- Email: [email protected]
Role of high resolution contrast-enhanced magnetic resonance angiography (HR CeMRA) in management of arterial complications of the renal transplant
European Journal of Radiology 79 (2011) e122–e127
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Role of high resolution contrast-enhanced magnetic resonance angiography (HR CeMRA) in management of arterial complications of the renal transplant M. Maged Ismaeel, Azza Abdel-Hamid ∗ Suez Canal University, Egypt
a r t i c l e
i n f o
Article history: Received 16 February 2011 Accepted 11 April 2011 Keywords: MRA Renal transplant Vascular complications Image guided intervention
a b s t r a c t Introduction: Transplant renal artery (RA) stenosis (TRAS) is the most frequent posttransplantation vascular complication. Contrast enhanced magnetic resonance (CeMRA) angiography has been established as the preferred imaging technique for the evaluation of TRAS because it does not require the use of iodinated contrast material and does not expose the patient to ionizing radiation. Digital subtraction angiography (DSA) is the gold standard in the evaluation of arterial tree of the renal allograft. Aim of the work: This study was carried out to assess the accuracy of CeMRA in the detection of arterial complications after renal transplantation. Patients and methods: Thirty renal transplant patients with suspected arterial complications in which both CeMRA and DSA were performed were included in the study. The HR CeMRA shows 93.7% sensitivity, 80% specificity, 88.2% positive predictive value, 88.9% negative predictive value and 88.5% accuracy. Conclusion: HR CeMRA is an accurate reliable tool in the assessment of arterial complications after renal transplantation. It may replace DSA as a diagnostic modality with reservation of interventional techniques for endovascular treatment of suitable cases. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Transplant renal artery (RA) stenosis (TRAS) is the most frequent posttransplantation vascular complication and may cause hypertension and allograft dysfunction. It occurs in up to 23% of allograft recipients. While the most frequent location is the site of the anastomosis, a stenosis may also appear at the iliac artery proximal to the anastomosis or at multiple locations within the RA [1]. Doppler ultrasonography (US) is routinely used as a reliable screening tool for TRAS [2] but the results are operator dependant. This is particularly true for the evaluation of tortuous vessels seen in transplant RAs, where kinks and curves may lead to the spurious elevations of US measured peak systolic velocity, raising the chances for mistaken suspicion of TRAS [3]. Contrast enhanced magnetic resonance (CeMRA) angiography has been established as the preferred imaging technique for the evaluation of TRAS because it does not require the use of iodinated contrast material and does not expose the patient to ionizing radiation [1]. Digital subtraction angiography (DSA) is the gold standard in the evaluation of arterial tree of the renal allograft [1]. This study was carried out in accordance with The Code of Ethics of the World Medical Association to assess the accuracy of CeMRA in
∗ Corresponding author. Tel.: +20 10 052 4927. E-mail address: [email protected] (A. Abdel-Hamid). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.04.039
the detection of arterial complications after renal transplantation. All authors disclose any financial and personal relationships with other people or organizations that could inappropriately influence this work. 2. Patients and methods Thirty renal transplantation patients referred to radiology unit from July 2006 to October 2009 with suspected arterial complications based on the presence of non-rejection-related renal failure; alterations identified on color Doppler US and/or hypertension of unknown origin or worsening hypertension. Both CeMRA and DSA were performed. The levels of serum creatinine were determined on the day of the conventional angiogram. MR angiography was performed prior to DSA in all cases. The average time between CeMRA and DSA was 7.66 days (range 0–30). 2.1. Examination techniques 2.1.1. Magnetic resonance angiography (MRA) All exams were done on 1.5 T machine (Avanto, Siemens Medical Solutions, Malvern, PA) using a phased array coil to cover the lower abdomen and pelvis in all three planes. Patients were given oxygen via a nasal cannula to aid in breath holding and a peripheral intravenous line was connected to an automatic power injector for the administration of contrast medium (gadolinium).
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
The following sequences were performed: 1. Localizer – breath hold 3 plane scout. 2. Three plane true Fast Imaging with Steady Precession (FISP) – cover lower abdomen and pelvis in all three planes. It is a single shot breath hold sequence used as a fast localizer of the allograft for anatomy reference. 3. 3D CeMRA pre contrast mask – coronal orientation over distal aorta, iliac arteries, and transplant kidney. The entire transplanted kidney was included. 4. 3D CeMRA post contrast – parameters of pre-contrast mask was copied. Arterial phase and two venous phases were captured with a 30 s delay between the arterial phase and each of the two venous phases. 20 ml of gadobenate dimeglumine was used (Multihance in 14 cases and Magnevist in 16 cases) and was injected with a 20 ml flush of normal saline at 2 ml/s. HR CeMRA parameters for 3 T were: TR 2.92 ms; TE 1.18 ms; 384 × 384 matrix; (GRAPPA) acceleration factor 3; flip angle 23◦ ; voxel size 1.0 mm × 1.0 mm × 1.2 mm; acquisition time 10 s; 20 ml of gadobenate dimeglumine (Multihance® ) was injected at a flow rate of 2 ml/s. HR CeMRA parameters for 1.5 T were: TR 2.61 ms; TE 1.09 ms; 448 × 448 matrix; (GRAPPA) acceleration factor 2; flip angle 15◦ ; voxel size 1.4 mm × 0.9 mm × 1.0 mm; acquisition time 36 s; 40 ml of gadobenate dimeglumine (Magnevist® ) or gadodiamide (Omniscan® ) was injected at a flow rate 2 of ml/s. 5. 3D T1 Volumetric Interpolated Breath-Hold Examination (VIBE) post contrast – using the following parameters (for 1.5 T we used; TR 4.35 ms, TE 1.65 ms, voxel size 2.1 mm × 1.6 mm × 2.5 mm, flip angle 12◦ and for 3 T we used; TR 3.3 ms, TE 1.17 ms, voxel size 1.7 mm × 1.4 mm × 2.0 mm, flip angle 13◦ ). The two post-contrast scans were each subtracted from the precontrast “mask” sequence, using the technique of subtraction of magnitudes of signal intensities. All precontrast (“mask”), postcontrast and subtracted sequences were available for review at an independent MR workstation. The subtracted images were subjected to 3D reformatting to generate MIP for all patients. 