Khawar M. Gul, Song S. Mao, Yanlin Gao, Ronald J. Oudiz, M. Leila Rasouli, Ambarish Gopal, Matthew J. Budoff
Rationale and Objectives. Although the use of gadolinium (Gd)-based contrast agents for angiographic studies of abdom-inal aorta, renal vasculature, and digital subtraction angiography has been reported, no studies have shown their diagnosticefficacy and image quality in coronary computed tomography angiography (CTA). The aim of this study is to evaluate theimage quality of Gd-based contrast agents during coronary CTA.
Materials and Methods. A phantom study was done to evaluate the attenuation of serial dilutions of Gd and iodinatedagents. This study was done on a 64-slice multirow detector CT (MDCT) scanner and e-speed scanner and CT attenuationHounsfield units (CTHU) were compared. We evaluated 35 consecutive patients who underwent Gd-enhanced CTA.CTHU of aorta was measured at first, mid, and lowest slice levels with and without contrast administration. The imagequality was graded on the basis of visualization of the coronary arteries (scale I-III; III demonstrating diagnostic imagequality of the distal-most vessels). In a substudy, four patients with borderline renal insufficiency underwent CTA usingGd and iodinated contrast agents admixed in a 50:50 ratio.
Results. The phantom study showed that enhancement of various dilutions of Gd and iodine resulted in near identicalCTHU with both e-speed and 64-slice scanners (r 2 � 0.997). Mean CTHU with contrast at the top slice was 116 HU, atmiddle slice was 125 HU, and at the lower slice was 93 (111.14 � 22). Quality evaluation showed 2 grade III, 9 grade II,and 24 grade I images (average quality of images 1.35). Mean CTHU was 222.
Conclusion. Gd-enhanced contrast medium provides adequate enhancement of coronary vasculature, allowing for diagnos-tic evaluation of coronary arteries with new CT systems. Use of newer generations of multirow detector CT scannersshould further enhance the quality of images.
Noninvasive cardiac computed tomographic angiography(CTA) is rapidly becoming a widely accepted tool forevaluations of the coronary arteries (1). The use of con-trast agents has increased tremendously with the increase
Acad Radiol 2006; 13:840–849
1 From the Division of Cardiology, Los Angeles Biomedical Research Insti-tute at Harbor-UCLA, 1124 W. Carson Street, RB2, Torrance, CA 90502(K.M.G., S.S.M., R.J.O., M.L.R., A.G., M.J.B.). Received February 1, 2006;accepted April 12, 2006. Address correspondence to: M.J.B. e-mail:[email protected]
in diagnostic and interventional angiographic procedures.Contrast use leads to adverse reactions such as iodinerelated contrast nephropathy and renal failure, making itthe third leading cause of new acute renal failure in hos-pitalized patients and are responsible for 4.6%–8.5% ofanaphylactoid reactions (2– 4). This has led to avoidanceof these diagnostic and therapeutic procedures in patientswith renal insufficiency or allergies to iodine. Gadolinium(Gd) chelates have been widely used for magnetic resonanceimaging procedures over the past 10 years and are regardedas less nephrotoxic and thus can be used increasingly in pa-
tients with renal insufficiency. Recently, Gd chelates have
Academic Radiology, Vol 13, No 7, July 2006 NONINVASIVE GD-ENHANCED 3D CT CORONARY ANGIOGRAPHY
Figure 1. Phantom image demonstrating attenuation of different dilutions of gadolinium and iodine scanned on a multirow detectorcomputed tomography 64 scanner. Computed tomographic image showing attenuation of various dilutions of gadolinium (top) and iodine(bottom), ranging from 1:1 to 1:512 dilutions, with decreasing attenuation across these concentrations. The iodine remains significantly
brighter, approximating a fourfold increase in concentration compared with gadolinium.
