Blood Purif 1992:10:115-121
Baxter Healthcare. Renal Division Research. McGaw Park. 111.. USA
Pathways for Fluid Loss from the Peritoneal Cavity
Key Words
Abstract
Peritoneal cavity Lymphatics Starling forces Dialysis Fluid loss
During peritoneal dialysis, fluid is transported out of the peri toneal cavity by lymphatic and nonlymphatic pathways, thereby decreasing net ultrafiltration by 40-50% and reducing small solute clearance by 15-20%. The direct lymphatic path way consists of the diaphragmatic lymphatics, which directly connect the peritoneal cavity to the bloodstream. The intersti tial lymphatic and direct blood entry pathways convey fluid that has been driven into the interstitial space of the tissue surrounding the peritoneal cavity by the increased intraperito neal pressure, and return it to the bloodstream. Since flow through lymphatic pathways is only a portion of the flow through all pathways, total fluid loss is greater than lymph flow. The best technique for estimating lymph flow is direct measurement by cannulation of lymphatic vessels, a tech nique that is not clinically feasible. The tracer disappearance technique, which measures the rate at which macromolecules leave the peritoneal cavity, is an indirect measure of fluid loss. The tracer appearance technique, which measures the rate at which macromolecules reach the blood from the peritoneal cavity, slightly overestimates lymph flow because some tracer may enter the bloodstream directly from the tissues. Much of the previous controversy over the contribution of the lym phatic pathways to total fluid loss can be resolved by under standing the differences in what these techniques measure.
Introduction
In peritoneal dialysis (PD), fluid is instilled into the peritoneal cavity, is held there for a specified time (dwell time), and is subse
quently drained, taking with it uremic toxins that have entered the peritoneal cavity during the dwell. The use of hypertonic solutions, which draw fluid out of the body, into the peritoneal cavity, allows control of body fluid
Ty R. Shockley, ScD Renal Division Research. Mail Code MPR-DI Baxter Healthcare Corporation 1620 Waukegan Rd. McGaw Park IL 60085 (USA)
© 1992 S. Karger AG. Basel 0253-5068/92/ 0104-0115S2.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
T.R. Shockley Norma J. Ofsthun
116
Shockiev/Ofsthun
Pathways for Fluid Loss from the Peritoneal Cavity during PD
Experimental evidence suggests that fluid leaves the peritoneal cavity via three kinds of pathways: direct lymphatic, interstitial lym phatic and direct blood entry (fig. 1). The direct lymphatic pathway consists of the dia phragmatic lymphatics, which directly connect the peritoneal cavity to the bloodstream. The interstitial lymphatic and direct blood entry pathways take fluid that is transported into the interstitial space of the tissue surrounding the peritoneal cavity and return it to the blood stream, thereby reducing net ultrafiltration. In the normal state, extravasated plasma is directly returned to the blood as a result of Starling's forces (i.e., sum of local oncotic and hydrostatic pressure differences) and indi rectly through local interstitial lymphatics; the result is maintenance of minimal levels of interstitial fluid. These two pathways can be expected to transport peritoneal fluid from the interstitial space to the blood during PD since, at the local level, the local interstitial lymphatics and the capillaries cannot distin guish this fluid from that extravasated from the local blood supply. Thus, once dialysis fluid from the peritoneal cavity has entered the interstitial space, it can either be trans ported directly back into capillaries as a result of Starling’s forces or be indirectly returned to the blood via local interstitial lymphatics. Studies have shown that intraperitoneal (i.p.) pressure increases as a function of in stilled volume [4], indicating that the support structure of the peritoneal membrane. al though somewhat compliant, is unable to completely ‘stretch’ to maintain i.p. pressure in the normal range. Cannulation experiments in sheep have shown that flow rates through lymphatic vessels that drain the peritoneal cavity increase in response to an increased instilled fluid volume [5]. This increase in
Fluid Loss from the Peritoneal Cavity
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
volume as well as clearance of uremic toxins. The difference between the volume of fluid drained from the peritoneal cavity (drain vol ume) and that instilled into it (instilled vol ume) is net ultrafiltration. Net ultrafiltration is positive if fluid is removed from the patient and negative if fluid is gained by the patient. Because small uremic toxins equilibrate dur ing the dwell time, the clearance of these tox ins is equal to the drain volume. During the dwell, fluid is transported out of the peritoneal cavity by lymphatic and non lymphatic pathways, thereby decreasing both net ultrafiltration and solute clearances. There is a general consensus that the overall magnitude of this fluid loss averages 1.0-1.5 ml/min [1,2] ( 1.4-2.1 liters/day), with a re ported range (mean ± 2 SD) of 0.3-2.7 ml/ min [3]. This mean value for the patient in overall fluid balance is approximately equal to the daily net ultrafiltration. In other words, in the average patient, fluid loss from the peri toneal cavity adversely affects the efficacy of peritoneal dialysis by reducing the potential for net fluid removal on average by 40-50%, and by reducing small solute clearance by 1520%. If fluid loss from the peritoneal cavity during PD could be reduced, smaller instilled volumes would be required to achieve equiva lent therapeutic results. Although it is generally agreed that fluid loss from the peritoneal cavity is clinically rel evant, there has been controversy in the litera ture regarding the pathways for its exit. This controversy has resulted from unorthodox use of terminology and differences in experimen tal techniques. In this paper, we describe the pathways for fluid loss from the peritoneal cavity and discuss the techniques for estimat ing flow rates through lymphatic and nonlym phatic pathways.