2.1.2. Conventional angiography All digital subtraction angiography (DSA) examinations were performed with an angiographic installation (Multi-Star; Siemens Medical Solutions, Forchheim, Germany). Intravenous hydration with a saline solution injected at a rate of 1 ml/kg/h was performed for 6 h prior to and following DSA in all patients. For visualization of the distal portion of the aorta and the pelvic vasculature, a power injector was used to inject 25 ml of iodinated contrast material (Imeron 300; Bracco, Milan, Italy) at a rate of 15 ml/s. The anastomosis and the transplant renal artery were visualized following manual injection of 8–10 ml of contrast material per imaging series. A total of 60–110 ml of contrast material were used per examination. In three patients, intraarterial pressure measurements were obtained proximal and distal to the stenotic lesions to confirm the presence of a hemodynamically significant stenosis. Fourteen patients underwent therapeutic conventional angiography; stents were performed in eleven patients (six using balloon expandable stents and five using self expandable stents) and balloon angioplasty was done in three patients with good therapeutic results initially and in the follow up scans. 3. Image analysis Axial, sagittal, and coronal maximum intensity projection (MIP) reconstructions, as well as a three-dimensional MIP reconstruction, were performed with a workstation (Kodak; Carestream Health, Rochester, New York, USA). Four predefined segments were evalu-
e123
Table 1 distribution of patients according to their clinical presentation. Clinical presentation
Number
Worsening hypertension (HTN) Change in renal function test (rising creatinine) Pain over the graft Decreased UOP Total
19 5 3 3 30
ated: common iliac artery; external iliac artery before anastomosis; site of anastomosis and transplant renal artery (TRA). The presence and degree of stenosis was determined independently by two radiologists (with 3 and 9 years experience with MR angiography) who were blinded to the results of conventional DSA. Both readers had full information about the type of surgery, the number of accessory arteries, and the location of the anastomosis. In cases of divergent interpretations, an additional reading was undertaken by these two radiologists in consensus. The extent of transplant renal artery stenosis (TRAS) was graded as follows: grade 0 indicated patency of the vessel with no detected stenosis; grade 1, less than 50% stenosis; grade 2, ≥50% stenosis and grade 3, total vessel occlusion. Grades 2–3 were considered as relevant stenoses. The percentage of stenosis was assessed at the optimal projection angle along the vessel axis with a digital caliper. The degree of TRAS was calculated as [1 (S/R)] × 100, where S and R are the diameters of the stenosis and reference vessels, respectively. A normal-appearing portion of the transplant RA distal to the stenotic lesion was defined as a reference vessel. Analysis of image quality was evaluated on the same segments with the addition of another segment (segmental branches) on the basis of a four-point Likert scale: 1. Non-diagnostic (no signal within the vessel or not seen due to artifact). 2. Moderate (inhomogeneous signal within the vessel, no sharp vessel border delineation). 3. Good (homogenous signal within the vessel with slight flow artifacts, almost complete and sharp delineation of vessel border). 4. Excellent (completely homogenous signal within the vessel lumen without flow artifacts, sharp and complete delineation of vessel border). 4. Results We retrospectively analyzed 30 episodes of allograft dysfunction in 30 patients including 12 women and 18 men with average age 48.2 (range 11–71) all cadaver-donor kidney recipients. All patients had clinical suspicion of transplant renal artery stenosis based on: the presence of non-rejection-related renal failure, alterations identified on color Doppler US and/or hypertension of unknown origin. Some of these patients were undergoing a series of follow-up studies owing to difficult clinical situations. All patients provided their informed consent before the examination. A stenosis ≥50% was considered a positive test. The average time between CeMRA and DSA was 7.66 days (range 0–30). Most patients presented by worsening hypertension (63.3%) (Table 1). At the level of CIA, CeMRA overestimates the degree of stenosis in two patients (3/1) while it underestimates the degree of stenosis in two patients (2/4). One patient had complete occlusion due to arterial anastomotic thrombosis (Fig. 1). Table 2 showed that (25) patients were proved to have relevant (≥50%) stenosis, with pre-set four arterial segments, so 100 arterial segments were available for comparison. MRA grades the stenosis in accordance with DSA in 89 arterial segments distributed
e124
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127 Table 2 Comparison of MRA and DSA in the assessment of the degree of arterial stenosis. MRA grading
0 1 2 3 Total
Fig. 1. Anastomotic arterial thrombosis. CeMRA shows abrupt occlusion of the transplant renal artery at the site of anastomosis with the Rt. External iliac artery (arrow).
DSA grading Grade 0
Grade 1
Grade 2
Grade 3
Total
75 3 3 0 81
0 2 2 0 4
0 3 11 0 14
0 0 0 1 1
75 8 16 1
as follows: grade 0 in 75 segments, grade 1 in two segments, grade 2 in 11 segments and grade 3 in one segment. Fourteen patients underwent therapeutic conventional angiography. Three patients underwent percutaneous transluminal angioplasty (PTA) with satisfactory results initially (Fig. 2), except one case who was found to have anastomotic restenosis upon follow-up CeMRA 3 months later with repeat PTA gives good results on follow up study. Eleven patients underwent percutaneous transluminal angioplasty and stenting (PTAS) who did not respond initially to PTA, six of them were done using balloon expandable stents and five were done using self expandable stents with satisfactory post PTAS angiographic results initially and in follow up exams (Fig. 3). The degree of stenosis was overestimated in two patients (Fig. 4) and underestimated in one patient by MRA (Fig. 5). DSA with
Fig. 2. Successful PTA of significant mid-third TRA stenosis. (a) CeMRA shows mild anastomotic stenosis of the TRA with significant (90%) stenosis of the mid-third TRA. (b) DSA clearly demonstrates the presence of stenoses. (c) PTA was performed using 6.2 mm Viatrack balloon. (d) Post PTA shows no evidence of residual stenosis.
