Original Article MRI FOR EVALUATING LIVE KIDNEY DONORS EL-DIASTY et al.
In this section, authors from Mansoura describe their experience with MRI as the sole method for the morphological and functional evaluation of live kidney donors. They recommend this technique in both instances. A series of studies from Detroit show that oxalate and hyperoxaluria induced free-radical generation, which resulted in injury to renal tubular cells. In this paper they show for the first time that hyperoxaluria-induced injury promotes individual calcium oxalate crystal attachment in the renal tubules. They also showed that this was prevented by vitamin E treatment.
Magnetic resonance imaging as a sole method for the morphological and functional evaluation of live kidney donors TAREK A. EL-DIASTY, MOHAMED E. ABO EL-GHAR, AHMED A. SHOKEIR, HOSSAM M. GAD, EHAB W. WAFA, MOHAMED E. EL-AZAB, AHMED B. SHEHAB EL-DIN and MOHAMED A. GHONEIM Urology & Nephrology Center, Mansoura University, Mansoura, Egypt Accepted for publication 31 January 2005
To evaluate gadolinium-enhanced dynamic magnetic resonance imaging (MRI) as the sole method for the anatomical and functional assessment of potential live-kidney donors.
estimated by MRI or MAG3. For the right and left kidneys the mean isotope clearance was not significantly different from that of mean MRI clearance. MR urography allowed visualization of the urinary tract and the detection of any abnormality.
SUBJECTS AND METHODS
CONCLUSION
The study included 50 consecutive kidney donors; in addition to routine donor evaluation, the kidney was imaged with Gdenhanced dynamic MRI, which was also used for selectively determining the glomerular filtration rate (GFR) of each kidney. All donors had a m99Tc-mercaptoacetyltriglycine (MAG3) renal scan as the reference standard to measure GFR. The anatomical results of MRI were compared with the findings at donor nephrectomy, and the GFR estimated from MRI compared with that from MAG3 scintigraphy.
Gd-enhanced dynamic MRI can provide accurate information about the anatomy of the urinary tract and vasculature of the kidney, and can be used to accurately estimate the selective GFR of each kidney. Therefore, we recommend MRI as a single imaging diagnostic method for assessing potential live kidney donors.
OBJECTIVE
KEYWORDS kidney, transplantation, living donors, MRI, GFR
RESULTS INTRODUCTION MR angiography had 100% sensitivity, 94% specificity and 96% overall accuracy for detecting the number of renal arteries, and 100% sensitivity, 98% specificity and 98% overall accuracy for the number of renal veins. There was a close correlation (r = 0.54, P < 0.01) between the GFR of each kidney
©
2 0 0 5 B J U I N T E R N A T I O N A L | 9 6 , 111 – 11 6 | doi:10.1111/j.1464-410X.2005.05578.x
Renal transplantation is an effective treatment for patients with end-stage renal failure. Although cadaveric transplants continue to outnumber live-donor transplants by three to one, the number of live-donor renal transplants has increased 111
E L - D I A S T Y ET AL.
steadily over the past decade. Transplants from living-related donors have better shortand long-term survival rates than do transplants from cadavers. Selectively assessing the anatomy and function of each kidney is a fundamental part of evaluating potential donors. It is important for the urologist to have detailed anatomical information about the vasculature and morphology of the kidney and ureter, and to ensure that the donor has two wellfunctioning kidneys, and that renal function is evenly divided. If there is unusual asymmetry of function in the donor then measuring individual renal function should prevent the donor being deprived of the better kidney [1]. For decades, the morphology and function of the donor’s kidneys were traditionally assessed by several separate procedures, most commonly by ultrasonography (US), catheter angiography, excretory urography and radioisotope renal scans. In a recent study, contrast-enhanced spiral CT was recommended a single method for the anatomical and functional assessment of potential live-kidney donors [1]. However, CT subjects the patient to the risk of a high radiation dose. Moreover, the use of radiocontrast materials may increase the risk of renal and systemic toxicity. However, MRI offers donors the advantages of avoiding both radiation exposure and injection of potentially nephrotoxic iodinated contrast materials. A combined MR examination including Gdenhanced MR angiography (MRA), MR nephrography and MR urography (MRU) offers several potential advantages over conventional studies for anatomical evaluation [1,2]. In the present study, we describe a new ‘all-in-one’ MRI technique that provides both anatomical and functional information for each kidney. The diagnostic accuracy of the new technique for identifying the number of renal arteries and veins, and the morphology of the collecting system, was assessed by comparing the results with the findings at donor nephrectomy; the correlation between the GFR measured using the new MRI technique was also compared with that determined from a conventional radioisotope renogram. SUBJECTS AND METHODS Between January and December 2003, 50 consecutive potential kidney donors (30 men 112