Fig. 1. Pathways for fluid loss from the peritoneal cavity during PD.
investigators have explained this paradox by assuming that any tracer found in the intersti tium is transported there by diffusion, while the majority of tracer and fluid leaves the peritoneal cavity together via the lymphatic system [9], However, protein transport stud ies provide strong evidence for significant convection of fluid (and tracer) through the interstitial pathway [10]. An alternative ex planation for this paradox is the nonuniform nature of the peritoneal ‘membrane’ (e.g., vis ceral vs. parietal, diaphragmatic vs. mesenter ic). Local convection through a nonuniform membrane could be either into or out of the peritoneal cavity at different locations since net fluid transport depends on the gradients of hydrostatic and osmotic pressure.
Techniques for Measuring or Estimating Flowthrough the Various Pathways
A variety of techniques can be used to esti mate total fluid loss from the peritoneal cav ity or flow rates through individual pathways. These techniques can be grouped into direct methods such as cannulation of lymphatic vessels and indirect methods such as those that employ macromolecular tracers.
II7
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
lymphatic How rate could simply be a passive effect of the increased hydrostatic pressure. Alternatively, it could result from an active biological response (e.g.. increased lymphatic pumping) to the increased pressure [6], In addition to causing increased flow rates through lymphatic vessels, an increase in i.p. pressure as a result of fluid instillation leads to a gradient in hydrostatic pressure from the peritoneal cavity to the interstitial space of the tissues surrounding the peritoneal cavity. Transport studies in rodents have shown that i.p. administered macromolecules move into the tissues adjacent to the peritoneal cavity [7. 8], Although the exact mechanism for this macromolecular transport has not been delin eated, the transport is most likely due to con vection of fluid (and entrained solutes) which results from the increased pressure driving force. Macromolecular solute concentration would remain unchanged during such fluid movement. During PD with hypertonic dialysate. fluid passes out of the blood, through the interstitium, and into the peritoneal cavity. As dis cussed above, tracer studies suggest that fluid moves from the peritoneal cavity into the interstitium. How can fluid be convected si multaneously in opposite directions? Some
Indirect Methods Since direct cannulation of lymphatic ves sels is ethically unacceptable in humans and cannot be used for measuring flow rates through nonlymphatic pathways (direct blood entry), clinical investigators have relied upon various indirect methods. These techniques use macromolecules as tracers and assume that fluid and macromolecular tracer move together, thus, movement of the tracer serves as a surrogate for fluid flow. Investigators have measured both the disappearance from the peritoneal cavity and the appearance in the blood of intraperitoneallv instilled macro molecules.