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
e125
Fig. 3. Successful treatment of significant anastomotic stenosis by PTAS. (a) Color and spectral Doppler US of the interlobar branches of the TRA show tardus parvus waveform suggesting proximal stenosis of TRA. (b) CeMRA shows significant focal stenosis at the origin of the transplant renal artery anastomosis with the left external iliac artery. (c) DSA confirms the presence of significant anastomotic stenosis. (d) Balloon angioplasty was performed using a 6× 20 mm and 7× 15 mm balloon. Post PTA angiogram showed residual stenosis. (e) PTAS with pressure measurements were performed simultaneously showed a systolic gradient of 52 mm Hg. A balloon expandable genesis stent 7× 12 mm was deployed at the anastomosis site. PTA was then performed using a 8× 20 mm balloon. Completion angiogram showed no residual stenosis with improved flow to the kidney. Pressure measurements across the anastomosis showed a systolic gradient of 8 mm Hg, and mean gradient of only 1 mm Hg.
pressure gradient measurement was performed that accurately determine the degree of stenosis in such cases. The HR CeMRA shows 93.7% sensitivity, 80% specificity, 88.5% accuracy, 88.2% positive predictive value and 88.9% negative predictive value. The mean image quality was 3.96 ± 0.19 (standard deviation), 3.88 ± 0.58, 3.73 ± 0.72 and 3.79 ± 0.71 (for segments 1–4), respectively.
5. Discussion Although color Doppler imaging has been claimed to be useful for the detection of renal allograft artery stenosis, it is not possible to determine the degree of stenosis. Both contrast enhanced MRA (CeMRA) and intraarterial digital subtraction angiography (IA DSA) yield good detection of vascular insult to the renal allograft. However, even within selective arteriography using multiple pro-
Fig. 4. MRA overestimation of the degree of stenosis. (a) MRA shows at least moderate stenosis at the origin of the transplant renal artery (arrows) which originates from the left external iliac artery. (b) DSA shows widely patent TRA (arrows) with no evidence of angiographically significant anastomotic stenosis or inflow iliac artery stenosis.
e126
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
Fig. 5. MRA underestimation of significant arterial stenoses. (a) MRA shows mild narrowing of right common iliac artery (CIA) by an eccentric atherosclerotic plaque (arrow) and mild narrowing of the right external iliac artery (EIA) just proximal to the TRA anastomosis (arrowhead), both felt not to be hemodynamically significant. (b) DSA shows focal area of stenosis in the proximal right CIA (arrow) and another focal area of stenosis right EIA (arrowhead) just proximal to a widely patent TRA. Measurement of pressure gradient confirms that all were hemodynamically significant (systolic pressure gradient of 30–40 mm Hg). (c) PTA was performed and none of the stenoses responded to initial PTA. (d) Successful placement of overlapping Genesis stents in the right common and external iliac arteries, post dilated to 8 and 6 mm, respectively, was performed. Post procedure pressure gradient was performed and shows systolic gradient of (3–4 mm Hg).
jections, eccentric stenosis may be overlooked in an IA DSA study. The 3D nature of the CeMRA images overcomes this limitation by permitting interactive reformatting to find the optimal obliquity or angulations to demonstrate the eccentric arterial lesions. The interactive reformatting capability of CeMRA may also unfold an acutely angulated transplant renal artery by providing obliquity and perpendicular views of the anastomosis between the external iliac artery and allograft renal artery, an area where stenosis frequently occurs [4] (application of Gd MRA 2003). Renal artery stenosis (RAS) may occur due to atherosclerotic disease or surgical trauma [5,6]. This manifests as hypertension (as most of our patients) or loss of function in the absence of rejection, obstruction, or infection and accounts for more than 75% of all vascular complications [7]. Typically it occurs at the site of anastomosis that connects the transplant artery to the native iliac artery [3], which is in accordance to our results found anastomotic stenosis in 46% of cases followed by the inflow vessels (common iliac artery 23%) and then comes the external iliac artery proximal to the site of anastomosis and the TRA (both were 15.3%). MRA has been shown to be more sensitive than
US in the detection of native RAS [8]. It can demonstrate perfusion in areas where Doppler US could not [9]. Thromboses (either arterial or venous) occur less commonly, our results show transplant renal artery thrombosis to occur in 3% which is in accordance to the expected figures ranging from 1 to 6% usually early in the postoperative period, seen on CeMRA as a global lack of perfusion, non visualization of the transplant renal artery or abrupt occlusion of the TRA [3,10]. Predisposing causes include poor surgical technique, compression by fluid collections, and hypovolemia [3]. Early diagnosis and treatment is essential to prevent premature loss of allograft [7,11]. The age of onset of TRAS can be variable and, although the outcome of untreated RAS is not known, balloon angioplasty or arterial stenting is the accepted treatment. These procedures, however, carry a significant complication profile. An accurate imaging modality, which can assess the transplant vascular system noninvasively, is therefore highly desirable [12] Nonenhanced assessment of transplant RAs has previously been performed with phase-contrast [13] and time-of-flight MR angiography [14]. Phase-contrast MR angiography is extremely time
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
