SINCE the classical experiments of Fahraeus, 1

Transcription

1 Evidence for a Red Cell Shunting Mechanism in the Kidney By LAAVRENCE S. LILIENFIELD, M.D., PH.D., JOHN C. ROSE, M.D. AND FRANK A. POKFIDO, M.D. Simultaneous tagged red cell and tagged plasma albumin concentration-time curves were obtained from the renal vein without recirculation, following rapid injection of the indicators into a renal artery. The mean transit time of the red cells was consistently faster than that of the plasma albumin, but the washout downslopes of the indicators was found to be identical. This is interpreted as indicating that some red cells are shunted through those pathways of shortest length. The dynamic circulating hematocrit of the kidneys averaged 89 per cent of the large vessel hematocrit. Intrarenal circulating blood averaged ml./loo Gm. of dog kidney. SINCE the classical experiments of Fahraeus, 1 it has been generally assumed that capillary blood lias a lower hematocrit ratio than large vessel blood, as a result of axial streaming, and that red cells stream through the capillaries at a higher velocity than does the plasma. With the development of better plasma and red blood cell tagging technics, interest has been revived in the "small vessel" or "capillary hematocrit" problem. Studies of the human lung and forearm from this laboratory 2 ' have failed to show evidence of significant axial streaming in these structures when measured by comparing red cell and plasma transit times during a single circulation. The calculated circulating hematocrit of these structures was not significantly different from the hematocrit of the afferent or efferent blood. During the past 10 years, several groups of investigators' 1 " 8 have attempted to measure the mammalian intrarenal hematocrit. In all of these studies, the kidneys were excised after a period for equilibration and analyzed for tagged albumin and red cell components. Intrarenal hematocrits varied between 5 and 7 per cent of the large vessel hematocrit. The results of human forearm studies have suggested that hematocrits measured by equilibration technics might differ from the dynamic circulating hematocrit. This report is concerned with the circulating hematocrit of the dog kidney studied with a single circulation technic. From the Cardiovascular Research Laboratory, Department of Medicine, Georgetown University Medical Center, Washington, D. C. Received for publication, August 27, 56. METHODS Ten healthy mongrel dogs weighing between 10 and 14 Kg., were anesthetized with sodium pentobarbital, 25 mg./kg. A generous midlinc abdominal incision was made and the kidney on the left side exposed. The kidney was freed from its bed and the renal artery and vein dissected free. A ligature was placed around the renal artery and the vessel occluded for a few seconds. During this time the renal vein was incised and a polyethylene catheter (2.5 mm., inside diameter) tied into the proximal segment. The free end of the catheter was rapidly connected to a similar catheter which had previously been inserted in a femoral vein. The renal artery occlusion was then released, and blood was seen to surge out of the kidney into the femoral vein. Flow through the catheters was unrestricted all connections having internal diameters the same as the polyethylene tubing. In effect, the procedure had established a long exteriorized renal vein. A ureteral catheter was tied into the ureter and an intravenous infusion of 5 per cent glucose in water started. Before and during each experiment, urine was observed to flow freely. Significant radioactivity was not detected in urine collected up to 5 min. following the experiment in any instance. In 5 experiments, aortic pressure was monitored continually using a strain-gage transducer and direct-writing oscillograph. Significant falls in arterial pressure were not observed during the measurement period. Mean pressure averaged 10 mm. ofhg. Five milliliters of a saline suspension of the dogs red cells previously tagged with 100 /nc of Cr 51, and washed free of any extracorpuscular Cr 51 activity, was mixed with 20 juc. of I 1 ' tagged serum albumin. One half milliliter of this mixture then was injected over a 2 to 6 sec. period into the renal artery. Immediately prior to injection, the renal venous catheter was disconnected from its coupling to the femoral venous catheter and permitted to drain freely. From the start of injection until 60 sec. later, total renal blood flow was continuously collected at 2 sec. inter- 64 Circulation Research, Volume V", January 57

