SUMMARY
Protein content of tissue fluid and lymph is not uniform and depends upon regional
differences in capillary permeability modified by changes in capillary filtration pressure.
Whereas increased pressure in freely permeable liver sinusoids promotes formation of
excess liver and thoracic duct lymph and ascitic fluid high in protein, increased venous
pressure in less permeable beds promotes formation of excess lymph and edema fluid
Downloaded from http://circ.ahajournals.org/ by guest on April 6, 2017
Numbers at the bottom of each bar represent number lymph of control subjects are shown in figures
of patients studied. 1 and 2.
40 ,19
29t 9
3I-1
Figure 2
Figure 3
Total protein in liver and intestinal lymph (mean + SD)
in control subjects and in patients with acute and Total protein in ascitic, pleural, and peripheral edema
chronic congestive heart failure (CHF). Numbers at fluid in acute and chronic congestive heart failure
the bottom of each bar represent number of patients (CHF). Numbers on the bottom of each bar represent
studied. number of patients studied.
Circulation, Volume XL, November 1969
626 WITTE ET AL.
Table 2
Average Levels of Albumin (A) and Globulin (G) in Plasma and Lymph (G/100 ml) Based on the Total and
Fractional Protein Content in Patients with "Acute," "Chronic," and "Compensated" Congestive Heart Failure
(CHF) Compared with Control Subjects*
Control "Acute" CHF "Chronic" CHF "Compensated" CHF
Total Total Total Total
protein %A A/G protein %A A/G protein %A A/G protein %A A/G
Liver lymph 6.2 64.2 222 8.7 48.4 420 6.3 46.7
2.9
Intestinal lymph 4.1 62.3 2.55 5.2 45.8 2.39 3.0 40.7 1.76
Downloaded from http://circ.ahajournals.org/ by guest on April 6, 2017
* Numbers in parentheses represent the number of subjects whose data have been averaged for the value
tabulated.
Table 3
Average Levels of Albumin (A) and Globulin (G) in Ascitic, Pleural, and Leg Edema Fluid
(g/100 ml) Based on the Total and Fractional Protein Content in Patients with "Acute' and
"Chronic" Congestive Heart Failure (CHF)*
"Acute" CHF "Chronic" CHF
Total Total
protein %A A/G protein %A A/G
* Numbers in parentheses represent number of subjects whose data have been averaged for
the value tabulated.
In patients with "chronic" congestive heart (P > 0.4 and P > 0.2 respectively) from val-
failure, the protein content in thoracic duct ues in "acute" heart failure (fig. 3).
lymph was 36% lower (P < 0.01) and in In patients with "compensated" congestive
intestinal lymph 31% lower (P <0.02) than heart failure, the protein level in thoracic
control values and within normal limits in duct lymph was within normal limits (fig. 1).
liver lymph (figs. 1 and 2). The protein Tables 2 and 3 represent the average levels
content in ascitic fluid and right pleural fluid of albumin and gobulin in plasma, lymOp,
was 41% lower (P < 0.01 and P < 0.02 respec- and edema fluids in patients in congestive
tively) than values in "acute" heart failure, heart failure as compared with control sub-
but the protein level in leg edema and left jects. The albumin fraction was lower in
pleural fluid was not significantly different patients with heart failure but there were no
Cisrclaiton, Volsme XL, November 1969
PROTEIN CONTENT IN LYMPH AND EDEMA FLUIDS 627
Table 4
Estimated Oncotic Pressure (H in mm Hg) in Plasma (H P), Liver (HLL), Extrahepatic Portal (IL) and
Peripheral Tissues (He) in Control Patients and in Those with "Acute" and "Chronic" Congestive Heart Failure
(CHF) Calculated from Albumin and Globulin Concentrations (g%)* in Plasma, Liver Lymph, Intestinal
Lymph, and Peripheral Interstitial Fluid (Tables 2 and 3)
Effective IIP
Control "Acute" CHF "Chronic" CHF Control "Acute" CHF "Chronic" CHF
HP 19.6 23.6 18.3
IILL 19.1 26.4 17.4 Liver (HiP - ILL) +0.5 -2.8 +0.9
HlIL 11.4 13.6 7.1 Intestine (IIP - IL) +8.2 +10.0 +11.2
He 4.8t 0.7 0.8 Leg (iiP - He) +14.8 +22.9 +17.5
* Based on van't Hoff's law corrected for Donnan effects and protein-protein interaction'4
H albumin = 2.8c + 0.18c2 + 0.012c3
H globulin = 1.6c + 0.15c2 + 0.006c3
(c = g%).