consuming, as echo time and the velocity-encoded value need to be optimally adjusted to display the turbulent flow in stenotic lesions [1], whereas the CeMRA technique we used has an acquisition time of approximately 1 min 10 s–1 min 36 s. With time-of-flight MR angiography, peripheral segments of RAs are not usually visualized owing to saturation effects. These effects are more pronounced in tortuous vessels oriented perpendicularly to the acquisition plane. Owing to the location of renal allografts in the pelvic fossa, transplant arteries often show these artifacts [1]. CeMR angiography delivers a more consistent vessel signal, as its flow is compensated in all three spatial orientations because of their intrinsically balanced gradient structure. The grading of RA stenosis with MR angiography is challenging [15]. Overestimation of the degree of stenosis has been described for contrast-enhanced MR angiography [16] as well as nonenhanced MR angiography [17] as we found in our study where the degree of stenosis was overestimated in two patients proved by DSA with measurement of the pressure gradient across the suspected site of stenosis. DSA is considered as the diagnostic standard for TRAS [5]. Depending on posttransplantation anatomy, it may require imaging with multiple oblique angles to properly visualize the anastomosis and the kinking of transplant RAs [1]. Thus, given the limited number of views obtained, we cannot exclude that the anastomosed region and kink stenosis were not fully visualized and that the degree of stenosis was therefore not determined correctly at DSA. To overcome this, pressure gradient across the suspected stenotic region was done which helped identifying significant stenosis in two patients with non significant stenosis at their DSA images. Several investigators report a high diagnostic accuracy of 3D gadolinium-enhanced MRA for the detection of renal artery stenosis, using conventional angiography as the standard of reference, with sensitivities and specificities ranging from 88% to 100% and 71% to 98%, respectively [18]. Our study shows CeMRA to have a sensitivity of 93.7%, 80% specificity, 88.5% accuracy, 88.2% positive predictive value and 88.9% negative predictive value. Our study had some limitations. The number of patients included in this study was relatively small. In our study, only transplant recipients scheduled to undergo DSA were included. However, in view of its invasive character and the need for potentially nephrotoxic iodinated contrast material, DSA is only performed after serious consideration of the risks and benefits to the patient. In conclusion, HR CeMRA is an accurate reliable tool in the assessment of arterial complications after renal transplantation and it can thus become an appropriate substitute for diagnostic angiography in the detection of transplant renal arterial stenosis that can accurately, rapidly and non-invasively select patients for therapeutic angiography. CeMRA performed earlier help reducing the dose of nephrotoxic iodinated contrast materials used in conventional angiography and it should be considered the first choice
e127
modality for the evaluation of transplant recipients with suspected arterial complications. The information from CeMRA studies can significantly increase the diagnostic certainty, alter physicians’ initial diagnoses and immediate clinical management, and can help avoid invasive procedures [19]. In some circumstances, when there is discordance between the imaging diagnoses and the clinical scenario, the measurement of pressure gradient across the stenotic region is of significant clinical and diagnostic importance and should not be neglected. Conflict of interest statement None. References [1] Lanzman RS, Voiculescu A, Walther C, et al. ECG-gated nonenhanced 3D steady-state free precession MR angiography in assessment of transplant renal arteries: comparison with DSA. Radiology 2009;252:914–21. [2] Voiculescu A, Schmitz M, Hollenbeck M, et al. Management of arterial stenosis affecting kidney graft perfusion: a single-center study in 53 patients. Am J Transplant 2005;5:1731–8. [3] Browne RF, Tuite DJ. Imaging of the renal transplant: comparison of MRI with duplex sonography. Abdom Imaging 2006;31:461–82. [4] Ng KK, Cheng YF, Lui KW, et al. Application of GD-enhanced renal allograft MR angiography for evaluation of posttransplantation complications. Transplant Proc 2003;35:307–8. [5] Patel NH, Jindal RM, Wilkin T, et al. Renal arterial stenosis in renal allografts: retrospective study of predisposing factors and outcome after percutaneous transluminal angioplasty. Radiology 2001;219:663–7. [6] Hohenwalter MD, Skowlund CJ, Erickson SJ, et al. transplant evaluation with MR angiography and MR imaging. Radiographics 2001;21(November–December (6)):1505–17. [7] Sandhu C, Patel U. Renal transplantation dysfunction: the role of interventional radiology. Clin Radiol 2002;57:772–83. [8] Zhang H, Prince MR. Renal MR angiography. Magn Reson Imaging Clin North Am 2004;12:487–503. [9] Jakobsen JA, Brabrand K, Egge TS, Hartmann A. Doppler examination of the allografted kidney. Acta Radiol 2003;44:3–12. [10] Tarzamni MK, Argani H, Nurifar M, Nezami N. Vascular complication and Doppler ultrasonographic finding after renal transplantation. Transplant Proc 2007;39:1098–102. [11] Baxter GM. Imaging in renal transplantation. Ultrasound Q 2003;19:123–38. [12] Patel U, Khaw KK, Hughes NC. Doppler ultrasound for detection of renal transplant artery stenosis – threshold peak systolic velocity needs to be higher in low-risk or surveillance population. Clin Radiol 2003;58:772–7. [13] Gedroyc WM, Negus R, al-Kutoubi A, Palmer A, Taube D, Hulme B. Magnetic resonance angiography of renal transplants. Lancet 1992;339:789–91. [14] Johnson DB, Lerner CA, Prince MR, et al. Gadolinium-enhanced magnetic resonance angiography of renal transplants. Magn Reson Imaging 1997;15:13–20. [15] Schoenberg SO, Rieger J, Weber CH, et al. High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 2005;235:687–98. [16] Mallouhi A, Schocke M, Judmaier W, et al. 3D MR angiography of renal arteries: comparison of volume rendering and maximum intensity projection algorithms. Radiology 2002;223:509–16. [17] Coenegrachts KL, Hoogeveen RM, Vaninbroukx JA, et al. High-spatial-resolution 3D balanced turbo field-echo technique for MR angiography of the renal arteries: initial experience. Radiology 2004;231:237–42. [18] Spinosa DJ, Isaacs RB, Matsumoto AH, Angle JF, Hagspiel KD, Leung DA. Angiographic evaluation and treatment of transplant renal artery stenosis. Curr Opin Urol 2001;11:197–205. [19] Jain R, Sawhney S. Contrast-enhanced MR angiography (CE-MRA) in the evaluation of vascular complications of renal transplantation. Clinical Radiology 2005;60(11):1171–81.
