC in the detection of coronary aneurysms and stenoses. MDCT-A was superior for the visualization of calcified lesions. MRI and CCC showed 93% agreement for the detection of aneurysms. Visualization of coronary artery stenoses was difficult using MRI—one stenosis was missed.

Conclusion

MDCT-A has excellent correlation with CCC regarding all changes affecting the coronary arteries in the follow-up of childhood KS. In comparison to MDCT-A and CCC, MRI is less precise in the detection of stenotic lesions. Due to its high image quality and ease of performance MDCT-A should be the primary diagnostic modality in patients following childhood KS.

Keywords

Kawasaki syndrome Multidetector CT MRA Children 

Introduction

Kawasaki syndrome (KS) is an acute febrile vasculitis of unknown origin that was first described in 1967 [1]. During the acute phase 10–25% of patients develop coronary ectasia or coronary artery aneurysms (CAA) [2]. Most CAA are small to medium in size. The so-called “giant” aneurysms exceed 8 mm in diameter. In the long term these pathological changes may lead to thrombotic vessel occlusion, coronary stenosis or premature arteriosclerosis [3]. Patients with a significant reduction in coronary perfusion need ongoing medical treatment. In some patients even percutaneous revascularization or surgical procedures are required [3]. In order to detect pathological changes in coronary arteries serial radiographic conventional coronary angiography (CCC) is recommended in the follow-up of patients with KS. At present, CCC is still the gold standard for coronary imaging.

Over recent years the image acquisition speed and quality of cardiac multidetector CT angiography (MDCT-A) have improved [4]. The first studies in KS patients showed good correlations with CCC [5]. In order to reduce or even avoid radiation exposure, MR techniques have been introduced and developed for coronary imaging [6]. However, spatial resolution and sensitivity are lower than in MDCT-A or CCC. In KS, MDCT-A and MRI have occasionally been used in adult patients [5, 7, 8, 9, 10]. So far, no study comparing imaging modalities in juvenile patients has been reported. Therefore, the goal of our study was to evaluate the clinical usefulness of MDCT-A and MRI in children and juveniles after KS for the assessment of coronary vascular pathology and ventricular function. The two methods were compared with CCC as the diagnostic gold standard.

Materials and methods

Patient population

The study population was recruited from about 60 KS patients who were regularly seen as outpatients at the Paediatric Cardiology Department of the University Hospital Heidelberg. One patient was referred from an external hospital. Overall 16 consecutive paediatric patients treated for KS between 1988 and 2002 were enrolled in this prospective study (nine male, seven female). Patients were included in the study during their regular scheduled follow-up examinations (basic examination included ECG and echocardiography). For study inclusion, an aneurysm had to be present at the acute stage of the disease and at follow-up (Table 1).
Table 1

Individual patient findings of each modality in the diagnosis of KS: CCC, MDCT-A and MRI. The values are the absolute numbers of findings for each modality

Patient no.

Gender

Age (years)

Aneurysms

Ectasia

Stenoses

Calcificationa

CCC

CTA

MRI

CCC

CTA

MRI

CCC

CTA

MRI

CCC

MDCT-A

1

M

12

2

2

2

0

0

0

0

0

0

0

1

2

F

12

0

0

0

0

0

0

0

0

0

0

0

3

M

12

2

2

2

2

1

1

0

0

0

0

0

4

M

14

0

0

0

1

1

1

0

0

0

0

0

5

M

25

2

2

2

0

0

0

0

0

0

1

2

6

M

10

0

0

Failed

2

2

Failed

0

0

Failed

0

0

7b

M

5

3

3

3

0

0

0

0

0

0

0

2

8

M

18

2

2

2

3

3

1

0

0

0

1

1

9

F

11

0

0

Refused

0

0

Refused

1

1

Refused

0

0

10

F

21

1

2

1

0

0

0

0

0

0

0

0

11

F

11

0

0

0

1

1

1

0

0

0

0

0

12

M

18

2

2

2

1

1

0

1

1

1

1

2

13

M

22

2

2

2

0

0

0

0

0

0

2

2

14

M

15

5

5

4

0

0

0

0

0

0

2

3

15

F

27

4

4

3

0

0

0

3

3

2

2

2

16

F

18

5

5

5

0

0

0

0

0

0

0

0

aCalcification can only be assessed by CCC and MDCT-A.

bPatient 7 was sedated for the MRI examination.

