topeeng.gif (8416 bytes)

[ Scientific Activities - Actividades Científicas ]

X-Ray Methods to Investigate Intrarenal Distribution of Blood Flow and Tubular Fluid Flow Dynamics

Juan Carlos Romero M.D.
Laura I. Pelaez, M.D.

Department of Physiology and Biophysics and Division of Hypertension
Mayo School of Medicine, Mayo Clinic
Rochester, Minnesota. USA


Computed Tomography (CT) Image Reconstruction
Fast Computed Tomography
Dynamic Spatial Reconstructor
Electron Beam Computerized Tomography (EBCT)
Determinations of Renal Perfusion and Flow
Determination of Tubular Dynamics
Intrarenal Hemodynamics in the Normal and Hypertensive Human’s Kidney


Diseases such as hypertension have a major health impact on the population. Several pathophysiological mechanisms, e.g. changes in total and intrarenal blood flow have been suggested to play a role in such diseases. However, elucidation of the ultimate mechanisms involved remains unclear.

For almost 20 years computed tomography (CT) provided static representation of the internal organ structure. But it was not until recently that CT imaging techniques permit the dynamic investigation of an organ, like the kidney. More precisely, we have utilized these techniques to study regional renal blood flow and tubular function. Thus, use of these techniques may shed light into the mechanisms involved in diseases like hypertension and ischemic nephropathy.

Key words: electron beam computed tomography, renal blood flow, renal tubular fluid flow


Hypertension is one of the most common health conditions in industrialized societies. It is the single most important contributor to the high rate of cardiovascular disease, as well as being a major contributor to the increasing incidence of end stage renal failure. Thus its pathophysiology has been extensively studied and several mechanisms have been proposed. In 1951 Selkurt, et al. [1] and Shipley and Study [2] introduced the concept of pressure induced natriuresis. They observed that significant changes in renal blood flow (RBF) and glomerular filtration rate (GFR) did not necessarily accompany changes in renal perfusion pressure and sodium excretion, because of regulatory renal vascular response. Later Goodyer, et al. [3] suggested the possibility of intrarenal distribution of blood flow, to areas of the renal parenchyma where tubular and sodium reabsorption were not affected. This redistribution has been initially proposed to occur within the cortex [4], and later from the renal cortex to the medulla [5].

In other sodium retaining states, such as cardiac insufficiency [6] and cirrhosis with ascites [7] the concept of intrarenal blood flow distribution has been underscored. However in these conditions, the uncoupling or disengagement between renal hemodynamics and tubular sodium reabsorption could lead to significant sodium retention in the presence of normal RBF and GFR. This phenomenon may underlie regulation of urine sodium excretion and blood pressure, and may play a role in various pathophysiologic situations. Therefore, estimations of intrarenal RBF in the cortex and medulla, and the corresponding changes in tubular fluid flow could significantly contribute to the diagnosis and assessment of renal function in several disease states, such as hypertension, renovascular hypertension and ischemic nephropathy.

In this review we will attempt to summarize the advances recently made in evaluation of total RBF, regional cortical and medullary perfusion, and tubular fluid flow dynamics by the application of electron beam computed tomography (EBCT).

Computed Tomography (CT) Image Reconstruction

CT image introduced a great advance in imaging diagnosis because of its three-dimentional image acquisition. This tomographic image results from achieving multiple cross-sectional images of a single projection (transaxial reconstruction), and applying the back projection reconstruction method. This process is illustrated in Fig. 1 which shows the basic operations underlying reconstruction of a transverse section of a cross-shaped object using two (orthogonal) x-ray projections oriented perpendicular (90° rotation) to the arms of the cross. Variations in the opacity of the projections, which are determined by the thickness of the cross segment interposed in the x-ray pathway, can be recorded in an x-ray film represented in Fig. 1 by an open square. The opacity of the projection can also be recorded as an analog signal by photo cells aligned along the same transverse plane. This is shown in Fig. 1 where the analog displays curves reaching the values 3 and 1, corresponding to the same x-ray density of the back projections recorded in the film. The summation of the opacities recorded from the two projections illustrated in Fig. 1 can also be represented (digitized) by a numerical set of values from which the transversal section of the cross can be reconstructed. Once a tomographic image is obtained manipulation of a series of contiguous cross-sectional slices (three-dimensional reconstruction) can provide a volume image of an organ, but does not provide any functional information.

fig1.gif (15604 bytes)
Fig. 1: Back projection procedure for generation of multidimensional image reconstruction.