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Role of high resolution contrast-enhanced magnetic resonance angiography (HR CeMRA) in management of arterial complications of the renal transplant M. Maged Ismaeel, Azza Abdel-Hamid ∗ Suez Canal University, Egypt
a r t i c l e
i n f o
Article history: Received 16 February 2011 Accepted 11 April 2011 Keywords: MRA Renal transplant Vascular complications Image guided intervention
a b s t r a c t Introduction: Transplant renal artery (RA) stenosis (TRAS) is the most frequent posttransplantation vascular complication. Contrast enhanced magnetic resonance (CeMRA) angiography has been established as the preferred imaging technique for the evaluation of TRAS because it does not require the use of iodinated contrast material and does not expose the patient to ionizing radiation. Digital subtraction angiography (DSA) is the gold standard in the evaluation of arterial tree of the renal allograft. Aim of the work: This study was carried out to assess the accuracy of CeMRA in the detection of arterial complications after renal transplantation. Patients and methods: Thirty renal transplant patients with suspected arterial complications in which both CeMRA and DSA were performed were included in the study. The HR CeMRA shows 93.7% sensitivity, 80% specificity, 88.2% positive predictive value, 88.9% negative predictive value and 88.5% accuracy. Conclusion: HR CeMRA is an accurate reliable tool in the assessment of arterial complications after renal transplantation. It may replace DSA as a diagnostic modality with reservation of interventional techniques for endovascular treatment of suitable cases. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Transplant renal artery (RA) stenosis (TRAS) is the most frequent posttransplantation vascular complication and may cause hypertension and allograft dysfunction. It occurs in up to 23% of allograft recipients. While the most frequent location is the site of the anastomosis, a stenosis may also appear at the iliac artery proximal to the anastomosis or at multiple locations within the RA [1]. Doppler ultrasonography (US) is routinely used as a reliable screening tool for TRAS [2] but the results are operator dependant. This is particularly true for the evaluation of tortuous vessels seen in transplant RAs, where kinks and curves may lead to the spurious elevations of US measured peak systolic velocity, raising the chances for mistaken suspicion of TRAS [3]. Contrast enhanced magnetic resonance (CeMRA) angiography has been established as the preferred imaging technique for the evaluation of TRAS because it does not require the use of iodinated contrast material and does not expose the patient to ionizing radiation [1]. Digital subtraction angiography (DSA) is the gold standard in the evaluation of arterial tree of the renal allograft [1]. This study was carried out in accordance with The Code of Ethics of the World Medical Association to assess the accuracy of CeMRA in
∗ Corresponding author. Tel.: +20 10 052 4927. E-mail address: [email protected] (A. Abdel-Hamid). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.04.039
the detection of arterial complications after renal transplantation. All authors disclose any financial and personal relationships with other people or organizations that could inappropriately influence this work. 2. Patients and methods Thirty renal transplantation patients referred to radiology unit from July 2006 to October 2009 with suspected arterial complications based on the presence of non-rejection-related renal failure; alterations identified on color Doppler US and/or hypertension of unknown origin or worsening hypertension. Both CeMRA and DSA were performed. The levels of serum creatinine were determined on the day of the conventional angiogram. MR angiography was performed prior to DSA in all cases. The average time between CeMRA and DSA was 7.66 days (range 0–30). 2.1. Examination techniques 2.1.1. Magnetic resonance angiography (MRA) All exams were done on 1.5 T machine (Avanto, Siemens Medical Solutions, Malvern, PA) using a phased array coil to cover the lower abdomen and pelvis in all three planes. Patients were given oxygen via a nasal cannula to aid in breath holding and a peripheral intravenous line was connected to an automatic power injector for the administration of contrast medium (gadolinium).
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
The following sequences were performed: 1. Localizer – breath hold 3 plane scout. 2. Three plane true Fast Imaging with Steady Precession (FISP) – cover lower abdomen and pelvis in all three planes. It is a single shot breath hold sequence used as a fast localizer of the allograft for anatomy reference. 3. 3D CeMRA pre contrast mask – coronal orientation over distal aorta, iliac arteries, and transplant kidney. The entire transplanted kidney was included. 4. 3D CeMRA post contrast – parameters of pre-contrast mask was copied. Arterial phase and two venous phases were captured with a 30 s delay between the arterial phase and each of the two venous phases. 20 ml of gadobenate dimeglumine was used (Multihance in 14 cases and Magnevist in 16 cases) and was injected with a 20 ml flush of normal saline at 2 ml/s. HR CeMRA parameters for 3 T were: TR 2.92 ms; TE 1.18 ms; 384 × 384 matrix; (GRAPPA) acceleration factor 3; flip angle 23◦ ; voxel size 1.0 mm × 1.0 mm × 1.2 mm; acquisition time 10 s; 20 ml of gadobenate dimeglumine (Multihance® ) was injected at a flow rate of 2 ml/s. HR CeMRA parameters for 1.5 T were: TR 2.61 ms; TE 1.09 ms; 448 × 448 matrix; (GRAPPA) acceleration factor 2; flip angle 15◦ ; voxel size 1.4 mm × 0.9 mm × 1.0 mm; acquisition time 36 s; 40 ml of gadobenate dimeglumine (Magnevist® ) or gadodiamide (Omniscan® ) was injected at a flow rate 2 of ml/s. 5. 3D T1 Volumetric Interpolated Breath-Hold Examination (VIBE) post contrast – using the following parameters (for 1.5 T we used; TR 4.35 ms, TE 1.65 ms, voxel size 2.1 mm × 1.6 mm × 2.5 mm, flip angle 12◦ and for 3 T we used; TR 3.3 ms, TE 1.17 ms, voxel size 1.7 mm × 1.4 mm × 2.0 mm, flip angle 13◦ ). The two post-contrast scans were each subtracted from the precontrast “mask” sequence, using the technique of subtraction of magnitudes of signal intensities. All precontrast (“mask”), postcontrast and subtracted sequences were available for review at an independent MR workstation. The subtracted images were subjected to 3D reformatting to generate MIP for all patients. 