During their visits, patients underwent MDCT-A and MRI. MDCT-A was performed instead of CCC as the recent literature indicates good agreement between CCC and MDCT-A of the coronary arteries. In one patient a close follow-up CCC was necessary. The study was carried out according to the institutional ethical guidelines and in conformity with the Declaration of Helsinki. The study was approved by the local ethics committee and all patients/parents gave informed consent.

For MDCT-A and MRI no β-blockers or sedatives were given, with the exception of a 5-year-old boy who received mild sedation for MRI. A 20/22-G needle was placed in an antecubital or more peripheral vein. On the day of examination the patients’ demographics were: mean age 16 ± 6 years (median 14.5 years; range 5–27 years), mean height 161 ± 23 cm (range 110–195 cm), and mean weight 53 ± 20 kg (range 20–84 kg). The follow-up interval, calculated between documented onset of acute KS and MDCT-A/MRI, was 15 ± 5 years. The mean time between CCC and MDCT-A/MR was 3 ± 3 years.

CCC

CCC procedures were performed using an Integris BH3000 machine (Philips, Best, The Netherlands) for biplane angiography. Diagnostic quality was achieved in all cases. Digitally stored data of the initial and the most recent follow-up CCC were reviewed by one reader (H.U.) blinded to the MDCT-A and MRI findings. Quantification of vessel diameter was done in both planes and expressed as a mean value.

MDCT-A

MDCT was performed on a 16-row-detector scanner (Aquilion 16, Toshiba Medical Systems, Tokyo, Japan). For cardiac gating the integrated scanner ECG gating unit was used. Retrospective ECG gating mode was chosen. The scanner did not allow ECG pulsing or ECG-based tube current modulation. Mean heart rate was 79 ± 19 beats/min (range 61–125 beats/min). Contrast medium was injected automatically (Injektron CT2, Medtron, Saarbrücken, Germany). CT parameters were (no dose modulation): 120 kVp, 17–69 mAs, gantry rotation time 0.4 s (n = 12) or 0.5 s (n = 4), beam pitch 4 (n = 5) or 3.2 (n = 11). The mean length of the scan was 105 mm. The reconstructed field-of-view (FOV) was adjusted to the individual’s heart size. The iodine concentration used was 300 mg/dl (n = 14) or 400 mg/dl in (n = 2) depending on the patient’s age and needle size (mean amount of contrast medium 1.6 ± 0.6 ml/kg body weight, 485 ± 172 mg iodine/kg body weight). A 20-ml saline flush was injected after injection of contrast medium. The injection rate for contrast medium and saline was 2 ml/s (n = 4) or 3 ml/s (n = 12).

Image acquisition was started manually when the contrast medium arrived in the ascending aorta. Patients were instructed to hold their breath in inspiration. Slice thickness was adjusted to the time needed to cover the whole heart balanced by the ability of the patient to suspend respiration. Thus, 0.5 mm (n = 10) or 1 mm (n = 6) slices were acquired. Reconstruction was done with 20% image overlap. Gated CT datasets were reconstructed every 10% of the cardiac cycle and all ten resulting 3-D datasets were transferred to a dedicated workstation (Vitrea v.3.5, Vital Images, Plymouth, Minn.).

