Copyright, 1956, by the Society for Experimental Biology and Medicine. Reprinted from PROCEEDINGS OF THF: SOCIETY FOR EXPERIMENTAL BIOLOGY AND MEDICINE, 1956 vo", 14x-194 Transcapillary Loss, Equilibrium Time, Half-Return Time of Thiocyanate and Heavy Water in the Forearm.* (22424) EDWARD D. FREIS AND HAROLD W. SCHNAPER. Cardiovascular Research Laboratory, Georgetown University School of Medicine and Veterans Administration Hospital, Washington, D.C. We previously described a method for esti- mating the transcapillary exchange of perme- able substances in the human forearm( 1). The method later was extended to the pul- monary capillaries of man( 2). No attempt was made in these experiments to estimate the length of time the permeable substances re- main in the tissues or their rate of return to the circulation. More recently we have de- scribed a method for determining the late washout slopes of injected tracer materials in the human forearm(3). The method takes advantage of the fact that the forearm circu- lation is small in comparison to the general circulation. Thus, brachial arterial injection of tracer materials in dosages sufficient to produce significant concentrations in the ef- fluent ipsilateral veins produces negligible con- centrations when diluted in the general circu- lation. In this way contamination and conse- quent distortion of the late washout slopes due to recirculation of significant amounts of labelled material can be avoided. The present report is concerned with the transcapillary loss and later return of the ex- tracellular electrolyte, thiocyanate, and of heavy water. We believe that these studies provide a more complete picture of the cir- culation of extravascular substances under physiological circumstances than has been available heretofore. * Supported in part by research grants from the N.H.I., National Institutes of Health, U. S. Public Health Service and the Squibb Institute for Medical Research, New Brunswick, N. J. Method. Preparation of labelled materials and manner of injection and sampling have been described in other communications(4,1, 3). The method of analysis of thiocyanate and deuterium oxide also has been described previously( 1) . The subjects were young or early middle-aged males who were on the wards of the Veterans Administration and Georgetown University Hospitals. They all were ambulatory and afebrile at the time of testing and were not suffering from circula- tory disease which might interfere with nor- mal transcapillary exchanges in the fore- arm. Twenty-two tests were carried out using T-1824 and thiocyanate. In 8 subjects D20 was determined. In 13 tests forearm and hand capillary beds were included. In one of these subjects (M.W., Table I) simultaneous sampling was carried out from cephalic and median veins. In 8 subjects tracer materials were limited to the forearm by inflating a cuff about the wrist to pressures of 100 mm Hg above systolic pressure. In the remaining 2 cases, permeability characteristics of the hand alone were studied, injection being made into a radial artery with subsequent sampling from a vein near the wrist. The site of trans- capillary exchange (forearm or hand or both) did not appear to make any significant dif- ference in the results obtained. Definition of Terms. Methods for deter- mining rate of net return have not been de- scribed previously and, therefore, requires some explanation. The method of determin- ing transcapillary percentage loss has been THIOCVANATE AND HEAVY WATER IN THE FOREARN described previously( 1) so that its principles only will be briefly outlined here. Mere in- jection of a permeable substance with later sampling is not adequate because it is im- possible to determine the extent of its dilu- tion by the blood. The present technic com- bines the permeable tracer materials with an impermeable substance (T-1824). A portion of this mixture is injected into the brachial artery and another portion saved for analysis of the relative concentrations of the various substances in the injectate. 