Clinical Science and Molecular Medicine (1975) 48, 51-60.
Measurement of intrarenal blood-flow distribution in the rabbit using radioactive microspheres
D. 1. WARREN
AND
J. G. G. LEDINGHAM
Department of the Regius Professor of Medicine, and Nuffield Institute for Medical Research, University of Oxford
(Received 8 May 1974)
Summary
was limited by these intrarenal effects of microspheres. 8. Total renal blood flow measured in six rabbits in acute experiments by the microsphere technique was 107 ± 12 (mean ± SD) ml/rnin and by p-aminohippurate clearance was 100 ± 10 ml/min. 9. Total renal blood flow in twelve conscious, chronically instrumented rabbits was 125 ± 11 ml/min, of which 92 ± 6 ml/min was distributed to the superficial cortex and 33 ± 4 mljmin to the deep cortex.
1. Total renal blood flow and its distribution within the renal cortex of the conscious rabbit were studied with radioactive microspheres of 15 and 25 pm diameter. 2. The reliability of the microsphere technique was influenced by microsphere diameter and number (dose). The optimum microsphere diameter for determination of flow distribution in the rabbit kidney was 15 pm and dose 100-150000 spheres. 3. Spheres of 15 pm nominal diameter were randomly distributed within the renal cortex of adult rabbits. The larger spheres in batches nominally 15 pm in diameter in young rabbits and 25 pm diameter in adult rabbits were preferentially distributed to the superficial cortex. 4. In adult rabbits 15 pm diameter spheres lodged in glomerular capillaries. Larger spheres occasionally lodged in interlobular arteries causing intrarenal haemorrhage. 5. Microspheres of 15 pm caused a decrease in renal clearance of creatinine and of p-aminohippurate when the total injection dose was about 200 000 spheres. These effects were greater when the injection dose was increased to 500 000 spheres. 6. The reduction in total renal blood flow observed with large doses of spheres largely reflected decreased outer cortical flow, as measured by a second injection of spheres, and confirmed by a decrease in p-aminohippurate extraction. 7. The reproducibility of multiple injection studies
Key words: microspheres, renal blood flow.
Introduction Investigation into the rates of blood flow to different zones of the kidney was stimulated by the countercurrent hypothesis (Wirtz, Hargitay & Kuhn, 1951), and its requirement for low renal medullary blood flow. The inert-gas washout technique has been widely used in the investigation of renal blood flow distribution in animal and human studies (Hollenberg, Epstein, Rosen, Basch, Oken & Merrill, 1968; Blaufox, Fromowitz, Lee, Meng & Elkin, 1970; Hollenberg & Merrill, 1970), but recent criticism (Mowat, Lupu & Maxwell, 1972; Brand & Cohen, 1972) has stimulated interest in alternative techniques. The radioactive microsphere technique for measurement of intrarenal blood-flow distribution does not require catheterization of the renal artery, exploration of the kidney or collection of urine,
Correspondence: Dr David J. Warren, Department of Medicine, Royal Infirmary, Edinburgh, Scotland.
