3D computed tomography angiography as a novel post-processing approach in diagnosis of pediatric malignant bone tumors

Abstract

OBJECTIVE:

This study aimed to investigate the application of 3D computed tomography (CT) angiography with a novel post-processing technique in diagnosis of malignant bone tumors in children.

METHODS:

Twenty-seven pediatric patients (15 males and 12 females; average age: 10±3.4 years old, with a range from 2 months to 14 years old) with suspected bone tumors were evaluated histopathologically using 3D CT angiography and a multislice scanner. CT angiography image data were analyzed with a novel post-processing technique that included separating, fusing opacifying false-coloring, and volume rendering.

RESULTS:

Among 27 cases, 20 (74%) osteosarcoma, 6 (22%) Ewing’s sarcoma, and 1 (4%) non-Hodgkin lymohoma were diagnosed by histological examination of surgical specimens. The tumor features, including size, location, invasion into the adjacent tissue as well as distant metastases, were clearly visualized with the regular volume rendering method and rotational and stereoscopical videos. The post-processing technique provided the reconstructed structure images without any overlap or shelter independently and collectively. Special colors represented different tissue structures, aiding in identification of various anatomical structures and pre-surgical planning.

CONCLUSIONS:

Compared to traditional 3-D CT methods, 3-D CT angiography with rotational and stereoscopical videos provides more detailed information of bone tumor lesions. It offers a superior and effective imaging technique in pediatric patients with malignant bone tumors.

Keywords

Malignant bone tumor post-processing technique computed tomography angiography imaging vivisection

1 Introduction

Bone tumors occurring in skeletal tissues or components include benign and malignant tumors. Malignant bone tumors are the seventh leading form of cancer in children and adolescents (4% of all cancers), with an increasing incidence rate in recent years [1]. There are three subtypes of malignant bone tumors: primary, secondary and metastatic. Bone and surrounding tissues are destroyed by these tumors, leading to persistent pain, loss of motion, malformation, and/or pathologic fracture. These patients suffer from physical pain, psychological distress and financial burden. Chemotherapy and surgical excision are the main treatments for bone malignant tumors [2]. The prognosis for pediatric patients could be substantially improved with earlier diagnosis and treatment [3].

The diagnosis of bone malignant tumor is usually based on clinical manifestations and imaging features [4]. Ultrasonography, standard X-ray, computed tomography (CT), magnetic resonance imaging (MRI) and bone scan are the most common imaging techniques in the detection of bone tumors [5 –7], as described in Table 1. Ultrasonography is a safe, simple, and inexpensive test that can also be used. However, ultrasonography is limited in skeletal imaging due to the inability of ultrasound waves to penetrate bone [8].

Table 1
The imaging methods in detection of malignant bone tumors

Advantages Disadvantages

Ultrasonography Simple, safe and inexpensive, play an important role in guiding needle biopsy and follow-up Difficulty in diagnosing the changes of bone tumor

X-ray High spatial resolution benefiting to display bone changes, calcifications and thin periosteal reaction. Difficulty in identifying the mini tumors from calcifications, and the complicated structures from soft tissue masses.

3D CT Display accurately the difference between mini tumor lesions and calcifications; identify most parts of soft tissue masses. Difficulty in demonstrating thin periosteal reaction.

3D CT Angiography Display rotating stereoscopic videos from various angles, provide accurate information and imaging vivisection Request an invasive procedure, side effects of contrast agent

MRI Sensitive in case of tumors permeating the bone medullary spaces without severely destroying the bone trabeculae. Display soft tissue masses, the relation between tumors and adjacent organs or soft tissues. Difficulty in demonstrating calcifications, high costs.

Bone scan Help in detecting any ‘active’ area in the bone, depict lesion quiescence or activity Unable to demonstrate the characterization of the ‘active’ area, high costs.