Fast Computed Tomography

CT can provide tomographic imaging of any cross-section of the human body, thereby enabling characterization of their internal structure. However this significant advance in disclosing anatomical detail provides a little dynamic information regarding organ function. Dynamic imaging allows assessment of organs, whose function is related to the flow of blood or other locally formed fluids, like the kidney. Measurements of these dynamic parameters necessitate very fast sequential scanning of the same cross-section in so that tissue blood perfusion or tubular fluid flow could be determined by external detection of the transit of a bolus of x-ray filterable contrast media. The accuracy with which the change in tissue density, consequent to transit of contrast, can be recorded is closely dependent on the number of images that can be obtained per unit of time, and the speed by which each can be acquired (temporal resolution).

Dynamic Spatial Reconstructor

The first three-dimensional volume scanning (simultaneous acquisition of multiple contiguous slices) CT scanner with high temporal resolution (scan repetition rate of up to 60 times/sec), called the dynamic spatial reconstructor (DSR), was presented in 1980 [8].

This machine allowed to examine the renovascular anatomy, detecting arterioles as small as 1 mm in diameter [9] and providing detailed dilution curves that showed the transit of contrast in the four zones of the renal cortex (superficial, middle, inner and juxtamedullary) and in two different areas (outer and inner) of the renal medulla. These dilution curves allowed a precise calculation of intrarenal RBF. Despite the great potential of the DSR, it was not extensively used because of its limited availability, and its high operating and maintenance costs. However, further studies on intrarenal hemodynamics were made possible when Imatron, a California company, marketed the first commercially available EBCT, which is described in the next section.

Electron Beam Computerized Tomography (EBCT)

This instrument represents a novel concept in the use of x-ray to obtain fast tomographic scanning. In contrast to the DSR and conventional CT, EBCT has no mechanical parts (x-ray tubes and/or TV cameras) moving around the patients, resulting in lower heat production and enabling fast scanning. An electron beam, originating from an electron gun located behind the patient is magnetically deflected sequentially onto four tungsten target rings, producing eight fan beams (two from each target ring) of x-ray radiation that pass through the patient. Eight almost simultaneous renal tomographic sections can thereby be obtained, that are thicker (8 mm) than those produced by the DSR. Alternatively, consecutive 1.5, 3, or 6 mm thick tomographic slices can be obtained by using a single target ring and moving the patient table at pre-determined increments. Although its temporal resolution is lower than that offered by the DSR (50 or 100 msec/image), it is nonetheless sufficient to obtain adequate evaluation of renal function. Furthermore, because of the slightly longer scan duration and lower image noise compared to the DSR, its spatial resolution is superior [10].

Jaschke, et al. [11, 12] were the first to demonstrate the potential of the EBCT in measuring RBF, and establish the basic principles for that calculation which correlated highly with measurements obtained with radioactive microspheres. Subsequent validation studies demonstrated the accuracy of EBCT-derived measurements of renal, cortical and medullary (compared to their in vitro) volumes [13] and perfusion (compared to electromagnetic flowmetry) within a wide range of RBF values [14], as well as changes in blood flow distribution [15].

Determinations of Renal Perfusion and Flow

CT flow studies are usually based on sequential scanning of the same tomographic level during the transit (first pass) of a bolus of contrast media. Since tissue density is linearly related to contrast concentration [11], and the contrast is assumed to be well-mixed with blood [16] (which is especially true using intra-venous injections), flow of contrast to a perfused tissue will follow the distribution of blood and will result in a parallel increase in the density (CT numbers) of the tissue. The changes in tissue density (representing a change in contrast concentration) can be plotted against time, and through the application of mathematical algorithms, regional perfusion (in units of ml blood/min/cc tissue) can be assessed according to the principles of the indicator-dilution theory [16].