Table 1In Vitro Hounsfield Unit (HU) Measurements of Serial Gadolinium (Magnevist) and Iopromide (Ultravist) Using E-Speed Scanner(EBCT) and 64-Slice Multirow Detector CT
GUL ET AL Academic Radiology, Vol 13, No 7, July 2006
Figure 2. (a) Correlation of attenuation values measured on multirow detector computed tomography (MDCT) and electron beam com-puted tomography (EBCT) scanners for various dilutions of iopamidole from 1:1 to 1:512 concentrations. Dilutions of iopamidole arelisted on the chart. (b) Correlation of attenuation values measured on MDCT and EBCT scanners for various dilutions of gadolinium from
1:1 to 1:512 concentrations. Dilutions of iopamidole are listed on the chart.
842
Academic Radiology, Vol 13, No 7, July 2006 NONINVASIVE GD-ENHANCED 3D CT CORONARY ANGIOGRAPHY
been used for digital subtraction angiography and CT imag-ing of pulmonary, abdominal, and renal arteries. There havealso been case reports about the successful use of Gd forcoronary catheterization in patients with renal insufficiency.These studies have shown the efficacy of use of Gd for di-agnostic purposes. However, no study has been done tocompare the efficacy of Gd to iodine-based agents in pa-tients undergoing cardiac CTA. The purpose of this studywas to show the efficacy of Gd in terms of image qualityand diagnosis in patients undergoing coronary CTA.
MATERIALS AND METHODS
In Vitro StudyAn in vitro study was performed to determine relative CT
enhancement of Gd-based contrast agent (Magnevist [gado-pentetate dimeglumine]: Berlex Imaging, Wayne, New Jer-sey) compared with iodinated contrast agent (Iopromide,Ultravist). Using disposable tubes, various dilutions of Gdand iodine with normal saline were prepared, ranging from1:1 to 1:512 dilutions for each agent. CT scanning was doneand attenuation of each dilution was recorded as meanHounsfield units using region of interest (ROI) analysis (Fig1). For each dilution, three ROI measurements were doneand mean Hounsfield units were calculated. The same imag-ing was then done using a 64-slice multirow detector CT(MDCT); for each dilution, three ROI measurements weredone and mean CT attenuation Hounsfield units (CTHU)were calculated.
CT TechniqueCTA studies were performed using an e-speed com-
puted topographic scanner (GE, San Francisco, CA) (1).First, a noncontrast 40-slice study (3-mm slice thickness,3-mm table increment, 100-millisecond acquisition time,130 kVp, and 900 MA) was obtained with the patientssupine and no couch angulations. Subsequently, a flowstudy was performed for evaluation of the contrast agenttransit time (scan delay time) to the descending aorta.The injection rate was 4 mL/second with 8 mL Gd and30 seconds’ image time. Finally, 60–70 contiguous axialCTA images were obtained with injection protocols asdescribed in the following section. Electrocardiographytriggering was employed, corresponding to end-systole aspreviously described. The slice thickness was 1.5 mmwith two-row detectors, 140 kVp, and 1,000 MA. The imageacquisition time was 100 milliseconds per image and total
scan time was 20–35 seconds (mean 27 seconds).
Two phase injection protocols were employed in thisstudy. The first contrast media infusion rate was 3–5 mL/second with 30 mL of contrast media; the second contrastmedia was infused at rate of 2–3 mL/second with30–50 mL, followed by a 30–40 saline injection with aninjection rate of 3 mL/second. The total injection timewas 20 seconds with average of 68 mL Gd infused (range50–80 mL). A dual injector (Stellant Dual Injector, Med-rad, Inc, Indianola, PA) was used. Arterial opacificationbased on CTHU of the ascending or descending aortawere measured at three prespecified levels: first, mid, andlatest slice in all studies (Insight Workstation, NeoImag-ery Technologies, City of Industry, CA). In a subset offour patients, Gd-based contrast agents were mixed withiodinated contrast agents in various concentrations. Thesepatients were selected on the basis of borderline renalinsufficiency. All patients were selected based on requir-
Figure 3. High-grade stenosis of the mid-right coronary artery(arrow) diagnosed using gadolinium enhancement on cardiaccomputed tomography.
ing a CTA for diagnosis of chest pain, with at least bor-
843
GUL ET AL Academic Radiology, Vol 13, No 7, July 2006
derline renal dysfunction (creatinine �1.3 mg/dL) orhigher. The study was approved by the InvestigationalReview Board of the Human Investigations Committee atour institution.