1. Expected con- centration of each permeable tracer is the concentration of the substance in each sam- ple that would be expected if there were no transcapillary gain or loss as the material passed through forearm circulation. The per- meable labelled substance is mixed completely with the impermeable tracer prior to injec- tion. It is assumed that the two travel to- gether and are equaily diluted by the circu- lating blood subsequent to injection into the brachial artery. It does not matter if the pattern of blood flow distribution is uneven in the forearm so long as the impermeable and permeable tracer are distributed to each vas- cular bed in the same proportionate concen- tration as was present in the injectate, The plasma samples collected at intervals of 2 to 4 seconds from an effluent vein are analyzed for their concentration of the impermeable tracer (T-1824). Concentrations of both permeable and impermeable substances also are determined in an aliquot of the injected mixture. The expected concentrations of the permeable substance in each sample are cal- culated as follows: C: I_ $ x C",r, where Ct x is the expected concentration of the permeable substance in each sample, C, the concentra- tion of this substance in the injectate, C, the injectate concentration of the impermeable tracer and C', the respective sample concen- tration of the impermeable material. Suit- able correction must be made for red cell penetration as has been described in a previ- ous report ( 1). The venous samples also are analyzed for their actz~l concentrations of each permeable tracer. The values of ex- pected and actual concentrations then are CO~CENlRATiDN OF PERUEABLE SUBSTANCE I3CYCLE LOG SCALEI -.*EXPECTED* CONCEMTRATION .--.. ACTUAL CONCENTRATION FIG. 1. Graph of "expected" and actual COIICPIL- trations of a hTpotheticn1 permeable labelled sub- ntauce illust~ratmg the te~~~~inolog~ employed. See " Definition of Terms " sectiou "ilk tllc test for details. plotted on semi-log paper as indicated in Fig. 1. 2. Eq~~i~~~~~~~~ Tinze. It will be seen from Fig. 1 that values of actual concentration at first lie below and later cross over to remain above the curve of expected concentrations. The difference between, the expected and ac- tual concentrations to the left or prior to the point of crossing represents net losses from blood to the extravascular tissues whereas the difference between the actual and expected concentrations to the right of the crossing in- dicates net return from the tissues to the blood. The point of crossing therefore repre- sents an equilibrium and the time from ap- pearance to the point of crossing is called the Equilibrium Time of the respective permeable tracer. It also represents the total period of net transcapillary loss. 3. Halj-Return Time. The area enclosed by the respective curves of expected and actual concentrations between the appearance and the point of equilibrium is proportional to the total net loss in the vas- cular bed drained by the effluent vein used for sampling. The latter is calculated by in- tegrating the net losses per unit time from ap- pearance to equilibrium time (expected minus actual concentrations per unit time). The net returns per unit time (actual minus ex- pected concentrations beginning at the equili- brium point) then are added together until the sum equals half the total net loss. Since TKIOCVANATE AND HEAVY WATER IN THE FOREARM net loss begins at the equilibrium point the time of half return minus the time of the equilibrium point is designated as the half- return time. It represents the duration of time in seconds from the beginning of net re- turn to completion of half-return. Since there is no significant recirculation of the tracer materials it is possible to estimate the return to the blood stream of the substances which have permeated into the tissues. Following the passage of the main bolus of injectate through the capillaries of the forearm the subsequent blood entering this vascular bed contains negligible quantities of the perme- able tracer materials. This produces a favor- able gradient across the capillary wall for de- termining the rate of return of the permeable tracer substances to the circulation. Results. Early per cent loss of thiocyanate and deuterium oxide. During the early por- tion of the transit curves of the labelled ma- terials up to and including the peak values the concentrations in the blood are continu- ously rising and exceed those in the tissues. Thus, during this period there is a gradient between blood and tissues favoring outward migration of the permeable substances and opposing their inward return. The percentage loss of the SCN and D20 during this early period, therefore, probably represents a fair approximation of the permeability of the cap- illary wall to these substances. The percent- age losses of SCN and D20 were determined by dividing the difference between the ex- pected and actual concentrations by the ex- pected concentrations ( 1) . The results listed in Table I represent the percentage losses at the peak. For thiocyanate the mean trans- capillary loss was 49 % 19%. For deute- rium oxide the mean loss was 90 + 4.15%. These values are not materially different from those previously reported in a smaller series (1). Equilibrium time of thiocyanate. The up- slope and early downslope of the SCN con- centration curve paralleled the T-1824 curve. However, after the early downslope SCN de- viated to produce a washout slope with a smaller gradient than that of T-1824. When the expected concentrations of thiocyanate FIG. 2. (Upper graph.) The curves of expected (solid line) and actual (broken line) coneentra- tions are plotted against time. Simultaneously de- termined curves for deuterium oxide are plotted in lower graph. 