2 LILIENFIELD, ROSE AND PORFIDO 65 vals into paraffined test tubes containing dried heparin solution. Only those cases in which the renal blood flow remained constant during the collection period were utilized for this study. The mean flow was 100 ml./min. with a range from 77 to 0 ml./min. except for 2 cases in which theflowswere 20 and ml./min. Following collection of the samples,.5 ml. aliquots of whole blood were pipetted into glass vials. These were centrifuged and the red blood cells washed three times with normal saline solution care being taken not to remove any of the red cells. The Cr 51 activity of these waslied aliquots were counted in a well-type scintillation counter and the activity expressed as Cr 51 counts per milliliter of whole blood. The remaining blood in the collection tubes was centrifuged and.5 ml. aliquots of the supernatant plasma was counted in a similar manner, the activity being expressed as I 11 counts per milliliter whole blood after correction for hematocrit obtained by centrifugation of at least three representative specimens (2000 G for 60 min.). No correction was made for the 2 per cent or less trapped plasma. From the Cr 61 counts per milliliter in each sample and its time of collection (corrected for collecting system catheter delay and using the midpoint of the injection period as 0 time) the mean circulation time of the red cells was determined. Similarly, from the I 11 albumin counts the plasma mean circulation time was calculated according to the method of Hamilton and co-workers. 10 Since the experimental design eliminated recirculation and the washout of indicator was rapid, it was only necessary to integrate to 0 sec. in all but one case (dog ). Beyond this time the radioactivity in the samples was negligible. The curves obtained from dog were integrated to GO sec. RESULTS Mean Circulation Time. In all experiments, the red cell mean transit time was shorter than that of the plasma. Mean circulation time of the red blood cells averaged 6.4 sec, while the mean circulation time of the plasma albumin averaged 7.6 sec. (table 1). The ratio of red cell to plasma mean transit time averaged 0.8 ±.05 (S.D.) in these experiments. Kidney Hematocrit. The circulating dynamic intrarenal hematocrit was calculated directly from the ratio of mean circulation times and the large vessel hematocrit employing the following formulas; HCT J-t <-> -L I.K1D) V v PL (KID) F HC2\L.V.) F KBC = F PL F V R y X x : X? X 1 Rh,c-\ + V 1 TPL 1 - F PL PL FPL 1 HCTa.v.) By substitution of (4) into () J I 2Vi T 1 J PL V 7' S * PL ' J^v^ -t (L. HCT^L.v.) (1) (2) () (4) (5) The ratio of the kidney hematocrit to large vessel hematocrit averaged.89 ±.0 in this series (table 2). Intravascular Kidney Volume. Circulating intravascular of the kidney was determined by summing the calculated red cell and plasma (albumin) s (table ). Red cell was obtained by multiplying the red cell mean circulation time by the measured red cell flow (total flow times large vessel hematocrit). was obtained from the albumin mean circulation time and plasma flow (total flow minus red cell flow). The red TABLE 1. Red CM and (Albumin) Mean Circulation Times Average S.D Mean circulation time Red cell (T r) (sec.) albumin <T P) (sec.) S.. 2.S S.9 S Tr/Tp.SI.S ±.05

3 EVIDENCE FOR A RED CELL SHUNTING MECHANISM IN THE KIDNEY TABLE 2. Circulating Kidney Hematocrit Tr/Tp Hct. (large vessel) Hct. (kidney) Ht/Hiv 1 IS 2.SI SO 5.S S.9,S S5 Average.S SO S.D db.05 TABLE Z. Ltitravascular lied Cell and (Albumin) Volume* Average.. Blood flow (ml./ min.) S S l.s S Total S ±.0 Kidney weight (Gm.) cell averaged.5 ml. and the plasma 6.0 ml. Total intrarenal circulating blood averaged 9.6 ml. and average kidney weight Gm. Thus the mean total intrarenal circulating blood was ml/100 Gm. kidney. DISCUSSION Application of a single circulation technic permits the measurement of a dynamic circulating hematocrit. All studies which have demonstrated organ or body hematocrits much lower than the large vessel hematocrit, have employed equilibration technics. 