t Estimated from animals. 15
Downloaded from http://circ.ahajournals.org/ by guest on April 6, 2017
essential differences between the "acute" and low-protein lymph.3 16 When the rate of
"chronic" stages. Note that there is a slight formation of capillary filtrate exceeds the rate
discrepancy between the average values in of lymph reabsorption, low-protein edema
total protein shown in figures 1 to 3 because fluid accumulates.7' 17
not all samples were fractionated for albumin. The liver, however, does not behave in this
Table 4 represents estimated oncotic pres- fashion. Hepatic sinusoids are perfused pri-
sure in the plasma, liver, intestine, and lower marily by low-pressure portal venous blood
extremity in control patients and in those with and are lined by a very tenuous endothelium
congestive heart failure. Whereas venous with large gaps,'8 allowing molecules the size
hypertension in heart failure immediately of plasma protein to pass freely.6 18 As a
tends to shift fluid into the liver, an increase in result, there is little or no oncotic gradient
effective plasma oncotic pressure in extrahepa- across the sinusoidal wall. Small increments in
tic portal and peripheral capillaries requires a the normally low venous pressure produce a
rise of at least 4 mm Hg in filtration pressure great increase in production of hepatic capil-
to produce a net fluid flux into these tissues. lary filtrate and hepatic lymph high in
Discussion protein.'9 20 Accordingly, in dogs with con-
striction of the inferior vena cava above the
Starling recognized the importance of plas- diaphragm and in patients with "early"
ma oncotic pressure in regulating capillary hepatic cirrhosis, restriction to hepatic venous
fluid exchange; he reasoned further that outflow greatly increases flow of high-protein
elevation of capillary hydrostatic pressure lymph from periportal hepatic lymphatics and
would lead to a fall in tissue oncotic pressure: the thoracic duct.20 2' When transfer of fluid
"With increased capillary pressure there must out of the sinusoid exceeds the capacity of
be increased transudation until equilibrium is hepatic lymphatics to absorb the fluid, excess
established at a somewhat higher point, when liver lymph spills into the peritoneal cavity to
there is a more dilute fluid in the tissue spaces
and therefore a higher absorbing force to form ascitic fluid high in protein.20 21 In
balance the increased capillary pressure."' contrast, in dogs with an aorto-portal vein
In general, mild elevations in venous shunt combined with restriction to transhepa-
pressure are initially counterbalanced by a tic portal blood flow, and in patients with far-
reduction in tissue oncotic pressure (table 4), advanced hepatic cirrhosis, marked elevation
but a further sustained rise in venous pressure in extrahepatic portal pressure (>30 cm of
shifts fluid into tissues and increases flow of saline) promotes the formation of large