118
Shocklev/Ofsthun
Tracer Disappearance Technique Tracer disappearance from the peritoneal cavity can be used as an indirect measure of total fluid loss from the peritoneal cavity. This technique assumes that fluid and tracer always leave the cavity together, regardless of the pathway. To perform this technique, the investigator infuses into the peritoneal cavity a known volume of solution containing a tracer and immediately draws a sample to determine the initial i.p. volume (instilled volume plus residual volume). At the end of dialysis, fluid is drained from the peritoneal cavity, the drain volume is recorded, and the tracer concentration is measured. The final residual volume is determined by tracer dilu tion after addition of a known quantity of fresh (tracer-free) dialysate; the final i.p. vol ume is calculated from the sum of the drain volume and the final residual volume. Given initial and final values for i.p. volume and tracer concentration, the mass of tracer lost from the peritoneal cavity during dialysis is calculated. Then the total volume of fluid lost during the dwell is computed by dividing the mass of tracer lost by its average concentra tion during the dwell (geometric mean of the initial and final values). Finally, the rate of fluid loss through all pathways is calculated from the total volume lost by assuming that the rate is constant throughout the dwell. The assumption that fluid and tracer al ways leave the peritoneal cavity together may be incorrect. Struijk et al. [13] have shown that the concentration of dextran in hypoosmotic i.p. fluid increases with time. This sug gests that fluid leaves the peritoneal cavity at a somewhat higher rate than tracer, as would be expected if macromolecular transport is hindered by the mesothelium. In the past, some investigators have pre sumed that all fluid exits the peritoneal cavity via lymphatic pathways, and have used the tracer disappearance technique to estimate
Fluid Loss from the Peritoneal Cavity
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
Direct Methods Currently, there is no direct method for measuring total fluid loss from the peritoneal cavity, and the only individual pathway whose flow rate can be measured directly is uptake through lymphatic vessels. In large an imals like sheep, lymph How rates have been measured by cannulation [5, 11]. Unfortu nately, each small lymphatic vessel that drains the peritoneal cavity cannot be cannulated, so instead, investigators cannulate larger vessels further downstream from the peritoneal cavity. However, since the lym phatics that drain the peritoneal cavity are fed not only from the peritoneal space but also from other tissues and organs, the actual lymph flow rate from the peritoneal cavity is estimated using i.p. macromolecular tracers (by monitoring the mass of i.p. tracer that en ters a given lymph compartment and dividing by the concentration of the tracer in the peri toneal cavity) [11], One drawback with this technique is that the cannulation process may affect the flow rate [ 12], Moreover, since can nulation is not ethically feasible in humans, use of this technique to estimate lymph flow during PD requires extrapolation from ani mal data.
dure for the tracer appearance technique is similar to that for the tracer disappearance technique, but in this case both blood and dialysate samples are collected for determina tion of tracer concentration. Over a given time interval, the apparent lymph flow rate is estimated by dividing the change in mass of tracer in the plasma by the average i.p. tracer concentration during that time interval. As suming that the tracer appearance rate is con stant, the change in mass of tracer over the course of an entire experiment is simply cal culated by multiplying plasma volume by the final concentration of tracer in the plasma. Thus, lymph flow rate can be estimated by multiplying the final concentration of tracer in the plasma by the plasma volume, and di viding the result by the average i.p. tracer con centration (geometric mean of the initial and final values). Although not discussed here, some investigators correct for tracer that is lost from the blood by distribution into other body compartments during the course of the experiment [15], The assumption that fluid and tracer ap pear in the blood only as a result of movement through lymphatic pathways may be incorrect because tracer can enter the bloodstream through the direct blood entry pathway [ 16], The magnitude of this error is small because the capillary wall significantly hinders the passage of macromolecules. Therefore, the tracer appearance technique slightly overesti mates lymph flow.