MRI

Studies were performed on a commercial 1.5-T whole-body MR system (Magnetom Symphony or Magnetom Avanto, Siemens Medical Systems, Erlangen, Germany) equipped with cardiac software. The MR protocol consisted of adequate localizers and navigator-gated, noncontrast bright-blood angiography of the coronary arteries (true FISP 3-D, TR 243.6 ms, TE 2.02 ms, flip angle 65°, matrix 384 × 216, fat saturation, resolution 1.2 × 0.9 × 1.5 mm3). Each main coronary artery was planned separately using the three-point planning tool. After this, contrast medium (0.1 mmol/kg body weight of Gd-DTPA) was injected. For myocardial functional imaging and detection of scars, 2-D cine sequences (true FISP 2-D, ECG-gated, TE 1.82 ms, TR 29.12 ms, flip angle 65°, slice thickness 6 mm, matrix 208 × 256) were obtained in conventional fashion. After 15–20 min, late enhancement studies of the myocardium were performed (turbo-flash 2-D, TR 600 ms, TE 5.4 ms, slice thickness 8 mm, TE individually determined, flip angle 30°, matrix 192 × 256).

Image analysis

MDCT-A and MRI were reviewed in consensus by three readers (one cardiologist, R.A.; two radiologists, S.L. and J.L.-Z.). Evaluation was done in random order between MDCT-A and MRI in different patients, and names were not displayed on the monitor. Image analysis comprised localization and quantification of coronary arterial pathologies. The main left and right coronary artery diameters as well as the diameters before and after a vascular pathology were measured. The length and diameter of the aneurysms were measured. The same information was extracted from CCC. Ectasia was defined as noncircumferential enlargement of the vessel diameter to 3–4 mm or an internal diameter of a segment measuring more than 1.5 times that of an adjacent segment. An aneurysm was defined according to the American Heart Association guidelines, i.e. circumferential enlargement of 5 mm or more. Stenosis was defined as a vessel diameter <50% of the proximal segment diameter [3].

Myocardial wall motion analysis was performed on CCC and MRI. Regional motion was classified as normal or abnormal. Late enhancement was assessed on MRI and noted if present. Calcifications were noted on CCC and MDCT-A.

Radiation dose calculation

To allow a rough comparison of radiation dose between CCC and MDCT-A, the modality-specific radiation dose was converted to effective dose (mSv). This was done for CCC using the commercial software Refdose v1.00.04 (IBA Scanditronix Wellhoefer, Schwarzenbruck, Germany) and for MDCT-A using CT-Expo v1.4 (G. Stamm, Hannover, Germany) [11]. Radiation dose conversion calculations can be severely affected by estimation errors. For CT-Expo, for example, the authors quote an error range of ±20–30%. As the CCC data were collected retrospectively, the individual field size and voltage had to be estimated. Therefore, only mean values are given to allow an approximate comparison between the two modalities.

Statistical analysis

Measurements from MDCT-A and MRI were compared with those from CCC as the gold standard using the Pearson r coefficient.

Results

Representative CCC, MDCT-A and MRA images are shown in Figs. 1 and 2.
Fig. 1

Patient 5, male, 25 years of age. Representative images: a CCC, b MDCT-A, c MR. The disease occurred at the age of 2 years and resulted in one giant aneurysm of the right coronary artery and one fusiform aneurysm in the left coronary artery. a CCC of the right coronary artery shows a giant aneurysm arising at the ostium of the vessel. The distal part of the vessel does not show any pathological changes. b Curved multiplanar reconstruction of the MDCT-A shows both coronary arteries: a giant aneurysm arises just beyond the ostium of the right coronary artery and there is a small calcified aneurysm of the left anterior descending artery. The right coronary artery aneurysm is 25 mm long with a diameter of 12.5 mm. c Maximum intensity projection of the MRA of the right coronary artery shows the giant aneurysm. Spatial resolution is high enough to visualize the small ostium of the right coronary artery proximal to the aneurysm. The distal parts of the vessel cannot be visualized in this particular image

Fig. 2

Patient 16, female, 18 years of age. a, b Curved multiplanar reconstruction reformatted MR images show a giant aneurysm in the proximal left anterior descending artery (LAD) and a small aneurysm more peripherally. The right coronary artery (RCA) shows two small aneurysms. The disease occurred at 4 years of age and at presentation there were two aneurysms of the LAD and three fusiform aneurysms at the RCA. LCA left coronary artery

CCC

All CCC datasets were of diagnostic quality and eligible for evaluation. The mean age of the patients was 12 ± 6 years. For the complete examination, 69 ± 35 ml of contrast medium was used. Solely for visualization of the coronary arteries 26 ± 22 ml of contrast medium was used. The mean duration of the examinations was 95 ± 34 min with a radiation dose of 26 ± 21 dGy/cm2. The calculated effective dose was 2 ± 1.7 mSv.