19th line of Table I. were plotted against the actual concentrations the curves crossed at a point in their down- slopes as shown in Fig. 2a. This crossing represents the time at which the net loss of thiocyanate equals the net gain (see under "definition of terms," above). The equili- brium time or period of net loss for SCN in the 22 subjects averaged 111 -t 66 seconds (Table I). The equilibrium time for SCN varied directly with th'e mean circulation time (T) of T-1824. When the equilibrium times of SCN were plotted against the respective mean circulation times of T-1824 the points were grouped along the line shown in Fig. 3a. The formula for this line indicated that the equilibrium time of SCN was approximately equal to 3 times the mean circulation time of THIOCYANATE AND HEAVY WATER IN THE FORELXRM TABLE 1. Per Cent Losses, Equi~briun~ Times and Half-Return Times for Thioeyanate and Deute- rium Oxide in 22 Subjects. ,-Thiocyimate-, r- Deuterium oxide -1 TTelm,$ 5% loss E.T.,* Age sec. at peak sec. OJRT, 0.5RTt ?T;Wsk E.T., 0.5RT, 9;E:T sec. /E.T. 1. sec. sec. . . Comment -___-.---_ ___-- --- 27 28 30 30 33 43 31 31 28 27 34 30 9, 31 40 36 45 33 32 38 30 28 24 72 68 35 61 79 70 27 69 78 25 50 155 1596 1.3 2.0 10.3 87 24 30 1.25 2:: 567 69 iii:; 29 53 125 49 51 56 27 215 21 98 43 151 103 265 1218 38 283 113 1.6 97 .7 83 4.9 6.8 8.4 1.6 I 2; so 61 55 63 80 38 44 92 56 5: 145 24 193 39 203 587 2.9 Cephalic vein drainage Median vein drainage Hand circulation O&Y Idem Forearm only Idem 65 46 106 147 1375 1.4 11.0 41 50 125 67 40 169 224 1.4 35 56 33 92 2.8 114 67 220 418 1.9 102 61 160 224 32 56 36 209 70 48 146 $ 90 42 185 389 39 4G 57 97 61 54 so 216 1.4 87 5.8 91 185 518 2.8 ,* 85 196 2.3 ,I w 2.1 1.7 91 2.4 91 v* 158 1.9 ,* 320 2.2 ,I Mean 61 49 111 370 3.0 90 130 306 2 & .6 S.D. 329 +19 266 *lSQ t3.0 24 `-t86 -4256 - .-.- -.--__ * E.T. = Eauilibriur~~ time. f 0.5 RT = Half-return time. $ Too long to calculate. 0 T,,,,, I Mean circulation time of T-1824. the plasma (T-1824) minus 75 seconds. ,411 of the values fell within the range of 3TT.lsaa-- (0 to 150 seconds). Half-return time of thibcyanate. The half- return time of SCN was determined in each case according to the method outlined under "definition of terms," above. The mean half- return time for SCN in the 22 subjects was 370 t 199 seconds (Table I). The period from appearance to half-return time averaged approximately 7 to 8 minutes. Since the equilibrium time represents the period during which net loss occurred it was of interest to compare it to the half-return time. In 20 cases the half-return time for thiocyanate ranged between 0.7 to 10.3 times the equili- brium or loss time with a mean of 3 t 3.0 (Table I). Thus, even with a favorable con- centration gradient the half return of this ex- tracellular substance to the circulation aver- aged 3 times longer than the period of trans- capillary loss. Equilibrium and half-return tames of heavy water. The great facility with which heavy water passes through ~mi-permeable mem- branes was shown not only by the early high percentage losses of this substance but also by the curves of expected and actual concen- trations. In Fig. 2b and in most of the other cases as well the peak of the curve of actual concentrations occurred approximately 10 seconds later than the peak of the expected. This was interpreted as indicating a massive buildup of Da0 outside the capillaries at the time of peak passage of the bolus of injectate followed by rapid re-entry of a portion of this extravascular DzO. This delayed peak of ac- tual concentrations never was observed with SC% in the forearm circulation suggesting more hindrance to back diffusion of SCN than of D?O. The larger space into which D20 permeates was indicated by the some- THIOCYANATE AND HEAVY WATER IN THE FOREARM Tea. SCN _ x)0- 200- too- 0 50 100 150 200 Tt-1824 Teq. _ p20 FIG. 3. (Upper graph.) The equilibrium times for thioc;vauate are plotted against the respective meau circulation times of T-1824. The equation of the milltlle liue is T,,;f.x = 3T,,,, minus 75 see- on&. (Lower