RESULTS
Phantom StudyFor each dilution of Gd and iodinated contrast, three
ROI were measured and mean Hounsfield units were cal-culated (Fig 1) . The same imaging was then done using a64-slice MDCT; for each dilution, three ROI measurements
Figure 4. Grade 2 image quality. Figure showsdescending artery (black arrow) with mild stenostion of diagonal branches.
were done and mean CTHU was calculated (Table 1). The
844
correlation between attenuation of Electron Beam Com-puted Tomography (EBCT) and MDCT was very high forboth Gd and Iopromide (r2 � 0.997, Fig 2). To achievesimilar enhancement (HU) as iodinated contrast, at leasttwofold doses of Gd are required for both CT systems.
Patient StudyThirty-five consecutive patients (22 male; mean age 60
years; range 37–86 years) underwent Gd-enhanced car-diac CTA between July 2004 and July 2005. The diag-nostic efficacy of the images was assessed by visualiza-tion of different segments of the coronary vasculature.The image quality was graded on the basis of visualiza-
mal intensity projection of a mid left anteriorcalcified plaque without adequate visualiza-
maxiis and
tion of right coronary artery, proximal to mid-left anterior
Academic Radiology, Vol 13, No 7, July 2006 NONINVASIVE GD-ENHANCED 3D CT CORONARY ANGIOGRAPHY
descending (LAD) and proximal to mid left circumflex(Grade I), distal segment of LAD, middle segment of leftcircumflex artery (grade II; Fig 3, 4) and visualization ofdistal left circumflex artery and diagonal branches ofLAD (grade III; Fig 5, 6). Images were read by a cardiol-ogist who was blinded to the use of Gd contrast dye forthe study. Mean CTHU without contrast was 52 (range47–60, SD 22.76) measured at the level of ascendingaorta (coronary ostium). Mean CTHU with contrast at thetop slice was 116 HU, at middle slice was 125 HU, andat the lower slice was 93 (111.14 � 22). Quality evalua-tion showed two grade III, 9 grade II, and 24 grade I im-ages (average quality of images 1.35) (Table 2). Themean CTHU obtained was 111 and was deemed sufficientto assess coronary segments in the right coronary arteryand proximal and middle segments of left anterior de-scending and left circumflex coronary artery. In all buttwo cases, we were unable to visualize distal segments ofLAD and left circumflex artery with confidence.
In the substudy in which Gd was mixed with iodinated
Figure 5. Grade 3 image quality. Distal left anterior descesignificant stenosis is seen.
contrast agents in four patients, the quality of images en-
hanced significantly. The image quality was greatly en-hanced by mixing iodine with Gd, the mean CTHU withGd � contrast was 249 at the top level, 222.5 at the mid,and 144.5 at the lower level (Table 2a). The mean CTHUincreased to 222 for all levels, significantly greater thanwith Gd only (P � .001).
SafetyAll 35 patients tolerated Gd without any sequelae. No
patient in the study experienced any allergic reaction,hives, anaphylaxis, and shortness of breath, itching, orrash. No patient experienced lightheadedness or throattightness.
DISCUSSION
CTA allows reliable diagnosis of various clinical con-ditions including coronary stenosis, cerebral aneurysm,vessel occlusion, cervical intra-arterial stenosis, aortic
artery and diagonal branches (D) are well visualized. No
nding
diseases, and evaluation of pulmonary embolism. Side
845
GUL ET AL Academic Radiology, Vol 13, No 7, July 2006
effects related to iodinated contrast agents have decreasedbecause of use of nonionic agents; however, the incidenceof allergic reactions secondary to use of iodinated contrastagents are still a big concern. Renal failure related to con-trast exposure is the third leading cause of acute renalfailure in hospitalized patients (2– 4).