Altogether 30 aneurysms were seen in 23 coronary arteries, 10 ectasias in 10 coronary arteries and 5 stenoses in 5 coronary arteries (Table 2). In the 14 patients who underwent MRI, CCC depicted 30 CAA, 8 ectasias and 4 stenoses. The mean diameter of the aneurysms was 7.9 ± 4.2 mm and the mean length was 14.1 ± 11.5 mm. All aneurysms within the left coronary artery were located in the proximal segment, whereas distal aneurysms were found in the right coronary artery.
Table 2

Overall comparison of CCC, MDCT-A and MRI in the diagnosis of coronary pathology

 

Aneurysms

Ectasiaa

Stenosesb

Patients who underwent MDCT-A (n = 16)

CCC

30

10

5

MDCT-A

31 (100%)

9 (90%)

5 (100%)

Patients who underwent MRI (n = 14)c

CCC

30

8

4

MRI

28 (93%)

4 (50%)

3 (75%)

aEctasia was defined as a vessel diameter larger than 4 mm or if the internal diameter of a segment measured more than 1.5 times that of an adjacent segment.

bA stenosis was defined as a vessel diameter less than 50% of the proximal segment diameter.

cTwo patients did not undergo MRA. Therefore the rate of detection was adjusted to the group of 14 patients who underwent CCC and MRI.

Regional wall motion abnormalities were seen in two patients. Calcifications were seen in 9 of 30 aneurysms (30%).

MDCT-A

All MDCT-A datasets were of diagnostic quality and eligible for evaluation (no artefacts, no breathing artefacts). Time of breath-hold during acquisition of ECG-gated MDCT-A datasets was 24 ± 12 s. In the CTA studies total mAs was 6,608 ± 2,960 with a dose length product (DLP) of 1,517 ± 617 mGy/cm. The calculated effective dose was 12 ± 4.8 mSv.

For analysis, ten phases of the cardiac cycle were reviewed and the phase with the fewest motion artefacts was chosen. Phases were named according to the time from onset of the cardiac cycle. Phase 30% was chosen once, 40% four times, 50% once, 60% once, 70% twice, 80% five times and 90% twice.

MDCT-A demonstrated all CAAs (n = 30) seen on CCC. One further heavily calcified CAA was seen solely on MDCT-A. Coronary ectasia was correctly identified in nine of ten patients, and was missed in one. All coronary vessel stenoses detected by CCC were successfully identified by MDCT-A (n = 5; Table 2). The mean diameter of the aneurysms was 8.5 ± 4.3 mm, and their mean length was 16.3 ± 11.7 mm. The linear regression coefficient for CCC versus MDCT-A for aneurysm length was r = 0.93 and for the diameter r = 0.89 (Figs. 3 and 4). Calcifications were seen in 15 of 31 aneurysms (48%).
Fig. 3

Linear regression for measured diameters of aneurysms: a CCC vs. MDCT-A, b CCC vs. MRI, c MDCT-A vs. MRI

Fig. 4

Linear regression for measured lengths of aneurysms: a CCC vs. MDCT-A, b CCC vs. MRI, c MDCT-A vs. MRI

MRI

One patient refused MR examination. In one patient image acquisition of the coronary arteries was not successful (very irregular breathing resulted in a signal detection problem from the navigators). The mean investigation time was 65 ± 16 min.