graph.) Similar values for D?O are l)lottcd. The equation of this line is Tegn,,, = 2.5 T Tm,R21 minus 31 seconds. I what longer equilibrium and half-return times for DzO than for SCN. The equilibrium times of D-0 averaged 130 -f 86 seconds (Table I). These were related to the mean circula- tion times of the plasma as is indicated in Fig. 3b. Equilibrium time D20 averaged 2.5 TT.ls21-31 seconds (range -7.5 to 70 sec- onds). In 7 cases SCN and DZO were in- jected simultaneously. In these the equi- librium times DsO averaged only 35 per cent more than the equilibrium times SCK. This again indicates the remarkable diffusion char- acteristics of heavy water through semi-per- meable membranes since D20 passes through cellular membranes and its extravascular space is approximately three times the SCN extravascular space. In the 7 cases in which SCN and D20 were injected simultaneously the half-return time of thiocyanate averaged 175 seconds and the half-return time of heavy water averaged 345 seconds. Thus, the mean half-return time D?O was approximately twice as great as that of SCN. The mean half-return time of D20 in the 8 cases studied was 306 + 256 seconds (Table I). This represented 2.1 + 0.6 times the equilibrium time. Although these values are somewhat lower than the corresponding means for SCN, inspection of Table I will show that the latter were heavily weighted by a few cases with greatly prolonged half-return times. In all of the 8 cases the time from appearance to half-return of D,O was less than 10 min- utes. Discussion. The validity of the method for determining percentage loss of the per- meable tracers has been discussed previously ( 1). The method now has been extended to provide information on the rate of net ex- change in both directions across the capillary membrane. The deviations of the actual from the ex- pected concentrations can only be explained by loss or gain to the circulating blood. Part of this loss could be into the walls of the blood vessels themselves which still would represent penetration into or through an endothelial membrane. It seems probable such losses are negligible compared to the transcapillary ex- change. The method does not provide information on the total exchanges in the forearm since the distribution of the labelled materials may not be evenly dispersed throughout all of the forearm vessels( 5). However, since imper- meable and permeable tracer substances are completely mixed prior to injection it is valid to assume that they will be delivered to the capillary beds of the area being sampled in the same relative concentrations as were present in the injectate. The present state of knowledge regarding capillary permeability has been reviewed re- cently by Pappenheimer (6). Comparison of the present results with those reported previ- THIOCYANATE AND HE.~VY WATER IN THE FOREARM ously in the literature is difficult because of differences in approach and technic. Much of the prior work has attempted to express the permeability characteristics of biological systems using the terminology employed in the physical sciences. The convention of using permeability constants representing the number of moles of a substance which cross unit cross-sectional area of the membrane in unit time under unit concentration difference may be useful in model systems where these variables are known. However, in biological systems they are not known and the attempt to estimate them introduces questionable as- sumptions and difficult complexities. In regard to the organ systems of man we essentially are desirous of knowing (1) the magnitude of transcapillary loss of various substances and the factors which influence them, (2) the extravascular distribution of these substances, and (3) their "half life" in the tissues or their rate of return to the blood. Progress in this field especially in relation to the study of diseas'e states depends upon the development of relatively direct and simple methods. Such methods need not mimic the approach used by the purely physical sciences where the experimental circumstances are completely different. The net transcapillary loss of heavy water was approximately 90% and its return to the circulation was relatively rapid. The high rate of loss is consistent with previous obser- vations from this laboratory and also with those of Chinard( 7). Such high percentage losses and rapid returns are consistent with a process of diffusion rather than "filtration in hulk"(8). The half-return time of water was short when one considers that it also pene- trates into cells. The present data suggest that more than half of the tissue