and 20 women, mean age 35 years, range 22–55) were prospectively included in the study. Apart from the routine evaluation of the donors, including abdominal US (to exclude renal stones or other abnormalities), the kidneys were imaged with Gd-enhanced MRI that was also used for selectively measuring the GFR of each kidney. All donors had a 99 mTc-MAG3 renal scan as the reference standard for measuring GFR; the two estimates of GFR were then compared. The anatomical results of MRI were compared with the findings at donor nephrectomy. All donors were evaluated clinically and had essentially normal biochemical clearance. All MRI was conducted on a 1.5 T scanner (Signa Horizon LX Echo speed, General Electric Medical Systems, Milwaukee, WI, USA) with the use of phased-array torso surface coil. Before the start of MRI, 10 mg of frusemide was administered intravenously. The procedure started by obtaining a coronal localizer (scout image) to identify the abdominal aorta and the origins of the renal arteries, followed by a coronal T2-weighted sequence for the whole of both kidneys, and six coronal fast-spoiled gradient (FSPGR) slices of the centre of the kidney. Gadodiamide (Omniscan 0.5 mmol/mL GdDTPA-BMA; Nycomed, Ireland) was injected via a wide-bore veno-catheter in the antecubital vein at 3–4 mL/s. The contrast medium in the abdominal aorta, at suprarenal level, was automatically detected using SmartPrep software (General Electric Medical Systems). MRA used a breath-hold, threedimensional (3D)-FSPGR acquisition in the coronal plane. The total amount of contrast medium was 20–30 mL, according to body weight, with a mean dose of 0.3 mmol/kg. The acquisition time was 12 s for each of the arterial and venous phases, with a 10-s gap between. After finishing the arterial and venous phases of MRA, the pre-contrast six-slice coronal FSPGR at the centre of the kidney was repeated 10 times every 30 s and then at 15 min from injection of contrast medium. Gd-enhanced excretory MRU was then generated from a coronal contrast materialenhanced 3D-FSPGR with imaging parameters identical to those of MRA. No donors had contraindications for MRI; all studies were completed with no major complications. Only four subjects had claustrophobia, overcome by assurance.
The imaging parameters for coronal T2 were 5 mm thickness, no interslice gap, repetition time 8000–1000 ms, time to echo 75–95 ms, field of view 40 ¥ 40 cm and matrix 256 ¥ 196; the respective values for coronal FSPGR were 4 mm, no interslice gap, 30–40 ms, 2–3 ms, flip angle 70∞, 42 ¥ 42 cm and 256 ¥ 160, for MRA were 2.6 mm, no gap, 0.9, 40 ¥ 32 cm, 256 ¥ 128 and slab thickness 30–50 mm. Reformatted maximum intensity projection (MIP) was used in different planes, e.g. coronal, sagittal oblique, axial and axial oblique, in the arterial and venous phases, to detect the number of vessels and to define any vascular abnormalities. Coronal and sagittal MIP for MRU was also used to identify the pelvicalyceal system and ureteric anatomy, and to define any abnormalities. Coronal T2 images were reviewed for any parenchymal or contour abnormalities, and to calculate parenchymal volume. The volume of the each renal unit was then calculated by drawing a manual region of interest (ROI) around each kidney at each T2 scan. The calculated surface area of pixels in each scan was transformed into millimetres automatically by the software. The total volume of the kidneys was calculated by adding the surface areas for each kidney and then the total surface area was multiplied by the slice thickness. For dynamic scans, we first started by visually interpreting the images, comparing the series before and after contrast medium, to determine the corticomedullary differentiation, degree of parenchymal enhancement and the excretory power of each renal unit. Renographic dynamic MRI was generated by drawing ROIs over the kidney, excluding the renal pelvis. Using the functional software tool (GE Medical System) that merges all series, a curve resembling that from isotope renography was obtained. The MR dynamic renographic curve plots the enhancement units vs time (Fig. 1), and from the curve the time to the peak, the relative maximum units of enhancement (total enhancement units for each kidney minus the total on the unenhanced scan) and the response to diuretic were obtained. Curves were then obtained for the cortex and medulla for each kidney by applying a manual ROI over each at the same scan level, and from these curves the time at which the medullary response exceeded the cortex was also calculated (Fig. 2). Other circular ROIs were obtained from the aorta to determine the peak relative enhancement of the aorta
©
2005 BJU INTERNATIONAL
MRI FOR EVALUATING LIVE KIDNEY DONORS
FIG. 1. Cortical (1) and medullary (2) curves of MR renography show the point of crossover. 189 180 160
1 2
140 NR Units (relative)
FIG. 3. MRA in the identification of arterial supply. (A) Double left renal arteries with closed ostia on reformatted coronal oblique MIP image. (B) Triple left renal arteries and double right renal arteries on coronal MIP MRA.