Of 30 CAAs diagnosed by CCC, 28 (93%) were also diagnosed by MRA. Of four coronary vessel stenoses detected by CCC, three were successfully identified by MRI. Coronary ectasia was detected in eight patients by CCC and in four patients by MRI (Table 2). the mean diameter of the aneurysms was 8.6 ± 3.8 mm; the mean length was 17.3 ± 12.3 mm. Wall motion abnormality was present in one patient. Late enhancement was not observed. The one patient who had known myocardial infarction (from CCC) refused MRI.

MDCT-A vs. MRI

The linear regression coefficient for MDCT-A versus MRI for aneurysm length was r = 0.99 and for aneurysm diameter r = 0.91 (Figs. 3 and 4).

Discussion

This study compared the diagnostic quality of MDCT-A and MRI with that of CCC for the investigation of coronary arteries after childhood KS. MDCT-A was performed in all patients, while MRI was performed in 14 of 16. Diagnostic information regarding aneurysms was always obtained by both methods with high precision regarding length and width. Additionally, MDCT-A showed high sensitivity in the detection of subtle lesions such as coronary stenosis (100%) and ectasia (90%). MDCT-A showed a higher detection rate for calcifications than CCC.

Due to the lifelong dynamic process of the coronary arteries following childhood KS, coronary artery imaging is essential in the follow-up of patients with KS [3, 12]. Serial CCC is regarded as the gold standard [13]. The timing of follow-up examinations is related to the individual severity of the coronary pathology and ranges from 1 to 10 years [3]. Based on CCC findings, medical, interventional or surgical treatment is considered. During recent years, MDCT-A and MRI have improved substantially for the clinical work-up of coronary arteries [5, 7, 8, 9, 10]. In this study, this was very obvious for image acquisition with MDCT-A. MRA was less well tolerated by the young patients and image quality was less reliable. From the patient’s perspective, MDCT-A was preferred to MRI regardless of radiation exposure, because of shorter examination times and greater convenience of the procedure [13].

MDCT-A showed almost 100% agreement with CCC concerning the visualization of pathological changes in affected vessels, including imaging of stenotic coronary lesions. These findings are in accordance with those reported by Kanamaru et al. [5]. In their study, β-blockers were mandatory to reduce the heart rate. We found that using a modern 16-row CT scanner with a 400-ms rotation speed the use of β-blockers was not absolutely necessary to obtain high-quality data. Installed multisegmental reconstruction algorithms ensure optimal reconstruction timing [14]. The optimal image acquisition technique regarding gantry rotation time and use of segmentation are determined automatically by the scanner.

With appropriate coaching, breath-hold examinations can be performed in all patients down to 5–6 years of age. In a three-case study, Sohn et al. [7] reported similar results. Sedatives must not be used in MDCT-A examinations, as breathing artefacts will severely degrade image quality.

In MDCT-A, even very small vessel diameters and subtle irregularities in the lumen of the coronary artery can be visualized. This is extremely important as coronary stenoses are potentially life threatening and must not be missed in any of the KS patients. By using small scan fields (FOV 240 mm) the spatial resolution is 0.3 × 0.3 mm. For reliable detection of a stenosis a calibre change of one or two voxels is required. Thus, MDCT-A can detect stenoses in the range of 0.3–0.6 mm. Calcification of coronary arteries can also be well detected [15]. In our study, calcifications seen on MDCT-A were often missed by CCC. Unfortunately, large calcified aneurysms lead to considerable artefacts in MDCT-A. Therefore, interpretation of heavily calcified aneurysms in MDCT-A is challenging and sometimes impossible [5]. In one patient of our series we had a similar experience; nevertheless the diagnosis was correct.

In order to detect coronary pathology it was often necessary to read every single phase of the cardiac cycle. It would have saved time if the optimum time point for coronary interpretation had been known. In adults, usually only two different phases of the cardiac cycle need to be reconstructed. For the right coronary artery the best phase is around 50% and for the left coronary arteries the recommended phase is around 70–80% [16]. Thus, it was concluded that reconstruction of raw data should be performed between 40% and 80%, as there the cardiac motion is relatively low [17]. In our series, two phases (40% and 80%) were used in nine patients for evaluation of the coronary arteries. Looking at the other phases that were used, a clear trend to diastolic phases was observed. Taking together the phases 40%, 70%, 80% and 90%, altogether 13 patients could be evaluated. Interestingly, in our series there was no difference in the optimum phase between left and right coronary arteries.