Reactions range from mild symptoms such as itching,nausea, or vomiting to severe life-threatening events such asdyspnea, circulatory shock, and even cardiac arrest (5). Pa-tients with renal insufficiency or allergic reactions to iodin-ated contrast dyes pose a challenge if they need imagingstudies requiring iodinated contrast dyes. This has led to anincreased interest in Gd-based studies, including digital sub-traction angiography, CTA, and x-ray angiography (5–18).
In experimental studies aimed at assessing the pharma-
Figure 6. Grade 3 image quality. The posterior descendvisualized with diagnostic image quality in this study.
cokinetics of Gd, it has been shown that it is exclusively
846
excreted through the kidney within 24 hours with normalrenal function. In patients with renal insufficiency, however,it takes a few days for excretion (19,20). The incidence ofadverse reactions to Gd agents is low with occurrence ofmoderate and severe reactions well below 1% (21–23). Al-though the safety of Gd-related contrast agents has beenadvocated in patients with normal renal function, there havebeen reports of adverse effects ranging from mild or negligi-ble renal insufficiency to acute renal failure in patients withdiabetes mellitus given large doses of Gd (24,25).Dose-related deterioration in renal function has alsobeen reported by Spinosa et al in patients withazotemia when they used 70 mL Gd (0.3 mmol/kg).However, in a study by Gemmete et al, dose in therange of 0.5–2.9 mmol per kilogram was used without
lack arrow) and posterolateral marginal branches are
ing (b
incidence of renal dysfunction (26).
Academic Radiology, Vol 13, No 7, July 2006 NONINVASIVE GD-ENHANCED 3D CT CORONARY ANGIOGRAPHY
Table 2Baseline Patient Data, Including Gadolinium Doses and Hounsfield Unit (HU) Measures, at Various Positions Throughoutthe Scan
GUL ET AL Academic Radiology, Vol 13, No 7, July 2006
The maximum dose of Gd compounds, according tomanufacturer’s recommendations, is 0.2 mmol per kilo-gram for gadopentate dimeglumine and 0.3 mmol per ki-logram for gadoterate and gadodiamide. In this study, weused either the upper limit of recommended dose orslightly exceeded the recommended dose. Further safetyand angiographic studies need to be done to validate theefficacy of this approach. Previous studies have suggesteda role of Gd in conventional angiography in patients withcontraindications to iodine (27,28).
In human CTA studies, uniform enhancement of 140HU was sufficient enough to identify vessel segments. Inour study we measured Hounsfield Units at three levelsalong the aorta. The maximum enhancement was obtainedat the mid level of aorta. The maximum enhancementobtained in our study was 209 HU and the mean enhance-ment was 125 HU, which was deemed optimal enoughfor diagnostic purposes by blinded readers. The range ofHU was between 90 and 209 at various levels of theaorta. The proximal portion of coronary tree including theleft main coronary artery, proximal to mid LAD artery,and the right coronary artery showed good enhancementin almost all cases.
Mixing StudyIn a phantom study by Badiola et al, the optical den-
sity of Gd mixed with iodinated contrast agent was com-pared with full-strength Gd, full-strength iodinated con-trast agent, and 20% solution of iodinated contrast agent.This study showed significant improvement in imagequality of Gd-iodinated contrast agents when comparedwith full-strength Gd agents alone and could be comparedfavorably with full-strength iodinated contrast media (29).In four patients with mild or borderline renal dysfunction,iodinated contrast was mixed with Gd (50:50 dilution),which greatly improved the enhancement compared withcases in which Gd alone was used. The mean CTHU was222 as compared with 110 with Gd alone. The imagequality greatly improved with use of iodine and Gd to-gether. The image quality was considered satisfactory fordiagnostic purposes for both the Gd alone and Gd-iodin-ated contrast studies. This study provides an added valuein patients with borderline renal dysfunction, becauselower doses of iodinated contrast agents are required.