Circulation, Volume XL, November 1969
628 WITTE ET AL.
amounts of both mesenteric and thoracic duct levels in intestinal and thoracic duct lymph
lymph very low in protein.21'22 When transfer and ascitic fluid seldom fall to the very low
of fluid out of splanchnic capillaries exceeds levels characteristic of far-advanced hepatic
the capacity of mesenteric lymphatics to cirrhosis in which portal hypertension is much
absorb the fluid, excess extrahepatic portal more prominent. The difference in portal
lymph very low in protein accumulates in the circulatory dynamics between cirrhosis and
peritoneal cavity.21' 22 congestive heart failure probably relates to
Failure of the heart to pump blood concomitant alterations in splanchnic arterial
effectively into the arterial circulation leads to blood flow. Whereas patients with cirrhosis
venous hypertension. The delicately balanced have normal or increased cardiac output28 in
liver sinusoid responds with a great outpour- conjunction with reduced transhepatic portal
ing of liver lymph high in protein that floods blood flow,29 patients with heart failure have
the thoracic duct. When the formation of decreased effective cardiac output3' 31 partial-
excess liver lymph is more rapid than its ly offsetting the tendency for portal pressure
reabsorption, ascitic fluid high in protein to rise from increased resistance to transhepa-
accumulates. (As in caval-constricted dogs, tic portal flow.
Downloaded from http://circ.ahajournals.org/ by guest on April 6, 2017
protein values in ascitic fluid are somewhat Controversy exists concerning the mecha-
lower than in liver lymph after osmotic nism of edema formation in congestive heart
equilibration with interstitial fluid in the failure-that is, does edema result primarily
extrahepatic portal bed).19 23 On the other from renal salt and water retention,3236
hand, venous hypertension outside the liver lymphatic obstruction,37 or changes in capil-
increases filtration of water more than pro- lary permeability due to decreased cardiac
tein.24 When regional lymphatics are over- output and hence arterial perfusion,31'37 or is
whelmed, peripheral edema fluid very low in it a direct result of systemic venous hyperten-
protein and pleural fluid relatively low in sion?38 39 Ischemia or trauma to capillaries as
protein result.* well as lymphatic obstruction alone results in
Starling believed that in the early stages of edema fluid that is consistently high rather
heart failure forced outflow of plasma from than low in protein.40'41 Plasma simply "leaks"
the venous side of the circulation into the liver into tissues in bulk or fails to be removed.
substance and its lymphatic network protected Lymph stasis, however, does eventually devel-
the failing heart and at the same time buffered op in cardiac failure as lymphatics are
the circulation outside the liver from venous overloaded and lymph flow into high pressure
hypertension.27 The data suggest, however, central veins is impeded.445 But the fact
that as this compensatory mechanism becomes that ascitic fluid from the liver is high in
insufficient, excess lymph can no longer return protein, and that edema fluid in other areas is
to central veins with elevated pressure as fast low indicates that the primary driving force is
as it forms, and resistance to transhepatic venous hypertension resulting from failure of
portal blood flow rises. Mesenteric venous forward venous flow. Retention of salt and
hypertension develops, and protein content in water further expands extracellular fluid vol-
extrahepatic portal lymph falls. Protein con- ume and imposes an ever increasing burden
centration of thoracic duct lymph also de- on an already overworkedb lymphatic circula-
creases, reflecting the greater contribution of tion. On the other hand, treatment with low-
extrahepatic lymph. However, the protein salt diet and diuretic drugs contracts extracel-
lular fluid volume and reduces the load on the
*In recumbent animals, protein content of leg lymphatic circulation.
lymph is normally 20 to 30% of plasma level,15 and In congestive heart failure, the protein
protein content in pleural fluid in diseases not content of lymph and edema fluids reflects the
characterized by venous hypertension (for example,
carcinoma and tuberculosis) is 60 to 80% of plasma hydrostatic pressure in different capillary beds
level.25. 26 and allows an estimation of effective plasma
Circulation, Volume XL, November 1969
PROTEIN CONTENT IN LYMPH AND EDEMA FLUIDS 629
oncotic pressure, a crucial unknown in Star- RADER, BERTHA, LEVINE, NoRnMAN, AND BREED,
lings law of transcapillary fluid exchange. E. S.: Lymph circulation in congestive heart
failure: Effect of extemal thoracic duct
High-protein ascitic fluid, relatively low-pro- drainage. Circulation 39: 723, 1969.
tein pleural fluid and very low protein 13. WITTE, C. L., WITTE, M. H., COLE, W. R.,
peripheral edema fluid arise primarily in CHUNG, Y. C., BLEISCH, V. R., AND DUMONT,
response to venous hypertension, which favors A. E.: Dual origin of ascites in hepatic
capillary filtration and eventually impedes cirrhosis. Surg Gynec & Obstet. In press.