	Advantages	Disadvantages
Ultrasonography	Simple, safe and inexpensive, play an important role in guiding needle biopsy and follow-up	Difficulty in diagnosing the changes of bone tumor
X-ray	High spatial resolution benefiting to display bone changes, calcifications and thin periosteal reaction.	Difficulty in identifying the mini tumors from calcifications, and the complicated structures from soft tissue masses.
3D CT	Display accurately the difference between mini tumor lesions and calcifications; identify most parts of soft tissue masses.	Difficulty in demonstrating thin periosteal reaction.
3D CT Angiography	Display rotating stereoscopic videos from various angles, provide accurate information and imaging vivisection	Request an invasive procedure, side effects of contrast agent
MRI	Sensitive in case of tumors permeating the bone medullary spaces without severely destroying the bone trabeculae. Display soft tissue masses, the relation between tumors and adjacent organs or soft tissues.	Difficulty in demonstrating calcifications, high costs.
Bone scan	Help in detecting any ‘active’ area in the bone, depict lesion quiescence or activity	Unable to demonstrate the characterization of the ‘active’ area, high costs.

Standard x-ray imaging is the first test in the diagnosis of bone tumors. X-rays distinguish tissue density and provide good contrast to distinguish bone and surrounding soft tissue. In addition, X-ray is able to detect thin periosteal reaction, a very important feature in the early stage of malignant bone tumors. However, it is less beneficial in the differential diagnosis of small tumors from calcifications, especially in the tumors of complicated bone structures such as spine, skull base and pelvis.

MRI is also an important imaging technique. It is sensitive in the examination of tumors that permeate the bone medullary spaces without destroying the bone trabeculae. MRI also serves as a valuable tool to differentiate between tumors and adjacent organs or soft tissues, but fails to display some bone changes, such as calcification [9].

Bone scans are helpful in detecting ‘active’ bone formation and destruction. Any process that is accompanied by active osteogenesis produces an ‘active’ area in the bone scan. Such changes are found in bone tumors with neoplastic or reactive osteogenesis, but also fracture, arthritis, infection, or rheumatoid arthritis [10].

CT with high density resolution possesses special advantages in the diagnosis of bone tumors [11]. Compared to other medical imaging technologies for oncology diagnostics, CT can display bone changes, including small changes in complicated structures. CT can satisfactorily image soft tissue masses. CT can measure the density of tissues through CT in contrast to standard X-ray. Moreover, three-dimensional (3D) CT can provide rotating stereoscopic reconstructed views, more comprehensive and accurate information, contributing favorably to the diagnosis and treatment planning of bone tumor patients. However, as anamorphic images, 3D CT images have still weakness. The overlapping of multi-layers may affect the veracity of reconstructed anatomical structures, since the reconstructed images are gained through routine volume rendering post-processing imaging technique. In addition, structures with similar densities cannot be obviously distinguished by naked e+ in gray-scale images.

On the basis of 3D CT imaging technique, 3D CT angiography can further process CT volume data with a novel post-processing technique. After completion of CT volume data acquisition, 3D images can be reconstructed by a novel and combined post-processing technique, including separating, fusing, pacifying and false-coloring. The 3D CT angiography can reconstruct any focus structure, and assemble two or more structures together in an image without any overlap or shelter. Moreover, distinctive colors marked on structures facilitate the differentiation of anatomical structures [12]. The 3D CT angiography is widely used to diagnose cardiopulmonary diseases, vascular disorders, liver tumors in pediatric patients and devise pre-surgery planning [13 –20]. In this study, we report 3D CT angiography with a novel post-processing technique to improve the diagnosis and pre-surgical planning of malignant bone tumors in children.

2 Material and methods

In our hospital, 27 patients (aged between 2 months and 14 years, mean age 10±3.4 years, 15 males and 12 females) with bone involvement presented with swelling pain, and/or mass (see Table 2). Routine CT and x-ray imaging, multiple-phase CT by 3D CT angiography (Philips Brilliance iCT 256, Philips Healthcare Inc., Cleveland, USA) and histopathologic examination were performed. CT scanning parameters were set as follows: tube voltage 100 kV, tube current 150 mAs, thickness 5.0 mm, pitch 0.984, standard resolution, rotation time 0.5 s, FOV 250 millimeter, reconstruction matrix 512×512, window width/level 410 Hounsfield unit (HU) /130 HU. Scanning range was 30–50 centimeters. All patients received 2 ml/kg of nonionic contrast medium (Visipaque, 320 mgI/ml, GE Healthcare Ireland) intravenously. The iDose 3 was used to reconstruct low dose images sets. After completion of CT volume data acquisition, 3D images were reconstructed by a novel and combined post-processing technique on the Workstation v6.0.4.02700 (Philips Intelli Space Portal, Philips Healthcare Nederland B.V., Netherland). “Clip and 3D Segmentation” to create the 3D images of tumor, vessels and bone were used, respectively. In addition, a VR tool to adjust opacity and attached colors was used to obtain fusing 3D images. The rotating 3D structures could be recorded as pictures and movies.