water ex- changes across the capillaries several times in an hour. It seems likely that this great movement which must include intracellular water serves as the medium through which much of the cellular metabolism is accom- plished. Thiocyanate met with more hindrance to free diffusion than did heavy water. This is consistent with the concept of restricted dif- fusion as advanced by Collander (9) and Weech and Michaelis( 10) and further devel- oped by Manegold ( 11) and Pappenheimer ( 12). The early percentage losses are less and its rate of return is no more rapid than D20 despite the fact that it is limited to a space which is not only smaller but also is in more intimate contact with the capillaries. Nevertheless, turnover of SCN also is rapid since in most cases total net transcapillary loss was complete and half of this loss had returned to the circulation in less than 10 minutes from the time of injection. The constant relationship between the mean circulation time of T-1824 and the equi- librium times of SCN and D20 show that the net transcapillary exchanges of these perme- able substances are blood flow dependent. In- deed, it is possible to predict the approximate equilibrium time of these substances when the rate of blood flow is known. The present method offers several advan- tages over previously used technics for study- ing transcapillary exchange. The first is that the studies are conducted in normally func- tioning untraumatized tissue without use of artificial perfusion fluids. Only tracer doses of the injected materials are used which do not upset the osmotic equilibria of the blood in the capillaries under study. Another de- sirable feature is that sampling is carried out at intervals of seconds during the early phase. This is important since percentage losses must be determined during the brief period when the concentrations in the blood are higher than thase in the tissues. The third advan- tage is the use of an impermeable tracer to cancel out the effects of dilution by the blood itself. The fourth is limitation of the study to a small vascular area in order to prevent Kgnificant recirculation of the tracer sub- stances. As a result following completion of injection the blood entering the forearm will be practically free of labelled materials pro- ducing favorable concentration gradients for studying net rates of return of the substances under study. Thus, rate of net loss and rate of net return can be compared in a single experiment. Finally, the method is applicable to any organ or body area where an afferent THIOCYANATE AND HEAVY \~ATER IN THE FOREARM and efferent vessel are available for injection and sampling. A modification applicable to animals for preventing significant recircula- tion in organs with a large blood flow (such as the kidney) will be described in a subse- quent communication. Summary and conclusions. A method is presented for determining under physiological conditions the net bidirectional exchange of permeable tracer substances across the capil- lary walls of the human forearm. 1. In the early period, maximum net transcapillary loss for thiocyanate was 49 k 19% and for deu- terium oxide 90 & 4%. 2. The equilibrium time or period of net loss averaged 111 k 66 seconds for SCN and 130 i 86 seconds for DZO. Equilibrium times for both SCN and D-0 were related directly to blood flow. 3. The half-return time for SCN averaged 3 F 3.0 times the equilibrium time. In the 1. Freis, E. D., Higgins, T. F., and Morowitz, H. J., J. Applied Physiol., 1953, v5, 526. 2. Lilienfield, L. S., Freis, E. D., Partenope, E. A., and Morowitz, H. J., J. Clin. Invest., 1955, ~34, 1. 3. Freis, E. D., Schnaper, H. W., Lilienfield, L. S., Wilson, I. M., and Kovach, R. D., Circulation, 1955, 1712, 508. 4. Freis, E. D., Stanton, J. R., and Emerson, C. P., Am. J. Physiol., 1949, vlj7, 153. 5. Andres, R., Zierler, K. L., Anderson, H. M., Stainsby, W. N., Cader, G., Chrayyib, A. S., and Lilienthal, J. L., Jr., J. C&n. Invest., 1954, 1733, 482. 6. Pappenheimer, J. R., Physiol. Rev., 1953, ~33, 387. 7. Chinard, F. P., Vostburgh, G. J., and Enns, T., Am. J. Physiol., 1955, ~183, 221. 8. Saperstein, L. A., Buckley, N. M., and Ogden, E., ibid., 1955, ~183, 1iS. 9. Collander, R., Kolloid-Berhefte, 1924, ~19, 72. 10. Weech, A. 4., and Michaelis, L., J. Gen. Physiol., 1928, ~12, SS. 11. Manegold, E., Kolloid Z., 1937, ~81, 164. cases in which simultaneous measurements 12. Pappenheimer, J. R., Renkin, E. M., and Bor- were made the half-return time of D?O aver- rero, L. M., Am J. Physiol., 1951, ~167, 13. aged twice that of SCN. Received April 9, 1956. P.S.E.B.M., 1956, ~92.