Determination of Tubular Dynamics

An important step towards evaluating renal function using tomographic imaging was the description of the dynamic characteristics of tubular fluid transit through different nephron segments, using an inulin-like contrast medium as a marker [15]. The scanning sequence was designed so as to first follow the transit of x-ray contrast through renal vasculature in the cortex and in the medulla. After completion of the first pass of the bolus through the kidney (5-6 seconds) a residual opacity is observed (Fig. 2) that, corresponds mainly to contrast which has been filtered in the glomeruli and is contained in the proximal tubules. Continuing the scanning sequence at a few second intervals, displacement of this contrast medium through the nephron can be traced. Potts, et al. [17] were the first to measure the transit times between two different regions of the canine kidney from EBCT contrast attenuation curves. Their recent report and ours [15] concluded that these displacements indicate the flow of contrast medium along the different nephron segments (Fig. 2). Furthermore, the transit time of contrast media displaced through all these renal segments agreed closely with those measured in dogs by Steinhouser, et al. [18, 19] using lisamine green. We have further shown that the contrast also showed changes in density similar to those described for inulin, whose concentration progressively increases towards the tip of the papilla (because of the osmotic gradient) and re-dilutes as it returns to the cortex through the thick ascending loop of Henle. This suggested that changes in tubular fluid reflecting mostly sodium reabsorption in a given tubular segment could be detected by corresponding changes in the density of contrast medium and by proportional alteration in the transit time (Fig2). To test this assumption, the kidney was scanned during administration of furosemide [15], a loop-diuretic that decreases sodium chloride reabsorption in the ascending loop of Henle [20]. Fig. 3 demonstrates a corresponding significant contrast dilution (or fall in x-ray density) in the loop of Henle and collecting system, which was detected with EBCT.

These results show that EBCT constitutes an excellent method to determine accurate changes in intrarenal distribution of blood flow and the manner in which these changes are coupled to changes in tubular sodium excretion [15]. This method is likely to be extremely useful to evaluate ischemic nephropathy and some forms of renal sufficiency.

fig2.gif (14046 bytes)
Fig. 2: Changes in the displacement of x-ray contrast medium density through
different regions of renal cortex and medulla yielding dilution curves that
correspond to different nephron segments.

fig3.gif (25290 bytes)
Fig. 3: Changes in contrast media concentration in different tubule segments
produced by the administration of a diuretic (furosemide).

Intrarenal Hemodynamics in the Normal and Hypertensive Human’s Kidney

Using methodology developed in animal studies [13, 14, 15], EBCT estimates of the whole kidney, cortical, and medullary perfusions and volumes have been shown to be feasible and highly reproducible in normal humans under controlled conditions [21]. Similar principles were applied in recent years in a series of EBCT studies aimed to investigate intrarenal perfusion and volume in essential and renovascular hypertension. Normotensive humans with a family history of essential hypertension [22] as well as essential hypertensive patients [23], were found to have normal regional renal volumes, but significantly lower than normal cortical perfusion. In patients with renovascular hypertension and preserved renal function, the decrease in cortical perfusion in atherosclerotic renal artery stenosis exceeded in severity the degree of the stenosis, underscoring the systemic nature of atherosclerosis. Future studies of renal tubular dynamics in animal and human models of hypertension may potentially shed light on the nephron sites and degree of impairment of renal function in this disease.


In conclusion, novel CT technology provides a unique opportunity for the study of organ function. More specifically, it permits the study of renal regional blood flow and tubular function in a simultaneous and non-invasive fashion. This provides an almost unprecedented opportunity to study diseases that may involve mechanisms associated with shifts in regional blood flow as well as those that may involve a coupling between blood flow and tubular function. Even though this methodology is mainly used for animal research, we anticipate that in the near future this technology will be available for its use in routine clinical practice.


The authors would like to acknowledge the editing and typing of Ms. Kristy Zodrow as well as the critique of Dr. Luis Juncos. This review and the study reported herein were supported by grants from the National Institutes of Health, Heart Lung and Blood Institute R01 HL 16496 and K08 HL 03621.



1. Selkurt EE, Hall PW, Spencer MP. Influence of grades arterial pressure decrement on renal clearance of creatinine, p-aminohippurate and sodium. Am J Physiol 1949; 159:369-379.

2. Shipley RE, Study RS. Changes in renal blood flow, extraction of insulin, GFR, tissue pressure and urine flow with acute alterations of renal artery blood pressure. Am J Physiol 1951; 167:676-688.

3. Goodyer AVN, Mattic LR, Chetrick A. Renal response to non-shocking hemorrhage. Sodium retention at constant perfusion pressure. Proc Soc Exptl Biol Med 1958; 97:422-425.

4. Barger AC. Renal hemodynamic factors in congestive heart failure. Ann NY Acad Sci 1966; 139:273-284.

5. Roman RJ, Cowley AW Jr, Garcia-Estan J, Lombard JH. Pressure-diuresis in volume expanded rats: Cortical and medullary hemodynamics. Hypertension 1988; 12:168-172.