Table 2
Patients’ information

No. Age (yrs) Sex Symptoms Diagnosis Location Size (cm³) Invasion Metastasis

1 13 M pain, swelling OS L tibia 3.4×2.6×1.9 + –

2 12 F pain OS L tibia 4.9×3.0×1.8 + –

3 9 M mass OS L fibula 5.6×4.9×3.7 + –

4 6 F pain OS L femur 9.2×3.2×3.8 – –

5 13 M pain OS L femur 4.5×3.3×1.7 – –

6 8 F pain, dysfunction OS lumbosacral portion 5.9×4.9×4.3 – –

7 0.2 F mass OS R femur 7.6×6.8×7.0 + –

8 15 M pain, mass OS L femur 5.1×3.0×1.1 – –

9 13 M pain, swelling, dysfunction OS R femur 8.3×1.5×2.0 + –

10 11 M pain, swelling OS L femur 7.4×3.8×2.9 – –

11 11 F pain, mass OS L tibia 8.8×2.5×3.4 – –

12 12 F mass, fever OS L femur 5.5×4.0×3.5 + –

13 13 M pain OS R femur 8.0×5.6×7.2 – –

14 8 M mass OS L fibula 10×5.6×4.5 – –

15 11 F mass OS L fibula 7.6×6.2×4.6 + Lungs

16 7 M pain, swelling OS R femur 8.2×3.5×2.2 + –

17 10 F pain OS R femur 10.3×4.2×3.5 – –

18 14 M pain, swelling OS L femur 17×14×6 + –

19 13 F pain OS L humerus 10×15×7 + –

20 10 M no symptom OS L ischium 6.1×4.4×2.7 – –

21 5 M pain, dysfunction EWS lumbosacral portion 10×8.9×4.6 + Lungs

22 14 F pain, dysfunction EWS L ischium 6.2×4.3×3.7 + –

23 6 M pain, mass EWS L scapula 4.3×3.1×2.2 + –

24 10 M mass EWS L fibula 7.5×4.5×2.8 – –

25 12 M pain, swelling EWS R tibia 7.8×5.7×4.3 + Inguinal lymph nodes

26 5 F mass EWS L scapula 8.5×8.8×4.8 + Left humerus

27 8 F pain, mass NHL R humerus 10.3×6.7×7.9 + Axillary lymph nodes

No.	Age (yrs)	Sex	Symptoms	Diagnosis	Location	Size (cm³)	Invasion	Metastasis
1	13	M	pain, swelling	OS	L tibia	3.4×2.6×1.9	+	–
2	12	F	pain	OS	L tibia	4.9×3.0×1.8	+	–
3	9	M	mass	OS	L fibula	5.6×4.9×3.7	+	–
4	6	F	pain	OS	L femur	9.2×3.2×3.8	–	–
5	13	M	pain	OS	L femur	4.5×3.3×1.7	–	–
6	8	F	pain, dysfunction	OS	lumbosacral portion	5.9×4.9×4.3	–	–
7	0.2	F	mass	OS	R femur	7.6×6.8×7.0	+	–
8	15	M	pain, mass	OS	L femur	5.1×3.0×1.1	–	–
9	13	M	pain, swelling, dysfunction	OS	R femur	8.3×1.5×2.0	+	–
10	11	M	pain, swelling	OS	L femur	7.4×3.8×2.9	–	–
11	11	F	pain, mass	OS	L tibia	8.8×2.5×3.4	–	–
12	12	F	mass, fever	OS	L femur	5.5×4.0×3.5	+	–
13	13	M	pain	OS	R femur	8.0×5.6×7.2	–	–
14	8	M	mass	OS	L fibula	10×5.6×4.5	–	–
15	11	F	mass	OS	L fibula	7.6×6.2×4.6	+	Lungs
16	7	M	pain, swelling	OS	R femur	8.2×3.5×2.2	+	–
17	10	F	pain	OS	R femur	10.3×4.2×3.5	–	–
18	14	M	pain, swelling	OS	L femur	17×14×6	+	–
19	13	F	pain	OS	L humerus	10×15×7	+	–
20	10	M	no symptom	OS	L ischium	6.1×4.4×2.7	–	–
21	5	M	pain, dysfunction	EWS	lumbosacral portion	10×8.9×4.6	+	Lungs
22	14	F	pain, dysfunction	EWS	L ischium	6.2×4.3×3.7	+	–
23	6	M	pain, mass	EWS	L scapula	4.3×3.1×2.2	+	–
24	10	M	mass	EWS	L fibula	7.5×4.5×2.8	–	–
25	12	M	pain, swelling	EWS	R tibia	7.8×5.7×4.3	+	Inguinal lymph nodes
26	5	F	mass	EWS	L scapula	8.5×8.8×4.8	+	Left humerus
27	8	F	pain, mass	NHL	R humerus	10.3×6.7×7.9	+	Axillary lymph nodes