2.1.2. Conventional angiography All digital subtraction angiography (DSA) examinations were performed with an angiographic installation (Multi-Star; Siemens Medical Solutions, Forchheim, Germany). Intravenous hydration with a saline solution injected at a rate of 1 ml/kg/h was performed for 6 h prior to and following DSA in all patients. For visualization of the distal portion of the aorta and the pelvic vasculature, a power injector was used to inject 25 ml of iodinated contrast material (Imeron 300; Bracco, Milan, Italy) at a rate of 15 ml/s. The anastomosis and the transplant renal artery were visualized following manual injection of 8–10 ml of contrast material per imaging series. A total of 60–110 ml of contrast material were used per examination. In three patients, intraarterial pressure measurements were obtained proximal and distal to the stenotic lesions to confirm the presence of a hemodynamically significant stenosis. Fourteen patients underwent therapeutic conventional angiography; stents were performed in eleven patients (six using balloon expandable stents and five using self expandable stents) and balloon angioplasty was done in three patients with good therapeutic results initially and in the follow up scans. 3. Image analysis Axial, sagittal, and coronal maximum intensity projection (MIP) reconstructions, as well as a three-dimensional MIP reconstruction, were performed with a workstation (Kodak; Carestream Health, Rochester, New York, USA). Four predefined segments were evalu-
e123
Table 1 distribution of patients according to their clinical presentation. Clinical presentation
Number
Worsening hypertension (HTN) Change in renal function test (rising creatinine) Pain over the graft Decreased UOP Total
19 5 3 3 30
ated: common iliac artery; external iliac artery before anastomosis; site of anastomosis and transplant renal artery (TRA). The presence and degree of stenosis was determined independently by two radiologists (with 3 and 9 years experience with MR angiography) who were blinded to the results of conventional DSA. Both readers had full information about the type of surgery, the number of accessory arteries, and the location of the anastomosis. In cases of divergent interpretations, an additional reading was undertaken by these two radiologists in consensus. The extent of transplant renal artery stenosis (TRAS) was graded as follows: grade 0 indicated patency of the vessel with no detected stenosis; grade 1, less than 50% stenosis; grade 2, ≥50% stenosis and grade 3, total vessel occlusion. Grades 2–3 were considered as relevant stenoses. The percentage of stenosis was assessed at the optimal projection angle along the vessel axis with a digital caliper. The degree of TRAS was calculated as [1 (S/R)] × 100, where S and R are the diameters of the stenosis and reference vessels, respectively. A normal-appearing portion of the transplant RA distal to the stenotic lesion was defined as a reference vessel. Analysis of image quality was evaluated on the same segments with the addition of another segment (segmental branches) on the basis of a four-point Likert scale: 1. Non-diagnostic (no signal within the vessel or not seen due to artifact). 2. Moderate (inhomogeneous signal within the vessel, no sharp vessel border delineation). 3. Good (homogenous signal within the vessel with slight flow artifacts, almost complete and sharp delineation of vessel border). 4. Excellent (completely homogenous signal within the vessel lumen without flow artifacts, sharp and complete delineation of vessel border). 4. Results We retrospectively analyzed 30 episodes of allograft dysfunction in 30 patients including 12 women and 18 men with average age 48.2 (range 11–71) all cadaver-donor kidney recipients. All patients had clinical suspicion of transplant renal artery stenosis based on: the presence of non-rejection-related renal failure, alterations identified on color Doppler US and/or hypertension of unknown origin. Some of these patients were undergoing a series of follow-up studies owing to difficult clinical situations. All patients provided their informed consent before the examination. A stenosis ≥50% was considered a positive test. The average time between CeMRA and DSA was 7.66 days (range 0–30). Most patients presented by worsening hypertension (63.3%) (Table 1). At the level of CIA, CeMRA overestimates the degree of stenosis in two patients (3/1) while it underestimates the degree of stenosis in two patients (2/4). One patient had complete occlusion due to arterial anastomotic thrombosis (Fig. 1). Table 2 showed that (25) patients were proved to have relevant (≥50%) stenosis, with pre-set four arterial segments, so 100 arterial segments were available for comparison. MRA grades the stenosis in accordance with DSA in 89 arterial segments distributed
e124
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127 Table 2 Comparison of MRA and DSA in the assessment of the degree of arterial stenosis. MRA grading
0 1 2 3 Total
Fig. 1. Anastomotic arterial thrombosis. CeMRA shows abrupt occlusion of the transplant renal artery at the site of anastomosis with the Rt. External iliac artery (arrow).
DSA grading Grade 0
Grade 1
Grade 2
Grade 3
Total
75 3 3 0 81
0 2 2 0 4
0 3 11 0 14
0 0 0 1 1
75 8 16 1
as follows: grade 0 in 75 segments, grade 1 in two segments, grade 2 in 11 segments and grade 3 in one segment. Fourteen patients underwent therapeutic conventional angiography. Three patients underwent percutaneous transluminal angioplasty (PTA) with satisfactory results initially (Fig. 2), except one case who was found to have anastomotic restenosis upon follow-up CeMRA 3 months later with repeat PTA gives good results on follow up study. Eleven patients underwent percutaneous transluminal angioplasty and stenting (PTAS) who did not respond initially to PTA, six of them were done using balloon expandable stents and five were done using self expandable stents with satisfactory post PTAS angiographic results initially and in follow up exams (Fig. 3). The degree of stenosis was overestimated in two patients (Fig. 4) and underestimated in one patient by MRA (Fig. 5). DSA with
Fig. 2. Successful PTA of significant mid-third TRA stenosis. (a) CeMRA shows mild anastomotic stenosis of the TRA with significant (90%) stenosis of the mid-third TRA. (b) DSA clearly demonstrates the presence of stenoses. (c) PTA was performed using 6.2 mm Viatrack balloon. (d) Post PTA shows no evidence of residual stenosis.