51
52
D. J. Warren and J. G. G. Ledingham
and may be used in conscious animals. We have previously described techniques for measurement of total organ blood flow by the microsphere technique in conscious chronically instrumented rabbits (Warren & Ledingham, 1974b). The following studies were undertaken to evaluate the limitations of the microsphere technique for measurement of flow distribution within the kidney, and to investigate the effects of microspheres on renal function. Particular attention was paid to rheological considerations. Methods Materials
White New Zealand male rabbits were used, and prepared with chronically indwelling left atrial catheters, and aortic thermistors for measurement of cardiac output as described elsewhere (Warren & Ledingham, 1972, 1974a; Warren, 1974). Microspheres were obtained from the 3M Company (Minnesota Mining and Manufacturing Co., Riker Laboratories, Loughborough, U.K.) and labelled with a radioactive isotope (8SS r, SlCr or 46SC), which was incorporated into the structure of the spheres. After injection of microspheres into the left atrium, total renal blood flow was measured, after estimation of fractional distribution of microspheres to the kidney, with a total body counter. We have shown that no shunting of 15 JIm diameter spheres occurs through the rabbit kidney (Warren & Ledingham, 1974b). Microscopic studies of the distribution of microspheres and their pathological effects
After injection of 15 and 25 JIm diameter spheres into conscious rabbits of different ages, the animals were killed and the kidneys prepared for microscopic examination of 8-12 JIm thick sections with Haematoxylin and Eosin stain. Observations were made on (a) the anatomical location of spheres within the kidneys, (b) the distribution of spheres within the cortex, according to size, and (c) the pathological changes associated with spheres of different sizes within the kidney. Clearance studies
Renal plasma flow (RPF) was estimated by p-aminohippurate clearance (C PA U) and glomerular filtration rate (GFR) by exogenous creatinine
clearance (Ce,) in conscious rabbits. The techniques used were identical with those of Korner (1963), except that dextrose (277 mmol/l) rather than mannitol was used for infusion. Before clearance studies, rabbits were allowed free access to food and water. Blood samples were obtained from a central ear artery catheter, which was also used for blood pressure recordings. A marginal ear vein catheter in the opposite ear was used for infusions. At the conclusion of three 20 min clearance periods, cardiac output was measured and microspheres were injected immediately afterwards through the left atrial catheter. The dose and diameter of the microspheres varied with the purpose of the experiment. In animals used for measurement of p-aminohippurate extraction ratios (EPA u ) , a catheter was passed into the renal vein at laparotomy via the right external jugular vein. The rabbit was then studied 3-4 h after completion of surgery, which was done under pentobarbitone anaesthesia. p-Aminohippurate was determined on diluted urine samples and cadmium sulphate precipitates of plasma. Creatinine was determined after adsorption to Lloyd's reagent by reaction with alkaline picrate solution. Both assays are described by Varley (1963), and were modified for 0'5 mI. Techniquefor sectioning the kidney
At the end of microsphere experiments the animals were killed, the kidneys removed and their total radioactivity was counted. The two kidneys were weighed after removal of fat and extra-pelvic tissues and the three diameters measured in a standard fashion with a steel caliper and Vernier scale. The maximum diameters from pole to pole (d 1 ) and from the medial to the lateral border (d z) were measured with the kidney lying on its posterior surface. The maximum antero-posterior diameter was measured with the kidney lying on its lateral border (d 3 ) . In spite of the semi-fluid consistency of the fresh kidney, these measurements could be made reproducibly. Comparison of fifty triplicate measurements showed that the standard deviation was always less than 5% and usually 2-3% of the mean. The diameters d., d z and d3 were used for calculation of intrarenal volumes. After measurement, the kidneys were placed in 10% neutral formalin in an oven at 37°C for 48 h before sectioning. At the end of this period the
Intrarenal distribution of microspheres
53 SUPERFIC IAL CORTEX DEEP CORTEX
- -
SECTION A
MEDULLA
-r
5mm approx.
l
SECTION B
- -
-
LINES OF SECTION
ANTER I OR SURFACE
TRANSVERSE SECTION
FIG.,1. Diagrams, to show the techni9ue used for sectioning the formalin-fixed rabbit kidney after microsphere studies, The junction between superficial and deep cortex was at a point judged by eye to be mid-cortical thickness.