14. LANDIS, E. M., AND PAPPENHEIMER, J. R.:
lymphatic drainage. Exchange of substances through the capillary
walls. In Handbook of Physiology, Section 2,
References Circulation, vol. 2, edited by W. F. Hamilton
1. STARLING, E. H.: On the absorption of fluids and P. Dow. Washington, American Physiolo-
from the connective tissue spaces. J Physiol gic Society, 1963, p. 961.
(London) 19: 312, 1896. 15. YOFFEY, J. M., AND COURTICE, F. C.: Lymphatics
2. STARLING, E. H.: The Fluids of the Body. and Lymphoid Tissue. Cambridge, Massachu-
Chicago, University of Chicago Press, 1909, p. setts, Harvard University Press, 1956, p. 238.
67. 16. IBM, p. 86.
3. WIEDERHIELM, C. A.: Dynamics of transcapil- 17. FAHR, G., AND ERSHLER, I.: Studies of the factors
Downloaded from http://circ.ahajournals.org/ by guest on April 6, 2017
lary fluid exchange. In Biological Interfaces: concerned in edema formation: II. Hydrostatic
Flow and Exchanges. Proceedings of a pressure in the capillaries during edema
Symposium sponsored by the New York Heart formation in right heart failure. Ann Intern
Association. Boston, Little, Brown and Co., Med 15: 798,1941.
1968, p. 29. 18. BRAUER, R. W.: Liver circulation and function.
4. WOLF, A. V.: Volume and pressure imbalances in Physiol Rev 43: 115, 1963.
the genesis of edema. In Edema: Mechanisms 19. HYATT, R. E., LAURENCE, G. H., AND SMITH, J.
and Management, edited by J. H. Moyer and R.: Observations on the origin of ascites from
M. Fuchs. Philadelphia, W. B. Saunders Co., experimental hepatic congestion. J Lab Clin
1960, p. 23. Med 45: 274, 1955.
5. STEAD, E. A., AND WARREN, J. V.: Protein content 20. HYATT, R. E., AND SMITH, J. R.: Mechanisms of
of the extracellular fluid in normal subjects ascites: Physiologic appraisal. Amer J Med
after venous congestion and in patients with 16: 434, 1954.
cardiac failure, anoxemia and fever. J Clin 21. WITTE, C. L., WITTE, M. H., DUMONT, A. E.,
Invest 23: 283, 1944. FRIST, J., AND COLE, W. R.: Lymph protein in
6. MAYERSON, H. S.: - Physiologic importance of hepatic cirrhosis and experimental hepatic and
lymph. In Handbook of Physiology, Section 2, portal venous hypertension. Ann Surg 168:
Circulation, vol. 2, edited by W. F. Hamilton 567, 1968.
and P. Dow. Washington, D. C., American 22. WITTE, C. L., CHuNG, Y. C., WrrrE, M. H.,
Physiologic Society, 1963, p. 1035. STERLE, 0. F., AND COLE, W. R.: Observations
7. STARLING, E. H.: Physiological factors involved on the origin of ascites from experimental
in the causation of dropsy. Lancet 1: 1267, extrahepatic portal congestion. Ann Surg. In
1896. press.
8. WmTm, J. C., FIELD, M. E., AND DINKER, C. 23. COURTICE, F. C., AND STEINBECK, A. W.:
K.: On the protein content and normal flow of Absorption of protein from the peritoneal
lymph fromn the foot of the dog. Amer J cavity. J Physiol (London) 114: 336, 1951.