6. Vander AJ, Malvin RL, Wilde WS, Sullivan LP. Re-examination of salt and water retention in congestive heart failure (Editorial). Am J Med 1958; 25:497-502.

7. Atucha NM, Garcia-Estan J, Ramirez A, Perez M, Quesada T, Romero JC. Renal effects of nitric oxide synthesis inhibition in cirrhotic rats. Am J Physiol 1994; 267:R1454-R1460.

8. Ritman EL, Kinsey JH, Robb RA, Gilbert BK,Harris LD, Wood EH. Three -dimensional imaging of the heart, lungs, and circulation. Science 1980, 210:273-280,

9. Bentley MD, Hoffman EA, Fiksen-Olsen MJ, Knox FG, Ritman EL, Romero JC. Three-dimensional canine renovascular structure and circulation visualized in situ with the dynamic spatial reconstructor. Am J Anat 1988; 181:77-88.

10. McCollough CH, Robb RA. Ultrafast computed tomography: Principles and instrumentation. In: Marcus Cardiac Imaging. Edited by Skorton DJ, Schelbert HR. Philadelphia: WB Saunders Co.; 1996. pp. 793-819.

11. Jaschke W, Gould R, Assimakopoulos PA, Lipton MJ. Flow measurements with a high-speed computed tomography scanner. Med Phys 1987; 14:238-243.

12. Jaschke W, Cogan MG, Sievers R, Gould R, Lipton MJ. Measurement of renal blood flow by cine computed tomography. Kidney Int 1987; 31:1038-1042.

13. Lerman LO, Bentley MD, Bell MR, Rumberger JA, Romero JC. Quantitation of the in vivo kidney volume with cine computed tomography. Invest Radiol 1990; 25:1206-1211.

14. Lerman LO, Bell MR, Lahera V, Rumberger JA, Sheedy PF, Sanchez Fueyo A, Romero JC. Quantification of global and regional renal blood flow with Electron Beam Computerized Tomography. Am J Hypertens 1994; B:829-837.

15. Lerman LO, Rodriguez-Porcel M, Sheedy PFI, Romero JC. Renal tubular dynamics in the intact canine kidney. Kidney Int 1996; 50:1358-1362.

16. Rumberger JA, Bell MR, Feiring AJ, Behrenbeck T, Marcus ML, Ritman EL. Measurement of myocardial perfusion using fast computed tomography. In: Cardiac Imaging: A Companion to Braunwald’s Heart Disease. Edited by Marcus M, Schelbert H, Skorton D, Wolf G. Philadelphia: W.B. Saunders; 1991. pp. 688-702.

17. Potts DG, Brody AS, Shafik IM, Lumsden CJ, Zielinski A, Silverman M, Whiteside CI. Demonstration of renal tubular flow by selective angiographic computed tomography. Can Assoc Radiol J 1993; 44:364-370.

18. Steinhausen M, Tanner GA. Microcirculation and tubular urine flow in mammalian kidney cortex (in vivo microscopy). In: Sitzungsberichte der heidelberger Akademie der Wissenschafter, Mathematisch-naturwissenschaftliche Klasse. Edited by Abhandlung. New York: Jahrgang; 1976. pp. 3.

19. Steinhausen M, Hill E, Parekh N. Intravital microscopical studies of the tubular urine flow in the conscious rat. Pflugers Arch 1976; 362:261-264.

20. Mertz JI, Burnett JC Jr, Knox FG. Diuretic therapy in congestive heart failure. In: Cardiology: Fundamentals and Practice. Edited by Brandenburg R, Fuster V, Giuliani W, McGoon D. Chicago: Year Book Publishers; 1987. pp. 560-565.

21. Lerman LO, Flickinger AL, Sheedy PF, Turner ST. Reproducibility of human kidney perfusion and volume determinations with electron beam computerized tomography. Invest Radiol 1996; 31:204-210.

22. Flickinger AL, Lerman LO, Sheedy PF, Turner ST. The relationship between renal cortical volume and predisposition to hypertension. Am J Hypertens 1996; 9:779-786.

23. Lerman LO, Taler SJ, Textor ST, Sheedy PF, Stanson AW, Romero JC. CT-derived intra-renal blood flow in renovascular and essential hypertension. Kidney Int 1996; 49:846-854.


Address for correspondence:
Dr. J C Romero
Department Physiology and Biophysics
Mayo Clinic and Foundation
200 First Street SW
Rochester, MN 55905
Ph: 507-284-2322
Fax: 507-284-8566