Tracer Appearance Technique Tracer appearance in the plasma has been used as an indirect technique for estimating lymph flow from the peritoneal cavity during a single PD dwell [8, 12, 14. 15]. This tech nique assumes that fluid and tracer move out of the peritoneal cavity together and reach the bloodstream simultaneously only through lymphatic pathways. The experimental proce
Comparison oj Tracer Disappearance and Appearance Techniques One might expect that rates of tracer ap pearance would equal rates of tracer disap pearance. However, as discussed above, the tracer disappearance technique estimates to tal fluid loss, while the tracer appearance tech nique estimates (or slightly overestimates) lymph flow. Experiments which measure
119
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
‘lymph flow’ from the peritoneal cavity [2, 3]. However, as discussed above, other investiga tors have shown that both fluid and tracer macromolecules can pass into the interstitial tissues surrounding the peritoneal cavity, and can reenter the bloodstream at different rates [7, 8], Therefore, tracer disappearance from the peritoneal cavity overestimates lymph flow, but is a valid technique for estimating total fluid loss. Because tracer disappearance rates have been shown to remain constant following ap parent saturation of the interstitial space of peritoneal tissues by extended administration of i.p. tracer [9], one might conclude that the tracer disappearance technique does not mea sure the contribution of the interstitial path ways. and therefore is a valid measure of flow through the direct lymphatic pathway. This would be true if diffusion were the only mech anism by which tracer could be transported into and through the interstitium, and if fluid could not reenter the bloodstream (indepen dent of tracer) as a result of Starling’s forces. However, saturation of a tissue with a tracer does not prevent convection of fluid and tracer into and through the tissue. With con vection, some tracer and fluid would be re turned to the blood via interstitial lymphatics, but other fluid could enter blood capillaries directly, leaving behind its tracer. Thus, even with apparent saturation of the interstitium, tracer disappearance is not a measure of lymph flow.
120
Shockley/Ofsthun
in tracer appearance and disappearance rates can be accounted for by distribution of tracer in the interstitial tissues surrounding the peri toneal cavity. Particulate Tracers While the above discussion has focused on the use of macromolecular tracers, particulate tracers (e.g., labeled RBC) have also been used [8. 15]. The rates of disappearance and appearance of particulate tracers arc similar in value because these tracers cannot move into and through the interstitial pathway dis cussed above. The small observed differences (disappearance rates approximately 30% higher) can be explained by hindered passage of particulates through the lymphatic system [8], Since particulates leave the peritoneal cavity primarily via the direct lymphatic pathway, the disappearance rate of particu late tracers provides an estimate for flow through the direct lymphatic pathway, not total fluid loss. Modeling Various mathematical models have been developed for simulating fluid and solute transport in PD patients [19. 20]. These mod els assume that membrane hydraulic perme ability and lymph flow rate vary from patient to patient, but assume that other parameters (e.g.. hydrostatic pressure) are uniform among all patients. By fitting experimentally determined i.p. volume profiles, estimates for membrane hydraulic permeability and lymph flow rate are obtained. A major limitation of these models is the assumption that blood vessels are juxtaposed to the peritoneal cavi ty. with no interstitial tissue between them. Therefore, these models cannot predict fluid flow in the tissue space between the peritoneal cavity and the blood.
Fluid Loss from the Peritoneal Cavitv
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
rates of both tracer appearance and tracer dis appearance have shown that macromolecular tracers disappear from the peritoneal cavity at least twice as fast as they appear in the blood [8. 12, 15]. This difference in rates can be explained by the dependence of the results on the control volume, i.e., the ‘compartment' for analysis. In the tracer disappearance tech nique. the control volume is the peritoneal cavity, and therefore this technique does not differentiate among the possible pathways for tracer exit from the peritoneal cavity. On the other hand, in the tracer appearance tech nique, the control volume is the blood space, and therefore this technique only accounts for pathways through which tracer enters the blood. The difference between control vol umes using tracer disappearance and appear ance techniques is the volume of the intersti tial tissues that surround the peritoneal cavity. Can the difference in control volumes ac count for observed differences in tracer ap pearance and disappearance rates? In other words, is the missing tracer located within some fraction of the interstitial tissue volume surrounding the peritoneal cavity? Assuming a difference of 1 ml/min, 240 ml of fluid-con taining tracer at its i.p. concentration must be accounted for over the course of a 4-hour dwell. If this fluid (and its accompanying trac er) were distributed in interstitial tissue hav ing an average volume fraction of 30% (upper estimate) available to protein [17], it would occupy a tissue volume of 800 ml. Dividing by a peritoneal surface area of 1.75 m2 in humans [17, 18] yields an average tracer pene tration depth of approximately 450 pm. This calculated depth is similar to that observed in animal experiments [7] measuring tracer pen etration into the tissues surrounding the peri toneal cavity, as expected given the similari ties in structure among mammalian tissues. Thus, it appears that the observed differences
Conclusion
In this paper, we have described the path ways that transport fluid out of the peritoneal cavity during PD. The direct lymphatic path way consists of the diaphragmatic lymphatics, which directly connect the peritoneal cavity to the bloodstream. The interstitial lymphatic and direct blood entry pathways convey fluid that has been driven into the interstitial space of the tissue surrounding the peritoneal cavity
by the increased i.p. pressure, and return it to the bloodstream. Since flow through lym phatic pathways is only a portion of the flow through all pathways, total tluid loss is greater than lymph flow. The net effect of fluid loss through all pathways is a reduction in net ultrafiltration and small solute clearances. Strategies that reduce fluid loss (through any pathway) would therefore reduce required in stilled volume and improve PD as a therapy for end-stage renal disease.