poles of the kidneys were removed and three horizontal sections about 5 mm thick made through the body of the kidney (Fig. 1). Before subdividing these sections cortical thickness was measured in four places around the circumference of each of the three sections, and the mean of these twelve observations taken as mean cortical thickness (MCT). The cortex was easy to discriminate in the fixed sections, since its whitish colour contrasted markedly with the deeper colour of the medulla. With a sharp blade it was possible to dissect the medulla free from the cortex and then to section the ring of cortex at a point judged by eye to be mid-cortical thickness. Because of the potential error arising from this technique, the individual zones of superficial and deep cortex and of medulla from each of the three original sections were weighed and counted separately in an automatic gamma counter. In this way it was possible to estimate the reproducibility of the sectioning technique in different slices of the kidney. From the data obtained by measurement of the kidney, the volumes of superficial cortex, deep cortex and medulla were calculated as follows. (The specific gravity of the kidney is assumed to be unity; KW = kidney weight (g); d., ds, 3 = kidney diameters; MCT = mean cortical thickness.) If the kidney were a perfect sphere, its volume would be
d
4/3 n
(dl~2d3)
The kidney is not a perfect sphere, but, knowing kidney weight, a correction factor (C) may be calculated as follows: C=
KW nd!d2d3/6
This factor may then be used to correct calculations of the volumes of the kidney based on the formula for the volume of a sphere. Therefore the volume of the medulla is: n(d! - 2MCT) (d 2 - 2MCT) (d 3 - 2MCT)C 6
and the volume of the medulla + deep cortex is n(d! - MCT) (d 2 - MCT) (d 3 - MCT)C 6
The volume of the deep cortex can then be obtained by difference. Provided that the medial indentation of the renal pelvis makes a proportionally equal contribution to loss of superficial cortex, deep cortex and medulla, this method of calculation should be valid (McNay & Abe, 1970). To check this, the values obtained for total cortex and medulla by calculation were compared in ten kidneys with the results obtained by the standard histological technique of point counting (Anderson & Dunnill, 1965). Calculation of regional blood flow in the kidney
The following information was obtained for each
D. J. Warren and J. G. G. Ledingham
54
kidney: (a) total renal blood flow; (b) the volumes of superficial and deep cortex; (c) the ratio of radioactivity counts (expressed as c.p.m.Ig) for superficial and deep cortex. From these data it was possible to calculate superficial and deep cortical flow in both ml/min and ml min- 1 100 g- 1, and the ratio of the flow between superficial and deep cortex. Expression of results
All results are expressed as the mean (M) ± so. The significance of differences was determined by Wilcoxon's rank sum test.
Results Microscopic studies
Measurement of glomerular blood flow by the microsphere technique requires that all spheres should become lodged in glomerular capillaries during their first passage through the kidney (Fig. 2). Preliminary experiments in conscious and anaesthetized animals showed that only spheres of 15 Jim nominal diameter fulfilled this condition. Larger spheres became lodged in arterioles proximal to the glomeruli,and on many occasions impaction of 25 Jim spheres in this way led to local haemorrhage within the kidney (Fig. 3). Blockage of arteries or arterioles was also frequently followed by 1. Effect of microsphere size and kidney size on the anatomical distribution ofspheres within the kidney
TABLE
Mean body wt. (kg) Mean kidney wt, (g) No. of spheres counted Sphere diameter (nominal) (Jim)
Three adult rabbits
Three adult rabbits
Three young rabbits
2'40 9'8 1840
2·70 10'2 471
0'48 4'2 524
25
Distribution of spheres (%) Glomerulus 18 Afferent arteriole I1 Interlobular artery 59 Interlobar/arcuate artery 0 Medulla 0 Obstructed within arteries 12
15
15
92 3 3
52 9 28
0 0
0
2
11
0
accumulation of spheres behind the obstructed one (Fig. 3). Microspheres were not seen in any vessel within the medulla of any of the sixteen kidneys examined. In serial sections through many kidneys the ratio of microspheres to glomeruli was approximately 1: 10 when the total injected dose was about 100 000 spheres. The relationship between body weight, kidney weight, microsphere size and anatomical location of microspheres was investigated in conscious uninjured rabbits. At least twenty kidney sections were examined from three adult rabbits injected with 25 Jim diameter spheres, three adult rabbits injected with 15 pm diameter spheres, and three young rabbits injected with 15 Jim diameter spheres, and the anatomical location of every sphere within each section was recorded (Table 1). It is apparent from these data that sphere size and kidney size are important factors in determining the distribution of microspheres. Whereas 95% of 15 Jim spheres in adult rabbits lodged in either the glomeruli or afferent arterioles, only 29% of 25 Jim spheres did so in adult rabbits, and 61% of 15 Jim spheres in young rabbits. Distribution of spheres according to size
The symmetry of total renal counts with 15 and 25 Jim diameter spheres has been shown elsewhere (Warren & Ledingham, 1974b). Three further
studies were designed to investigate the uniformity of sphere distribution within individual kidneys. First, the poles were removed from both kidneys in five rabbits with a razor blade, and these were then weighed and counted in an automatic gamma counter. There was excellent correlation between the counts in the four renal poles from anyone rabbit, the so value in each case being less than 5% of the mean counts. Secondly, six punches were made through the cortex of the body of the kidney in a random fashion with a 5 mrn cork borer. The medulla was trimmed away, and the remaining small pieces of renal cortex were weighed and counted. The counts are shown in Table 2. There was excellent agreement between the counts from six cortical punches taken from the same kidney, and also between the twelve cortical punches taken from two kidneys in one rabbit. Thirdly, microsphere size distribution within the rabbit kidney was determined by plotting the relationship between microsphere size and depth
lntrarenal distribution of microspheres
FIG. 2. Section of adult rabbit kidney showing a 15 11m diameter microsphere lodged within a glomerulus.