MRA showed 93% correlation with CCC concerning CAA, while ectasia was found in 50% of the patients; one stenotic lesion was missed. These findings are in accordance with those of other groups [8, 10]. Concerning larger vessel diameters, such as is seen in CAA, MRA showed reliable measurements. Smaller vessel diameters, such as those seen in stenotic lesions and the distal part of coronary arteries, were not always detectable. This problem was not emphasized in previous reports [8, 9, 10].

The results of MRA depend largely on the experience of the investigator and the compliance of the patient. Procedure times up to 60 min may sometimes only be accomplished with the use of sedation, and still examinations can fail. This may be the reason why MRA is not yet widely used in the follow-up of patients after childhood KS. Furthermore, detection of coronary artery calcification is not possible or at least is very difficult by MRI. On the other hand, cardiac wall motion and myocardial viability can be evaluated during the examination. Lately, interpretation of wall motion has also been possible with CT datasets. ECG gating allows this as a side product of MDCT-A. Still the biggest advantage of MR is the possibility to assess myocardial viability.

We regard dose considerations of major importance, especially as the lifetime risk from radiation is higher in children than in adults. However, there are no proven facts about the amount of risk increase in a paediatric population. To date, there is no study dealing with radiation dose and gated cardiac CT in children. For the purpose of this discussion a few studies in adults are considered to get an idea of the dose range expected from MSCT-A and CCC. The results of our conversion calculation needed for comparison between MDCT-A and CCC can only be taken as approximate.

Frush and Yoshizumi [18] found that in adults the radiation dose range for gated cardiac CT is approximately 11 mSv, but possibly as high as 20–25 mSv. Looking at gated cardiac CT in children, the same authors mentioned that radiation doses could range from 7.0 mSv to >25 mSv (Caroline Hollingsworth, Durham, N.C., unpublished data). In the recent literature, the radiation doses for MDCT-A range from 5.3 mSv to 14 mSv depending on the scanner technology used [19, 20]. The lowest value was acquired using a 64-slice CT scanner facilitating dose modulation depending on the cardiac cycle. However, peak doses of 16.3 mSv were reached and dose modulation only works efficiently if the heart rate is low. In our study population the heart rates were as high as 125 bpm. Therefore, the administration of a β-blocker would have been mandatory. Using this technique we might have been able to reduce the CT energy and the effective dose (mean 12 mSv) even more. The information about the best cardiac phase for coronary imaging obtained from this study is additionally helpful. To reduce radiation dose, ECG-related dose modulation can be adjusted to the phase in which the coronary arteries are best visualized.

For CCC, clinical investigations have found meaningful dose ranges for paediatric procedures. Values vary from about 5.0 mSv to more than 20 mSv. One recent report gives an approximate dose of 6 mSv for diagnostic CCC [19]. Our values (mean 2 mSv) were obviously lower, which can be mainly explained by the experience of our cardiologists with KS patients.

We make one final comment on the absolute risk of radiation. In adults, one study compared the combined radiogenic and nonradiogenic risk of mortality from MDCT-A (0.07%) and diagnostic CCC (0.13%) [19]. As children are considerably more sensitive to the carcinogenic effects of ionizing radiation than adults, and have a longer life expectancy in which to express risk, these estimates cannot be applied to them. Several cohort studies of girls and young women exposed to multiple diagnostic radiation doses have been informative about increased mortality from breast cancer with increasing radiation dose, and case-control studies of childhood leukaemia and postnatal diagnostic radiation exposure have suggested increased risks with an increasing number of examinations [21]. It is obvious that the indication for coronary imaging should be considered carefully and, especially in female patients, the number of investigations should be reduced to a minimum.