Abbreviations: M-male; F-female; OS-osteosarcoma; EWS-Ewing’s sarcoma; NHL-non-Hodgkin Lymphoma. R-right; L-left; +-positive; –- negative.

3 Results

The regular X-ray files showed bone destruction. The 3D CT angiography provided further detailed information of the tumor, such as size, infiltration, blood supply, and the stereoanatomic relationship with surrounding blood vessels. The results of X-ray examinations were consistent with histopathologic examinations. Of the 27 cases, 20 (74%) were osteosarcoma, 6 (22%) were Ewing’s sarcoma, and 1 (4%) was non-Hodgkin lymohoma (Table 2).

As an example, imaging from a ten-year-old boy suffering from an ischial tumor is presented in Figs. 1 and 2. The patient was admitted due to crush-related injuries of the right leg. The x-ray images of lumbar vertebrae, pelvis and double lower limbs showed bone destruction in the left ischium without any fracture sign. He was suspected to suffer from an ischial tumor, and was further examined with 3D CT angiography. Prior to the injection of nonionic contrast medium, the patient first received a CT scan. 3D images were reconstructed and are shown in Fig. 1B. After the injection of nonionic contrast medium, the CT volume data were analyzed with a combined post-processing technique, including separating, fusing, opacifying and false-coloring technique. The reconstructed structures were recorded as pictures (Fig. 2) and movies (See Supplementary Electronic Video). As shown in Fig. 2 and the supplementary electronic video, the left ischial tumor lesions could be viewed clearly with the tumor in brown, blood vessels in red and skeletal system in silver taupe, rotationally and stereoscopically. The clinical manifestations and imaging examinations suggested the bone tumor was osteosarcoma, which was further confirmed by histopathologic examinations (Fig. 3).

Fig.1

3D image of CT scan of the affected hip. A), normal ischium (green circle). B), malignant bone tumor invading the left ischium (red circle), see the attached video.

Fig.2

Reconstructed images of 3D CT angiography: A). Left ischium osteosarcoma (brown color); B). Common iliac artery, internal iliac artery, external iliac artery and their branches (red color); C). Fusing image of left ischium osteosarcoma and blood vessels, D). Fusing image of tumor, blood vessels and the left hip joint bones.

Fig.3

Pathophysiology of malignant bone tumor in the left hip. Numerous pleomorphic anaplastic cells (short and thicker arrows) and some tumor-induced chondral tissues (a long and thinner arrow) are noted. H&E staining×200.

4 Discussion

Three-D CT images show rotating stereoscopic reconstructed views from various angles, providing better understanding of patients’ diseases and complex anatomy, compared to 2D CT images that require special skills to interpret the captured data [21]. Recent advances in computer processing and 3D computer graphics allow 3D reconstruction of target structures with various imaging modalities [22]. The most common post-processing method is the routine volume rendering imaging technique. Three-D CT images reconstructed by volume rendering imaging technique tend to be anamorphic [23]. Volume rendering imaging techniques reconstruct anatomical structures by overlapping multi-layer images. Accurate imaging of lesions may be reduced during this process. Furthermore, in gray-scale images, the structures with similar density may not be easily distinguished. Three-D CT angiography can overcome these disadvantages.

The 3D CT angiography displays rotating stereoscopic 3D images from various angles, helping provide better understanding of the lesions and complex anatomy nearby. In addition, it can be performed on imaging vivisection because its volume data is analyzed by a novel and combined post-processing technique, including separating, fusing, opacifying and false-coloring. By separating technique, any interested or suspected structure may be extracted from the entire object. By fusing and opacifying technique, two or more arbitrary structure images are able to be put together in an image without any overlap or shelter, and obtain satisfactory images showing target structures. It is possible to observe the whole anatomic structures and measure the size of target structures. Accurate evaluation of tumor morphologic characteristics and 3D anatomic relations may offer considerable advantages in the diagnosis and treatment planning of bone tumor patients. Moreover, unique colors can be attached to different structures and facilitated the identification of anatomical structures.