4 " 9 ' u " 1 Single circulation experiments have failed to show significant differences between dynamic circu- Fio. 1. Ratio of Cr 61 red cell activity to I 11 plasma activity in each sample (ordinate) plotted against time of collection for each experiment (abscissa). Ratios plotted on logarithmic scale. It is apparent that in each case the ratio falls off rapidly to a nearly constant value. lating hematocrit and large vessel hematocrit in the human lung and forearm. 2 - Application of this technic to the study of the dog kidney, however, reveals a dynamic circulating hematocrit significantly lower than the large vessel hematocrit. This would appear then to be a truly intravascular phenomenon not related to any possible slowly equilibrating albumin space. The mean circulation time of red blood cells was shorter than that of plasma. An analysis of the relationship of the tagged red cell concentration to tagged albumin in each experiment, with reference to time, is graphically summarized in figure 1. In each instance the ratio of red cell to plasma radioactivity was highest in the blood which first appeared in the renal vein. This ratio then quickly fell off to a nearly constant value. This implies that red cells are preferentially shunted, into those vascular pathways of shortest length, whereas in the remaining pathways, blood containing excess plasma circulates. The observed difference in mean circulation times cannot be explained by axial streaming of red cells in all the renal vessels. Under this circumstance, one would expect that the downslope of the tagged red cell concentration-time curve would be steeper than that of the tagged

4 LI.LIENFIELD, ROSE AND l'orfido plasma, rather than be identical to it as was observed. Preferential shunting of red cells in short vascular pathways was ingeniously surmised by Pappenheimer and Kinter 4 using less direct evidence. The renal to large vessel hematocrit ratio calculated by these authors averaged 0.48, compared to 0.89 obtained in the present body. However, the total kidney vascular (cats) was calculated by extrapolation of red cell data, which perhaps could just as easily have been extrapolated to yield results closer to the findings in this study. Though their extrapolated agreed well with the results obtained by adding the red cell to the albumin space, as these authors point out, their method of measurement of renal plasma is subject to error. Albumin spaces measured by equilibrium technics are larger than those obtained by single circulation measurements, and probably include some slowly circulating or perhaps extravascular component. The kidney, in common with other vascular beds, may contain capillaries which at times contain only plasma. The plasma in these capillaries would not mix with injected labelled albumin during a single circulation but would equilibrate with repeated circulations. thus measured by equilibration technics would be larger than when measured by single circulation technics. Hence, hematocrits calculated from equilibration methods (pseudodynamic) would be lower than when calculated from single circulation transit times (dynamic). The significance and possible mechanism of the red cell shunt in the kidney has been brilliantly discussed by Pappenheimer and Kinter. 4 The technics of simultaneous measurement of red cell and plasma transit times described in this communication permits more direct measurement of this phenomenon. SUMMARY Indicator dilution curves employing Cr 51 tagged red cells and I 1 ' albumin were obtained without recirculation in a functioning kidney of 10 mongrel dogs. Analysis of these curves and their red cell plasma activity relationships provide evidence that the mean transit time of red cells is significantly faster than that of the plasma, and that this phenomenon results from a fast red cell transit time in those vascular pathways that are of shortest length. The dynamic circulating hematocrit of the dog kidney averaged 89 per cent of the large vessel hematocrit. Circulating intrarenal blood averaged ml./loo Cm. of dog kidney. ACKNOWLEDGM ENT The authors gratefully acknowledge the encouragement of Dr. Edward D. Freis and the technical assistance of Mr. Thomas