One major limitation of this study was the long time interval between CCC and MDCT-A/MRI in some patients. However, no patient with a long time interval showed progression of clinical symptoms. Therefore, in those patients no additional CCC was performed. In one patient with marked discrepancies between the old CCC report and actual cross-sectional imaging, a close follow-up CCC was carried out.

Conclusion and recommendation for follow-up examinations

Shrinking of aneurysms, proliferation of the intima and vessel calcification may lead to the development of coronary stenosis in KS patients. Aneurysms show a lifelong dynamic disease process and patients are in danger of suffering from a potentially life-threatening event such as myocardial infarction. For follow-up examinations it is of utmost importance to visualize stenotic segments of the coronary arteries because they lead to impaired myocardial perfusion. Compared to CCC, this can be done less invasively by MDCT-A in adults. If the latest CT scanner technologies are available (allowing ECG-dose modulation), MDCT-A can be recommended for follow-up examinations. In all other settings, CCC should still be recommended as the first-line examination in juveniles. In younger children especially, new MRI techniques are very helpful because calcification does not play a role in the early stage of the disease. Higher spatial resolution allows detection of stenosis, but still there is concern that a potentially life-threatening pathology may be missed. Therefore, the use of MRI must also be considered. Considerable experience, both in MRI technique and care of children with KS, is necessary to provide reliable investigations and interpretation of the findings.

Notes

Acknowledgements

The authors acknowledge Toshiba and Siemens for providing imaging assistance and special on-site training. Patient I.E. was referred by Dr. M. Freund (UMC Utrecht, The Netherlands) and Dr. O. Krogmann (Duisburg Paediatric Heart Centre, Germany). Dr. Freund also kindly provided recent coronary angiograms of the patient. The MRI examinations were performed by Mrs. S. Yubai (RT) and Mrs. K. Knauer (RT) with enthusiasm and excellent patient care. The authors are very grateful to Dipl. Ing Mews (Toshiba) and Dipl. Ing. Knoch (University Heidelberg) for calculation of the effective radiation doses.