LimitationsAlthough the images obtained were of fair quality,
we were unable to effectively visualize distal segments
of the coronary tree. This problem could be handled by
848
increasing the dose of Gd used or mixing iodinatedcontrast with Gd. In our study, the quality of imagesimproved with mixing iodine and Gd. This approachmay prove feasible to limit doses of both agents, po-tentially effectively lowering the nephrotoxicity of ei-ther drug. With the advent of MDCT, it is expectedthat the quality of images will improve, because 64-rowdetector scanners will allow for faster coverage of thecoronary tree, decreasing the need for higher volumes ofcontrast, further diminishing the doses of both Gd andiodine-based agents.
Another limitation to use of Gd-based contrast agentsis cost. It is generally claimed that these agents are aboutfour to five times more expensive per milliliter than areiodinated nonionic compounds. It requires an approximatefourfold dose to obtain equal attenuation (Table 1, dem-onstrating enhancement with different dilutions of Gd andiodine on both e-speed and 64-MDCT). With equal-atten-uating doses of iodinated nonionic monomers, Gd-relatedagents may cost up to 20 times more than iodinatedagents per study.
No patients reported any adverse events. However, wedid not assess renal function at a specific interval after theCTA, so further studies in Gd safety specifically for renaldysfunction are warranted. Our study showed that use ofGd for imaging of coronary vasculature was satisfactoryfor diagnostic purposes. More studies are warranted withMDCT scanners to further evaluate the quality of imageswith Gd-related contrast agents. The diagnostic efficacyof CTA can be confirmed by comparing the results ofCTA with conventional cardiac catheterization. Becausee-speed and 64 MDCT have equal enhancement of differ-ent doses of Gd and contrast, the lower contrast require-ments of 64-MDCT should improve image quality evenfurther on coronary CTA.
REFERENCES
1. Chau A, Gopal A, Mao S, et al. Comparison of three generations ofelectron beam tomography on image noise and reproducibility, a phan-tom study. Invest Radiol 2006; 41: In press.
2. Hou SH, Bushinsky DA, Wisch JB, et al. Hospital acquired renalinsufficiency: a prospective study. Am J Med 1983; 74:243–248.
3. Sesso R, Roque A, Vicioso B, et al. Prognosis of ARF in hospitalizedelderly patients. Am J Kidney Dis 2004; 44:410–419.
4. Goldenberg I, Matetzky S. Nephropathy induced by contrast media:pathogenesis, risk factors and preventive strategies [review]. CMAJ2005; 172:1461–1471. (Erratum in CMAJ 2005; 173:1210.)
5. Ose K, Doue T, Zen K, et al. ‘Gadolinium’ as an alternative to iodinatedcontrast media for X-ray angiography in patients with severe allergy.Circ J 2005; 69:507–509.
6. Hoffmann U, Fischereder M, Reil A, et al. Renal effects of gado-
pentetate dimeglumine in patients with normal and impaired renal func-tion. Eur J Med Res 2005; 10:149–154.
Academic Radiology, Vol 13, No 7, July 2006 NONINVASIVE GD-ENHANCED 3D CT CORONARY ANGIOGRAPHY
7. Sancak T, Bilgic S, Sanldilek U. Gadodiamide as an alternative contrastagent in intravenous digital subtraction angiography and interventionalprocedures of the upper extremity veins. Cardiovasc Intervent Radiol2002; 25:49–52.
8. Remy-Jardin M, Dequiedt P, Ertzbischoff O, et al. Safety and effective-ness of gadolinium-enhanced multi-detector row spiral CT angiographyof the chest: preliminary results in 37 patients with contraindications toiodinated contrast agents. Radiology 2005; 235:819–826.