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
e125
Fig. 3. Successful treatment of significant anastomotic stenosis by PTAS. (a) Color and spectral Doppler US of the interlobar branches of the TRA show tardus parvus waveform suggesting proximal stenosis of TRA. (b) CeMRA shows significant focal stenosis at the origin of the transplant renal artery anastomosis with the left external iliac artery. (c) DSA confirms the presence of significant anastomotic stenosis. (d) Balloon angioplasty was performed using a 6× 20 mm and 7× 15 mm balloon. Post PTA angiogram showed residual stenosis. (e) PTAS with pressure measurements were performed simultaneously showed a systolic gradient of 52 mm Hg. A balloon expandable genesis stent 7× 12 mm was deployed at the anastomosis site. PTA was then performed using a 8× 20 mm balloon. Completion angiogram showed no residual stenosis with improved flow to the kidney. Pressure measurements across the anastomosis showed a systolic gradient of 8 mm Hg, and mean gradient of only 1 mm Hg.
pressure gradient measurement was performed that accurately determine the degree of stenosis in such cases. The HR CeMRA shows 93.7% sensitivity, 80% specificity, 88.5% accuracy, 88.2% positive predictive value and 88.9% negative predictive value. The mean image quality was 3.96 ± 0.19 (standard deviation), 3.88 ± 0.58, 3.73 ± 0.72 and 3.79 ± 0.71 (for segments 1–4), respectively.
5. Discussion Although color Doppler imaging has been claimed to be useful for the detection of renal allograft artery stenosis, it is not possible to determine the degree of stenosis. Both contrast enhanced MRA (CeMRA) and intraarterial digital subtraction angiography (IA DSA) yield good detection of vascular insult to the renal allograft. However, even within selective arteriography using multiple pro-
Fig. 4. MRA overestimation of the degree of stenosis. (a) MRA shows at least moderate stenosis at the origin of the transplant renal artery (arrows) which originates from the left external iliac artery. (b) DSA shows widely patent TRA (arrows) with no evidence of angiographically significant anastomotic stenosis or inflow iliac artery stenosis.
e126
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
Fig. 5. MRA underestimation of significant arterial stenoses. (a) MRA shows mild narrowing of right common iliac artery (CIA) by an eccentric atherosclerotic plaque (arrow) and mild narrowing of the right external iliac artery (EIA) just proximal to the TRA anastomosis (arrowhead), both felt not to be hemodynamically significant. (b) DSA shows focal area of stenosis in the proximal right CIA (arrow) and another focal area of stenosis right EIA (arrowhead) just proximal to a widely patent TRA. Measurement of pressure gradient confirms that all were hemodynamically significant (systolic pressure gradient of 30–40 mm Hg). (c) PTA was performed and none of the stenoses responded to initial PTA. (d) Successful placement of overlapping Genesis stents in the right common and external iliac arteries, post dilated to 8 and 6 mm, respectively, was performed. Post procedure pressure gradient was performed and shows systolic gradient of (3–4 mm Hg).
jections, eccentric stenosis may be overlooked in an IA DSA study. The 3D nature of the CeMRA images overcomes this limitation by permitting interactive reformatting to find the optimal obliquity or angulations to demonstrate the eccentric arterial lesions. The interactive reformatting capability of CeMRA may also unfold an acutely angulated transplant renal artery by providing obliquity and perpendicular views of the anastomosis between the external iliac artery and allograft renal artery, an area where stenosis frequently occurs [4] (application of Gd MRA 2003). Renal artery stenosis (RAS) may occur due to atherosclerotic disease or surgical trauma [5,6]. This manifests as hypertension (as most of our patients) or loss of function in the absence of rejection, obstruction, or infection and accounts for more than 75% of all vascular complications [7]. Typically it occurs at the site of anastomosis that connects the transplant artery to the native iliac artery [3], which is in accordance to our results found anastomotic stenosis in 46% of cases followed by the inflow vessels (common iliac artery 23%) and then comes the external iliac artery proximal to the site of anastomosis and the TRA (both were 15.3%). MRA has been shown to be more sensitive than
US in the detection of native RAS [8]. It can demonstrate perfusion in areas where Doppler US could not [9]. Thromboses (either arterial or venous) occur less commonly, our results show transplant renal artery thrombosis to occur in 3% which is in accordance to the expected figures ranging from 1 to 6% usually early in the postoperative period, seen on CeMRA as a global lack of perfusion, non visualization of the transplant renal artery or abrupt occlusion of the TRA [3,10]. Predisposing causes include poor surgical technique, compression by fluid collections, and hypovolemia [3]. Early diagnosis and treatment is essential to prevent premature loss of allograft [7,11]. The age of onset of TRAS can be variable and, although the outcome of untreated RAS is not known, balloon angioplasty or arterial stenting is the accepted treatment. These procedures, however, carry a significant complication profile. An accurate imaging modality, which can assess the transplant vascular system noninvasively, is therefore highly desirable [12] Nonenhanced assessment of transplant RAs has previously been performed with phase-contrast [13] and time-of-flight MR angiography [14]. Phase-contrast MR angiography is extremely time
M.M. Ismaeel, A. Abdel-Hamid / European Journal of Radiology 79 (2011) e122–e127