FIG. 3. Section of adult rabbit kidney showing impaction of a 25 11m diameter microsphere within an interlobular artery. Five microspheres have become trapped behind the impacted sphere. and an area of haemorrhage can be seen to the left of these spheres. Some fragmentation of the spheres has occurred during sectioning. (Facing p, 54)
Intrarenal distribution of microspheres TABLE
55
2. Relationship between radioactivity counts in twelve cortical punches taken randomly from both kidneys in five rabbits Cortical punches (c.p.m.rg)
Rabbit no. Kidney
2
3
4
5
6
Mean ±SD
so/mean
co
84
L R
19642 17466
18827 18643
16287 18927
16426 16949
17483 17386
18269 17829
17844 ±1040
5-83
88
L R
26506 27627
30392 28423
29330 29167
29678 28163
28496 28923
28763 28567
28669±993
3-46
86
L R
18653 19713
19358 19024
16484 19580
19945 20060
18327 19762
17825 19926
19054±1073
5'63
85
L R
28788 33231
28717 31550
30820 28253
30520 29573
29462 30687
28681 29832
30009 ±1435
4·80
81
L R
40547 38599
40159 37492
40102 36522
41242 37302
40793 40692
39972 41792
39601 ±1703
4'30
from the renal capsule in young and adult rabbits after injection of either 15 or 25 /lm diameter spheres. To eliminate the possibility that the same sphere might be counted more than once in successive sections, data were obtained only from every third section. Observations were made with a micrometer eyepiece of the maximum diameter (in micrometer units) and depth from the capsule of large numbers of intraglomerular spheres in at least twenty sections from each kidney. After completion of these observations, every sphere was allocated to one of four cortical zones, each representing 25% of cortical thickness. The aglomerular cortex was TABLE
ignored, and zone 1 represented the outermost 25%, and zone 4 the deepest 25% of total cortical thickness. The sphere size in vitro was also measured for each batch of spheres used. The results are shown in Table 3. In the adult rabbits there was no difference between sphere size distribution in any zone of the kidney and size distribution in vitro when 15 /lm diameter spheres were used. By contrast, there was a significantly greater (P< 0'01) proportion of large spheres in the superficial zone in the kidneys of adult rabbits injected with 25 /lm· spheres and young rabbits injected with 15 /lm spheres.
3. Effect of cortical depth and microsphere diameter on intrarenal distribution of spheres
Diameter of microspheres is expressed in micrometer units (I unit = 2·53/lm). P = significance ofdilference between intrarenal sphere diameter and diameter in vitro. NS = not significant.
No. of spheres counted Sphere diameter (nominal) (/lm)
Three adult rabbits
Three adult rabbits
Three young rabbits
2186
872
671
25
15
15
Mean±sD diameter
P
Mean Lso diameter
P
Mean±SD diameter
P
1 2 3 4
9·5 ±l-6 8·8±1·3 8·9±1·3 9'1 ±t'3