A
120 100 80
rt
60 40 20
0
40 80 12 160 200 240 280 320 360 400 440 480 50 0 7
0 -9
B
168
NR Units (relative)
140
2 1
120 100 60 40 20
40 80 12 160 200 240 280 320 360 400 440 480 500 7
0
s
(peak enhancement units at aorta minus peak units of the aorta on the unenhanced scan). To obtain an approximate GFR for each kidney, the total volume of each renal unit was multiplied by its peak relative enhancement, then divided by the peak relative enhancement of the aorta (to minimize the effect of differences in dose of contrast media and body weight of each subject, and differences in the rate of injection). The mean (range) post-processing time was 60 (45–70) min. The results for vascular anatomy were reviewed and compared with operative data, considered the reference standard for vascular anatomy, including the number of renal arteries and veins. Functional data were correlated with the results of renographic clearance and 24-h total creatinine clearance. MRU results were also correlated with operative findings.
©
2005 BJU INTERNATIONAL
Actual N at nephrectomy Single Multiple 30 –
1 19
45 –
1 4
kidneys, which included 30 single, 16 double and two triple renal arteries (Fig. 3a,b). One case was diagnosed as a single artery but was double, and one was diagnosed as double but was quadruple. Six cases with early branches were diagnosed accurately. MRA had 100% sensitivity, 94% specificity and 96% overall accuracy in identifying the arterial supply (Table 1). For renal veins depicted at the second pass of 3D-FSPGR, MRA accurately diagnosed three cases with a retro-aortic left renal vein, two with circumaortic left renal veins and one with a double inferior vena cava. For identifying the number of veins, 45 cases with a single vein and four with double veins were diagnosed accurately. Only one case with double veins was diagnosed as a single vein. The sensitivity, specificity and overall accuracy for identifying renal veins at MRA was 100%, 98% and 98%, respectively (Table 1).
80
0 -9
MRA Arteries Single Multiple Veins Single Multiple
s
FIG. 2. The whole kidney MR renographic curves (1, 2) indicate the normal rate of excretion.
160
TABLE 1 The agreement between the number of renal arteries and veins at nephrectomy and on MRA
The sensitivity, specificity and overall accuracy of MRI for detecting the number of arteries were calculated. The proportion of surgically confirmed single renal arteries or veins was defined as sensitivity and the proportion of surgically confirmed multiple renal arteries or veins was defined as specificity. The correlation between MR clearance and isotope clearance was assessed using simple linear regression analysis, and mean MR and isotope clearance compared using Student’s t-test.