Physiol 103: 34, 1933. 24. RUSZNYAK, I., FOLDI, M., AND SzABo, G.:
9. LANDIS, E. M., JONES, L., ANGEVINE, M., AND Lymphatics and Lymph Circulation: Physiol-
ERB, W.: Passage of fluid and protein through ogy and Pathology, ed. 2. New York,
the human capillary wall during venous Pergamon Press, 1967, p. 20.
congestion. J Clin Invest 11: 717, 1932. 25. JAMES, A. H.: Mechanism of pleural and ascitic
10. PAPPENHEIMER, J. R., AND SOTO-RIVERA, A.: effusions, with a suggested method for the
Effect of osmotic pressure on the plasma indirect estimation of portal venous pressure.
protein and other quantities associated with the Clin Sci 8: 291, 1949.
capillary circulation in the hind limb of cats
and dogs. Amer J Physiol 152: 471, 1948. 26. LUETSCHER, J. A., JR.: Electrophoretic analysis of
11. DUMONT, A. E., CLAuss, R. H., REED, G. E., AND the proteins of plasma and serous effusions. J
TICE, D. A.: Lymph drainage in patients with Clin Invest 20: 99, 1941.
congestive heart failure. New Eng J Med 269: 27. STARLING, E. H.: Some points on the pathology
949, 1963. of heart disease (The Arris and Gale
12. WITTE, M. H., DUMONT, A. E., CLAuss, R. H., Lectures). Lancet 1: 652, 1897.
Circulation, Volume XL, November 1969
630 WITTE ET AL.
28. MURRAY, J. F., DAWSON, A. M., AND SHERLOCK, 36. BRIGGS, A. P., FOWELL, D. M., HAMILTON, W.
S.: Circulatory changes in chronic liver disease. F., REMINGTON, J. W., WHEELER, N. C., AND
Amer J Med 24: 358, 1958. WINSLOW, J. A.: Renal and circulatory factors
29. MORENO, A. H., BURCHELL, A. R., ROUSSELOT, L. in the edema formation of congestive heart
M., PANKE, W. F., SLAFSKY, S. F., AND BURKE, failure. J Clin Invest 27: 810, 1948.
J. H.: Portal blood flow in cirrhosis of the 37. PAINE, R., AND SMITH, J. R.: Mechanism of
liver. J Clin Invest 46: 436, 1967. heart failure. Amer J Med 6: 84, 1949.
30. MERRILL, A. J.: Edema and decreased renal 38. DAvis, J. O., AND SMITH, J. R.: Pathogenesis of
blood flow in patients with chronic congestive peripheral cardiac failure. Amer J Med 3: 704,
heart failure: Evidence of forward failure as 1947.
the primary cause of edema. J Clin Invest 25:
389, 1946. 39. RAY, C. T., AND BURCH, G.: Vascular response in
31. BURCH, G. E., AND RAY, C. T.: Consideration of man to ligation of the inferior vena cava. Arch
the mechanism of congestive heart failure. Intern Med (Chicago) 80: 587, 1947.
Amer Heart J 43: 918, 1951. 40. YOFFEY, J. M., AND COURTICE, F. C.: Lymphatics
32. WARREN, J. V., AND STEAD, E. A., JR.: Fluid and Lymphoid Tissue. op. cit. (ref. 24), p. 90.
dynamics in chronic congestive heart failure. 41. IBID, p. 458.
Arch Intern Med (Chicago) 73: 138, 1944. 42. FOLDI, M., AND PAPP, N.: Role of the lymph
33. STARR, I., JEFFERS, W. A., AND MEADE, R. H.,
Downloaded from http://circ.ahajournals.org/ by guest on April 6, 2017
Circulation. 1969;40:623-630
doi: 10.1161/01.CIR.40.5.623
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright 1969 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/40/5/623