References

  1. 1.
    Kawasaki T (1967) Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children. Arerugi 16:178–222PubMedGoogle Scholar
  2. 2.
    Kato H (1996) Long term consequences of Kawasaki disease. Circulation 94:1379–1385PubMedGoogle Scholar
  3. 3.
    Newburger JW, Takahashi M, Gerber MA et al (2004) Diagnosis, treatment, and long-term management of Kawasaki disease: a statement for health professionals from the Committee on Rheumatic Fever, Endocarditis and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association. Circulation 110:2747–2771PubMedCrossRefGoogle Scholar
  4. 4.
    Goo HW, Park IS, Ko JK et al (2006) Coronary CT angiography and MR angiography of Kawasaki disease. Pediatr Radiol 36:697–705PubMedCrossRefGoogle Scholar
  5. 5.
    Kanamaru H, Sato Y, Takayama T et al (2005) Assessment of coronary artery abnormalities by multislice spiral computed tomography in adolescents and young adults with Kawasaki disease. Am J Cardiol 95:522–525PubMedCrossRefGoogle Scholar
  6. 6.
    Budoff MJ, Achenbach S, Duerinckx A (2003) Clinical utility of computed tomography and magnetic resonance techniques for noninvasive coronary angiography. J Am Coll Cardiol 42:1867–1878PubMedCrossRefGoogle Scholar
  7. 7.
    Sohn S, Kim HS, Lee SW (2004) Multidetector row computed tomography for follow-up of patients with coronary artery aneurysms due to Kawasaki disease. Pediatr Cardiol 25:35–39PubMedCrossRefGoogle Scholar
  8. 8.
    Greil GF, Stuber M, Botnar RM et al (2002) Coronary magnetic resonance angiography in adolescents and young adults with kawasaki disease. Circulation 105:908–911PubMedCrossRefGoogle Scholar
  9. 9.
    Flacke S, Setser RM, Barger P et al (2000) Coronary aneurysms in Kawasaki’s disease detected by magnetic resonance coronary angiography. Circulation 101:E156–157PubMedGoogle Scholar
  10. 10.
    Mavrogeni S, Papadopoulos G, Douskou M et al (2004) Magnetic resonance angiography is equivalent to X-ray coronary angiography for the evaluation of coronary arteries in Kawasaki disease. J Am Coll Cardiol 43:649–652PubMedCrossRefGoogle Scholar
  11. 11.
    Brix G, Lechel U, Veit R et al (2004) Assessment of a theoretical formalism for dose estimation in CT: an anthropomorphic phantom study. Eur Radiol 14:1275–1284PubMedCrossRefGoogle Scholar
  12. 12.
    Onouchi Z, Hamaoka K, Sakata K et al (2005) Long-term changes in coronary artery aneurysms in patients with Kawasaki disease: comparison of therapeutic regimens. Circ J 69:265–272PubMedCrossRefGoogle Scholar
  13. 13.
    Chu WC, Mok GC, Lam WW et al (2006) Assessment of coronary artery aneurysms in paediatric patients with Kawasaki disease by multidetector row CT angiography: feasibility and comparison with 2D echocardiography. Pediatr Radiol 36:1148–1153PubMedCrossRefGoogle Scholar
  14. 14.
    Blobel J, Baartman H, Rogalla P et al (2003) Spatial and temporal resolution with 16-slice computed tomography for cardiac imaging. Rofo 175:1264–1271PubMedGoogle Scholar
  15. 15.
    Thompson BH, Stanford W (2005) Update on using coronary calcium screening by computed tomography to measure risk for coronary heart disease. Int J Cardiovasc Imaging 21:39–53PubMedCrossRefGoogle Scholar
  16. 16.
    Achenbach S, Ulzheimer S, Baum U et al (2000) Noninvasive coronary angiography by retrospectively ECG-gated multislice spiral CT. Circulation 102:2823–2828PubMedGoogle Scholar
  17. 17.
    Pannu HK, Flohr TG, Corl FM et al (2003) Current concepts in multi-detector row CT evaluation of the coronary arteries: principles, techniques, and anatomy. Radiographics 23 Spec No:S111–S125PubMedCrossRefGoogle Scholar
  18. 18.
    Frush DP, Yoshizumi T (2006) Conventional and CT angiography in children: dosimetry and dose comparisons. Pediatr Radiol 36:154–158PubMedCrossRefGoogle Scholar
  19. 19.
    Zanzonico P, Rothenberg LN, Strauss HW (2006) Radiation exposure of computed tomography and direct intracoronary angiography: risk has its reward. J Am Coll Cardiol 47:1846–1849PubMedCrossRefGoogle Scholar
  20. 20.
    Delhaye D, Remy-Jardin M, Salem R et al (2007) Coronary imaging quality in routine ECG-gated multidetector CT examinations of the entire thorax: preliminary experience with a 64-slice CT system in 133 patients. Eur Radiol 17:902–910PubMedCrossRefGoogle Scholar
  21. 21.
    Kleinerman RA (2006) Cancer risks following diagnostic and therapeutic radiation exposure in children. Pediatr Radiol 36 [Suppl 14]:121–125PubMedCrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2007

Authors and Affiliations

  • Raoul Arnold
    • 1
    • 2
  • Sebastian Ley
    • 3
    • 4
  • Julia Ley-Zaporozhan
    • 3
  • Joachim Eichhorn
    • 1
  • Jens-Peter Schenk
    • 4
  • Herbert Ulmer
    • 1
  • Hans-Ulrich Kauczor
    • 3
  1. 1.Department of Paediatric CardiologyUniversity HospitalHeidelbergGermany
  2. 2.Department of Paediatric CardiologyUniversity HospitalFreiburgGermany
  3. 3.Department of Radiology, German Cancer Research CentreHeidelbergGermany
  4. 4.Department of Paediatric RadiologyUniversity HospitalHeidelbergGermany

Personalised recommendations