F. Doyle and Miss.lean B. Pietras. SUMMARIO IX LVTERLINGUA Curvas del dilution de indicator, establite per medio de erythrocytos etiquettate con Cr 51 e albumina a I" 1, esseva obtenite sin recirculation in un ren intacte in 10 canes mesticio. Le analyse de iste curvas e lor relationes de activitate erythrocytic e plasmatic indica que le tempore medie del transition de erythrocytos es significativemente plus breve que illo del plasma e que iste phenomeno resulta de un rapide transition del erythrocytos in le vias vascular que ha le plus breve long or. Le dynamic hematocrite circulante del ren canin habeva un valor medie de S9 pro cento del hematocrite del vasos grande. Le circulante volumine sanguinee intrarenal habeva un valor medie de ml per 100 g de ren canin. REFERENCES 1 FAHRAEUS, R.: The suspension stability of the blood. Physiol. Rev. 9: 1, LlLlENFIELD, L. S., KOVACH, R. D., MARKS, P. A., HERSHENSON, L. D., RODMAN, G. R., EBAUGH, F. P., JR., AND FREIS, E. D.: The hematocrit of the lesser circulation in man. J. Clinical Invest. In press.,, PORFIDO, F. A., AND FREIS, E. D.: The hematocrit of the human forearm capillary bed. Clinical Research Proceedings 4: 110, PAPPENHEIMER,.). R., AND KINTKH, W. B.: Hematocrit ratio of blood within mammalian kidney and its significance for renal hemodynamies. Am. J. Physiol. 185: 77, GIBSON, J. G., II, SELIGMAN, A. M., PJOACOCK, W. C, AUB, J. C, FINE, J., AND EVANS, R. D.: The distribution of red cells and plasma in large and minute vessels of the normal dog, determined by radioactive isotopes of iron and iodine. J. Clin. Invest. 25: S4S, 46.

5 68 EVIDENCE FOR A RED CELL SHUNTING MECHANISM IN THE KIDNEY 6 LEWIS, A. E., GOODMAN, R. D., AND SCHUCK, E. A.: Oxygen blood measurements in normal rats. J. Lab. & Clin. Med. 9: 704,52. 7 ALLEN, T. H., AND REEVE, E. B.: Distribution of extra plasma in the blood of some tissues in the dog as measured with P 2 and T-1S. Am. J. Physiol. 175: 8, 5. 8 EVERETT, N. B., SIMMONS, B., AND LASHER, E. P.: Distribution of blood (Fe 9 ) and plasma (I 11 ) s of rats determined by liquid nitrogen freezing. Circulation Research 4: 4, CHAPLIN, H., JR., MORRISON, P. L., AND VETTER, H.: The body venous hematocrit ratio: Its constancy over a wide hematocrit range. J. Clin. Invest. 2: 109, HAMILTON, W. F., MOORE, J. W., KINSMAN, J. M., AND SPURLING, R. G.: Studies on the circulation. IV. Further analysis of injection method, and of changes in hemodynamics under physiological and pathological conditions. Am. J. Physiol. 99: 54, VEREL, D.: Observations on the distributions of plasma and red cells in disease. Clin. Sc. 1: 51, EBERT, R. V., AND STEAD, D. A.: Demonstration that the cell plasma ratio of blood contained in minute vessels is lower than that of venous blood. J. Clin. Invest. 20: 7, BARNES, D. W., LOUTIT, J. F., AND REEVE, E. B.: A comparison of estimates of circulating red blood cell given by the Ashby marked red cell method and the T 1S hematocrit method in man. Clin. Sc. 7:, 48. The Left Atrial-Pulmonary Vascular Reflex Conclusions drawn from specific experimental data are often extrapolated to other conditions and situations. In Dexter, Gorlin and their associates assembled data obtained from different normal subjects and from patients with mitral stenosis. On the basis of this information they suggested that a fortuitous mechanism is provided by which pulmonary edema is avoided: When pulmonary capillary pressure approaches the colloid osmotic pressure of blood, a reflex from the left atrium reduces capillary pressure by constriction of pulmonary arterioles. This postulate could not be confirmed experimentally by comparing effects on isolated perfused cat's lungs and on chloralized cats in which a lobe of lung was separated and perfused at constant flows. The effects of increasing left atrial pressure on pulmonary arterial resistance in innervated lungs was similar to those in isolated lungs. The responses were not affected by atropinization, vagotomy or removal of the stellate ganglion. In short, no evidence for operation of a left atrial-pulmonary arteriolar reflex could be obtained in cats under conditions in which pulmonary and vascular reflexes pathways were intact and operative. For details see S. D. Carlill and H. N. Duke. J. Physiol. 1: 275, 56.