In this study, 3D CT angiography was applied in the diagnosis and treatment planning of bone tumor patients. By 3D CT angiography with the novel post-processing technique, the bone tumor lesions could be reconstructed independently to show the shape, size and invasion of bone tumor. The blood vessels also could be reconstructed independently to demonstrate the blood supply of tumor clearly. In addition, interrelationship of tumors vis-a-vis adjacent structures, such as blood vessels, could be clearly observed and intuitively measured by fusing them together. The results indicate that 3D CT angiography not only facilitates the imaging diagnosis, but also could be used to pre-surgery planning.

In angiography, contrast-induced nephropathy is a serious complication of diagnostic and interventional procedures. Recent study showed that the incidence of contrast-induced nephropathy was lower in the nonionic contrast medium (Visipaque) group (7.9%) as compared with the hexabrix group (17%) [24] in the patients with renal insufficiency after coronary angiography. In addition, risk of nephropathy induced by contrast medium reduced is increased in patients with renal insufficiency when as the contrast used is not low-osmolar or a nonionic contrast medium [25]. In the present study, Visipaque 320, a nonionic water-soluble contrast medium with the same osmolality as blood (290 mOsm/kg H₂O), was used as the contrast medium, and there was no contrast-induced nephropathy in all cases.

Shortcomings of 3D CT angiography include scanning the patients twice, resulting in increased effort for patient and staff and additional radiation exposure. It also requires xperienced radiologists because non-target structures may result in misdiagnosis or missed diagnosis. Finally, the contrast medium may induce nephropathy although it did not occur in this study.

5 Conclusions

Three-dimensional CT angiography allows detailed observation of malignant bone tumor lesions. Using a novel post-processing technique composed of separating, fusing, opacifying and false-coloring, 3D CT angiography displays rotating stereoscopic videos from various angles, providing accurate information and imaging vivisection compared to regular X-ray or 2D CT images. The 3D CT angiography offers an effective diagnostic technology in pediatric patients with malignant bone tumors.

Footnotes

The Supplementary one Piece of Electronic Video is available in the electronic version of this article: .

References

Ward

, DeSantis

, Robbins

, Kohler

and Jemal

, Childhood and Adolescent Cancer Statistics, 2014, Ca Cancer J Clin64 (2014), 83–103.

Chowdhury

M.M.

, Dagash

and Pierro

, A systematic review of the impact of volume of surgery and specialization on patient outcome, Br J Surg94 (2007), 145–161.

Grimer

R.J.

and Briggs

T.W.

, Earlier diagnosis of bone and soft-tissue tumours, J Bone Joint Surg Br92 (2010), 1489–1492.

Grimer

R.J.

, Carter

S.R.

, Spooner

and Sneath

R.S.

, Diagnosing musculoskeletal tumours, Sarcoma5 (2001), 89–94.

Morley

and Omar

, Imaging evaluation of musculoskeletal tumors, Cancer Treat Res162 (2014), 9–29.

Kransdorf

M.J.

and Bridges

M.D.

, Current developments and recent advances in musculoskeletal tumor imaging, Semin Musculoskelet Radiol17 (2013), 145–155.

Campanacci

, Mercuri

, Gasbarrini

and Campanacci

, The value of imaging in the diagnosis and treatment of bone tumors, European Journal of Radiology27 (1998), S116–S122.

Van der Woude

H.J.

and Vanderschueren

, Ultrasound in musculoskeletal tumors with emphasis on its role in tumor follow-up, Radiol Clin North Am37 (1999), 753–766.

Subhawong

T.K.

and Wilky

B.A.

, Value added: Functional MR imaging in management of bone and soft tissue sarcomas, Curr Opin Oncol27 (2015), 323–331.

10.

Tateishi

, Yamaguchi

, Seki

, Terauchi

, Arai

and Kim

E.E.

, Bone and soft-tissue sarcoma: Preoperative staging with fluorine 18 fluorodeoxyglucose pet/ct and conventional imaging, Radiology245 (2007), 839–847.