The MR nephrogram in the axial and coronal planes provided estimates of renal size and contour (Fig. 4a,b) that correlated with information from other complementary imaging. In one donor, a small ( 99% for detecting renal disease; otherwise, MRA with CT angiography is the most cost-effective strategy. The cost-effectiveness is not only direct but also includes the advantage of lack of exposure to ionizing radiation from conventional angiography, CT and renography, and avoiding the potential adverse reaction to iodinated contrast material, that may include anaphylaxis and nephrotoxicity. In the present study we tried to use a simple technique for easy quantification of parenchymal enhancement by multiplying the total volume of each renal unit by its peak enhancement, and then divided by the aortic enhancement. Using this technique there was a good correlation between GFR values obtained by MR and those by isotope renography. By multiplying the MR renographic value by 0.25 we estimated a corrected MR GFR equivalent to isotope GFR. In conclusion, Gd-enhanced dynamic MRI has several advantages for assessing potential live-kidney donors. In the vascular phase, MRA can clearly visualize the number of renal arteries and veins, and detect any congenital or acquired vascular anomaly. Moreover, in the parenchymal phase, MRI is as accurate as radioisotope renography in determining the relative GFR of each kidney, allowing selection of the kidney for nephrectomy. In addition, the MR nephrogram provides estimates of
renal size and contour, and can identify parenchymal defects. Finally, MRU allows the visualization of the pelvicalyceal system, ureter and bladder, with the detection of any abnormality. Therefore, we recommend Gdenhanced dynamic MRI as a single imaging method for assessing potential live-kidney donors. CONFLICT OF INTEREST None declared. REFERENCES 1
2
3
4
5
6
7
8
9
El-Diasty TA, Shokeir AA, Abo El-Ghar ME, Gad HM, Refaie AF, Shehab El-Din AB. Contrast enhanced spiral computerized tomography in live kidney donors. A single session for anatomical and functional assessment. J Urol 2004; 171: 31–4 Courlay WA, Yucel EK, Hakaim AG et al. Magnetic resonance angiography in the evaluation of living related renal donors. Transplantation 1995; 60: 1363–6 Wolf GL, Hoop B, Cannillo JA, Rogowska JA, Halpern EF. Measurement of renal transit of gadopentate dimmeglumine with echo-planar MR imaging. J Magn Reson Imaging 1994; 4: 365–72 Katzberg RW, Buonocore MH, Ivanovic M, Pellot-Barakat C, Ryan JM, Whang K. Functional dynamic and anatomic MR urography: feasibility and preliminary findings. Acad Radiol 2001; 8: 1083–99 Zhang H, Prince MR, Renal MR angiography. Magn Reson Imaging Clin N Am 2004; 12: 487–503 Low RN, Martinez AG, Steinberg SM et al. Potential renal transplant donors. Evaluation with gadolinium-enhanced MR angiography and MR urography. Radiology 1998; 207: 165–72 Halpern EJ, Mitchell DG, Wechsler RJ, Outwater EK, Mortiz MJ, Wilson GA. Preoperative evaluation of living renal donors. Comparison of CT angiography and MR angiography. Radiology 2000; 216: 434–9 Fink C, Hallscheidt PJ, Hosch WP et al. Preoperative evaluation of living donors. value of contrast-enhanced 3D magnetic resonance angiography and comparison of three rendering algorithms. Eur Radiol 2003; 13: 794–801 Rankin SC, Jan W, Koffman CG. Noninvasive imaging of living related 115
E L - D I A S T Y ET AL.
kidney donors. Evaluation with CT angiography and gadolinium-enhanced MR angiography. Am J Roentgenol 2001; 177: 344–55 10 Moritz M, Halpern E, Mitchell D et al. Comparison of CT and MR angiography for evaluation of living renal donors. Transplant Proc 2001; 33: 831–2 11 Rajab A, Khabiri H, Pelletier RP et al. Magnetic resonance angiography for preoperative evaluation of potential kidney donors. J Surg Res 2004; 120: 195–200 12 Hussain SM, Kock MC, Ijzermans JN, Pattynama PM, Hunink MG, Krestin GP. MR imaging. A ‘onestop shop’ modality for preoperative evaluation of potential living kidney donors. Radiographics 2003; 23: 505–20
116
13 El-Diasty TA, Ateia M, Kamal T, El-Azab M, Gad H, Ghoneim M. Comprehensive MR evaluation of potential kidney transplant donors. Eur Radiol 2000; 10 (Suppl): E1–26 14 El-Diasty T, Mansour O, Farouk A. Diuretic contrast-enhanced magnetic resonance urography versus intravenous urography for depiction of nondilated urinary tract. Abdom Imaging 2003; 28: 135–45 15 Li W, Chavez D, Edelman RR, Prasad PV. Magnetic resonance urography by breath-hold contrast enhanced threedimensional FISP. J Magn Reson Imaging 1997; 7: 309–16 16 Semelka RC, Hricak H, Tomei A, Floth A, Stoller M. Obstructive nephropathy: Evaluation with dynamic Gd-DTPA-
enhanced MR imaging. Radiology 1990; 175: 797–803 17 Liem YS, Kock MC, Ijzermans JN, Weimar W, Visser K, Hunink MG. Living renal donors optimizing the imaging strategy-decision-and costeffectiveness analysis. Radiology 2003; 226: 53–62 Correspondence: Ahmed A. Shokeir, Urology & Nephrology Center, Mansoura University, Mansoura, Egypt. e-mail:
[email protected] Abbreviations: US, ultrasonography; MRA(U), MR angiography (urography); (3D)-FSPGR, (three-dimensional)-fast-spoiled gradient (scan); MIP, maximum intensity projection; ROI, region of interest.
©
2005 BJU INTERNATIONAL