consuming, as echo time and the velocity-encoded value need to be optimally adjusted to display the turbulent flow in stenotic lesions [1], whereas the CeMRA technique we used has an acquisition time of approximately 1 min 10 s–1 min 36 s. With time-of-flight MR angiography, peripheral segments of RAs are not usually visualized owing to saturation effects. These effects are more pronounced in tortuous vessels oriented perpendicularly to the acquisition plane. Owing to the location of renal allografts in the pelvic fossa, transplant arteries often show these artifacts [1]. CeMR angiography delivers a more consistent vessel signal, as its flow is compensated in all three spatial orientations because of their intrinsically balanced gradient structure. The grading of RA stenosis with MR angiography is challenging [15]. Overestimation of the degree of stenosis has been described for contrast-enhanced MR angiography [16] as well as nonenhanced MR angiography [17] as we found in our study where the degree of stenosis was overestimated in two patients proved by DSA with measurement of the pressure gradient across the suspected site of stenosis. DSA is considered as the diagnostic standard for TRAS [5]. Depending on posttransplantation anatomy, it may require imaging with multiple oblique angles to properly visualize the anastomosis and the kinking of transplant RAs [1]. Thus, given the limited number of views obtained, we cannot exclude that the anastomosed region and kink stenosis were not fully visualized and that the degree of stenosis was therefore not determined correctly at DSA. To overcome this, pressure gradient across the suspected stenotic region was done which helped identifying significant stenosis in two patients with non significant stenosis at their DSA images. Several investigators report a high diagnostic accuracy of 3D gadolinium-enhanced MRA for the detection of renal artery stenosis, using conventional angiography as the standard of reference, with sensitivities and specificities ranging from 88% to 100% and 71% to 98%, respectively [18]. Our study shows CeMRA to have a sensitivity of 93.7%, 80% specificity, 88.5% accuracy, 88.2% positive predictive value and 88.9% negative predictive value. Our study had some limitations. The number of patients included in this study was relatively small. In our study, only transplant recipients scheduled to undergo DSA were included. However, in view of its invasive character and the need for potentially nephrotoxic iodinated contrast material, DSA is only performed after serious consideration of the risks and benefits to the patient. In conclusion, HR CeMRA is an accurate reliable tool in the assessment of arterial complications after renal transplantation and it can thus become an appropriate substitute for diagnostic angiography in the detection of transplant renal arterial stenosis that can accurately, rapidly and non-invasively select patients for therapeutic angiography. CeMRA performed earlier help reducing the dose of nephrotoxic iodinated contrast materials used in conventional angiography and it should be considered the first choice
e127
modality for the evaluation of transplant recipients with suspected arterial complications. The information from CeMRA studies can significantly increase the diagnostic certainty, alter physicians’ initial diagnoses and immediate clinical management, and can help avoid invasive procedures [19]. In some circumstances, when there is discordance between the imaging diagnoses and the clinical scenario, the measurement of pressure gradient across the stenotic region is of significant clinical and diagnostic importance and should not be neglected. Conflict of interest statement None. References [1] Lanzman RS, Voiculescu A, Walther C, et al. ECG-gated nonenhanced 3D steady-state free precession MR angiography in assessment of transplant renal arteries: comparison with DSA. Radiology 2009;252:914–21. [2] Voiculescu A, Schmitz M, Hollenbeck M, et al. Management of arterial stenosis affecting kidney graft perfusion: a single-center study in 53 patients. Am J Transplant 2005;5:1731–8. [3] Browne RF, Tuite DJ. Imaging of the renal transplant: comparison of MRI with duplex sonography. Abdom Imaging 2006;31:461–82. [4] Ng KK, Cheng YF, Lui KW, et al. Application of GD-enhanced renal allograft MR angiography for evaluation of posttransplantation complications. Transplant Proc 2003;35:307–8. [5] Patel NH, Jindal RM, Wilkin T, et al. Renal arterial stenosis in renal allografts: retrospective study of predisposing factors and outcome after percutaneous transluminal angioplasty. Radiology 2001;219:663–7. [6] Hohenwalter MD, Skowlund CJ, Erickson SJ, et al. transplant evaluation with MR angiography and MR imaging. Radiographics 2001;21(November–December (6)):1505–17. [7] Sandhu C, Patel U. Renal transplantation dysfunction: the role of interventional radiology. Clin Radiol 2002;57:772–83. [8] Zhang H, Prince MR. Renal MR angiography. Magn Reson Imaging Clin North Am 2004;12:487–503. [9] Jakobsen JA, Brabrand K, Egge TS, Hartmann A. Doppler examination of the allografted kidney. Acta Radiol 2003;44:3–12. [10] Tarzamni MK, Argani H, Nurifar M, Nezami N. Vascular complication and Doppler ultrasonographic finding after renal transplantation. Transplant Proc 2007;39:1098–102. [11] Baxter GM. Imaging in renal transplantation. Ultrasound Q 2003;19:123–38. [12] Patel U, Khaw KK, Hughes NC. Doppler ultrasound for detection of renal transplant artery stenosis – threshold peak systolic velocity needs to be higher in low-risk or surveillance population. Clin Radiol 2003;58:772–7. [13] Gedroyc WM, Negus R, al-Kutoubi A, Palmer A, Taube D, Hulme B. Magnetic resonance angiography of renal transplants. Lancet 1992;339:789–91. [14] Johnson DB, Lerner CA, Prince MR, et al. Gadolinium-enhanced magnetic resonance angiography of renal transplants. Magn Reson Imaging 1997;15:13–20. [15] Schoenberg SO, Rieger J, Weber CH, et al. High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 2005;235:687–98. [16] Mallouhi A, Schocke M, Judmaier W, et al. 3D MR angiography of renal arteries: comparison of volume rendering and maximum intensity projection algorithms. Radiology 2002;223:509–16. [17] Coenegrachts KL, Hoogeveen RM, Vaninbroukx JA, et al. High-spatial-resolution 3D balanced turbo field-echo technique for MR angiography of the renal arteries: initial experience. Radiology 2004;231:237–42. [18] Spinosa DJ, Isaacs RB, Matsumoto AH, Angle JF, Hagspiel KD, Leung DA. Angiographic evaluation and treatment of transplant renal artery stenosis. Curr Opin Urol 2001;11:197–205. [19] Jain R, Sawhney S. Contrast-enhanced MR angiography (CE-MRA) in the evaluation of vascular complications of renal transplantation. Clinical Radiology 2005;60(11):1171–81.