9. Pena CS, Kaufman JA, Geller SC, et al. Gadopentetate dimeglumine: apossible alternative contrast agent for CT angiography of the aorta.J Comput Assist Tomogr 1999; 23:23–24.
10. Gupta AK, Alberico RA, Litwin A, et al. Gadopentetate dimeglumine ispotentially an alternative contrast agent for three-dimensional com-puted tomography angiography with multidetector-row helical scan-ning. J Comput Assist Tomogr 2002; 26:869–874.
11. Kim MD, Kim H. Gadolinium dimeglumine as a contrast agent for digi-tal subtraction angiography: in vitro Hounsfield unit measurement andclinical efficacy. Yonsei Med J 2003; 44:138–142.
12. Robert A, Bonneville F, Miralbes S, et al. Gadolinium in arteriographyand interventional radiology: 39 patients. J Radiol 2002; 83:1759–1763.
13. Hammer FD, Malaise J, Goffette PP, et al. Gadolinium dimeglumine: analternative contrast agent for digital subtraction angiography in patientswith renal failure. Transplant Proc 2000; 32:432–433.
14. Albrecht T, Dawson P. Gadolinium-DTPA as X-ray contrast medium inclinical studies. Br J Radiol 2000; 73:878–882.
15. Haustein J, Niendorf HP, Krestin G, et al. Renal tolerance of gadolinium-DTPA/dimeglumine in patients with chronic renal failure. Invest Radiol1992; 27:153–156.
16. Prince MR, Arnoldus C, Frisoli JK. Nephrotoxicity of high-dose gadolin-ium compared with iodinated contrast. J Magn Reson Imaging 1996;6:162–166.
17. Arsenault TM, King BF, Marsh JW Jr, et al. Systemic gadolinium toxic-ity in patients with renal insufficiency and renal failure: retrospectiveanalysis of an initial experience. Mayo Clin Proc 1996; 71:1150–1154.
18. Rieger J, Sitter T, Toepfer M, et al. Gadolinium as an alternative con-trast agent for diagnostic and interventional angiographic procedures in
patients with impaired renal function. Nephrol Dial Transplant 2002;17:824–828.
19. Joffe P, Thomsen HS, Meusel M. Pharmacokinetics of gadodiamideinjection in patients with severe renal insufficiency and patients under-going hemodialysis or continuous ambulatory peritoneal dialysis. AcadRadiol 1998; 5:491–502.
20. Normann PT, Joffe P, Martinsen I, et al. Quantification of gadodiamideas Gd in serum, peritoneal dialysate and faeces by inductively coupledplasma atomic emission spectroscopy and comparative analysis byhigh-performance liquid chromatography. J Pharm Biomed Anal 2000;22:939–947.
21. Katayama H, Yamaguchi K, Kozuka T, et al. Adverse reactions to ionicand nonionic contrast media. A report from the Japanese Committeeon the Safety of Contrast Media. Radiology. 1990; 175:621–628.
22. Niendorf HP, Dinger JC, Haustein J, et al. Tolerance data of Gd-DTPA:a review. Eur J Radiol 1991; 13:15–20.
23. Thomsen HS. Frequency of acute adverse events to a non-ionic low-osmolar contrast medium: the effect of verbal interview. PharmacolToxicol 1997; 80:108–110.
24. Sam AD 2nd, Morasch MD, Collins J, et al. Safety of gadolinium con-trast angiography in patients with chronic renal insufficiency. J VascSurg 2003; 38:313–318.
25. Gemery J, Idelson B, Reid S, et al. Acute renal failure after arteriogra-phy with a gadolinium-based contrast agent. AJR Am J Roentgenol1998; 171:1277–1278.
26. Gammete JJ, Forauer AR, Kazanjian S, et al. Safety of large volumeGadolinium angiography. J Vascular Interv Radiol 2001; 12(Suppl):S28.
27. Quin AD, O’Hare NJ, Wallis FJ, et al. Gd-DPTA: an alternative contrastmedium for CT. J Comput Assist Tomogr 1994; 18:634–636.