11.

Fayad

L.M.

, Johnson

and Fishman

E.K.

, Multidetector CT of musculoskeletal disease in the pediatric patient: Principles, techniques, and clinical applications, Radiographics25 (2005), 603–618.

12.

Duan

, Zhong

and Ye

, Three-dimensional CT imaging with volume rending of separating, fusing, opacifying and false-coloring technique, Applied Mechanics and Materials Vols423-426 (2013), 2606–2609.

13.

Manolis

A.J.

, Chamodraka

and Zacharopoulou

, Modern Diagnostic Approach for the Assessment of Cardiac Damage in Hypertension: 3D CT and MRI, Assessment of Preclinical Organ Damage in Hypertension (2015), 25–37.

14.

Wadaa

, Nawashirob

, Ohkawac

, Arimotoc

, Takeuchia

and Moria

, Feasibility of the combination of 3D CTA and 2D CT imaging guidance for clipping microsurgery of anterior communicating artery aneurysm, British Journal of Neurosurgery29 (2015), 229–236.

15.

Bruce

A.U.

, Lloyd

E.R.

and Elliot

K.F.

, Three-dimensional Volume-rendered CT Angiography of the Renal Arteries and Veins: Normal Anatomy, Variants, and Clinical Applications, Radio Graphics21 (2001), 373–386.

16.

Zhang

, Zeng

, Zhang

, Liu

, Wu

, Hua

, Liu

, Lu

, Feng

, Qin

, Cai

, Zhang

, He

, Lin

and Wei

, 3D Reconstruction Computed Tomography Scan in Diagnosis of Bilateral Wilm’s Tumor with Its Embolus in Right Atrium, J Xray Sci Technol24 (2016), 657–660.

17.

, Sun

, Zhang

, Li

, Hu

and Song

, 3D Reconstruction Images of Cone Beam Computed Tomography in Dental Medicine Application: A Case Study and Mini-review, J Xray Sci Technol24 (2016), 673–680.

18.

Peng

, Wang

, Zhang

, Fu

, Fan

and Zhang

, Diagnosisvalue of multi-slice spiral CT in renal trauma, J Xray SciTechnol24 (2016), 649–655.

19.

, Li

, Guo

, Gu

, Xiao

, Ye

, Pan

, Zhang

and Huang

, Bilateral duplex urinary collecting systems accompanied with horseshoe kidneys deformity and right renal ureteral calculi and hydronephrosis: Diagnosis in magnetic resonance urography, J Xray Sci Technol25 (2017), 667–672.

20.

Zhang

, Li

, Mo

, Peng

, Yang

, Tang

, Zhang

and Wu

, Benign mixed epithelial and stromal tumor of the kidney with inferior vena cava tumor thrombus: A rare case report and review of literature, J Xray Sci Technol25 (2017), 831–837.

21.

Ferretti

G.R.

and Jankowski

, Volumetric CT scanning: 2D and 3D reconstruction, Rev Mal Respir27 (2010), 1267–1274.

22.

Chen

, Zeng

, Li

, Gong

, Peng

, Wu

, Zhang

and Zhang

, Evaluation of a large adrenal carcinoma with 3D reconstruction of computed tomography images: A case report and literature review, J Xray Sci Technol24 (2016), 665–671.

23.

Fleischmann

, Chin

A.S.

, Molvin

, Wang

and Hallett

, Computed Tomography Angiography: A Review and Technical Update, Radiol Clin North Am54 (2016), 1–12.

24.

S-H.

, Youn

T-J.

, Koo

B-K.

, Park

J-S.

, Kang

H-J.

, Cho

Y-S.

, Chung

W-Y.

, Joo

G-W.

, Chae

I-H.

, Choi

D-J.

, Oh

B-H.

, Lee

M-M.

, Park

Y-B.

and Kim

H-S.

, Renal toxicity evaluation and comparison between Visipaque (iodixanol) and Hexabrix (ioxaglate) in patientswith renal insufficiency undergoing coronary angiography: Therecover study: A randomized controlled trial, J Am CollCardiol48 (2006), 924–930.

25.

Aspelin

, Aubry

, Fransson

S-G.

, Strasser

, Willenbrock

and Berg

K.J.

, Nephrotoxic effects in high-risk patients undergoing angiography, N Engl J Med348 (2003), 491–499.