References 8 Nagy JA: Lymphatic and non-lym phatic pathways of peritoneal ab sorption in mice: Physiology versus pathology. Blood Purif 1992:10: 148-1621 9 Struijk DG. Koomen GCM. Krediet RT, Arisz L: Indirect measurement of lymphatic absorption in CAPD patients is not influenced by trap ping. Kidney lnt 1992:41:16681775. 10 Flessner MF: Net ultrafiltration in peritoneal dialysis: Role of direct fluid absorption into peritoneal tis sue. Blood Purif 1992:10:136-147. 11 Tran LP. Rodela H. Abernethy NJ. Yuan Z-Y. Hay JB. Oreopoulos D. Johnston MG: Lymphatic drainage of hypertonic dialysis solution from the peritoneal cavity: Comparison between anesthetized and conscious sheep. Am J Physiol, in press. 12 Johnston MG: Studies on lymphatic drainage of the peritoneal cavity in sheep. Blood Purif 1992:10:122131. 13 Struijk DG. Krediet RT. Koomen GCM, Arisz L: Fluid kinetics in CAPD patients during dialysis with a hvpo-osmotic solution. Periton Dial Int 1992:12:164.
14 Rippe B. Stelin G. Ahlmen J: Lymph flow from the peritoneal cavity in CAPD patients; in Maher JF, Winchester JF (eds): Frontiers in Peritoneal Dialysis. New York, Field, Rich, 1986,pp 24-30. 15 Flessner MF. Parker RJ. Sieber SM: Peritoneal lymphatic uptake of fi brinogen and erythrocytes in the rat. Am J Physiol I983:244:H89-H96. 16 Bent-Hansen L, Svendsen JH: Tis sue to plasma capillary permeability of ,5ll-albumin in the perfused rab bit ear. Microvasc Res 1991:41: 141-148. 17 Bell DR. Watson PD. Renkin EM: Exclusion of plasma proteins in interstitium of tissues from the dog hind paw. Am J Physiol 1980:239: H532-H538. 18 Henderson LW: The problem of peritoneal membrane area and per meability. Kidney lnt 1973:3:409—
410. 19 Rippe B. Zakaria ER: Lymphatic vs. non-lymphatic fluid absorption from the peritoneal cavity' as related to the peritoneal ultrafiltration ca pacity and sieving properties. Blood Purif 1992:10:189-202. 20 Vonesh EF. Rippe B: Net fluid ab sorption under membrane transport models of peritoneal dialysis. Blood Purif 1992:10:209-226.
121
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
1 Stelin G. Rippe B: A phenomeno logical interpretation of the varia tion in dialysate volume with dwell time in CAPD. Kidney Int 1990:38: 465-472. 2 Krediet RT, Struijk DG, Koomen GCM, Arisz L: Peritoneal fluid ki netics during CAPD measured with intraperitoneal dextran 70. ASAIO Trans 1991;37:662-667. 3 Mactier RA. Khanna R, Twardowski Z. Moore H. Nolph KD: Contribution of lymphatic absorp tion to loss of ultrafiltration and sol ute clearances in continuous ambu latory peritoneal dialysis. J Clin In vest 1987,80:1311-1316. 4 Gotloib L. Mines M. Garmizo L. Varka I: Hemodynamic eflcctsofincreasing intra-abdominal pressure in peritoneal dialysis. Periton Dialy sis Bull 1981:1:41-43. 5 Drake RE. Gabel JC: Abdominal lymph flow response to intraperito neal fluid in awake sheep. Lymphology 1991:24:77-81. 6 Johnston MG. Elias R: The regula tion of lymphatic pumping. Lymphologv 1987:20:215-218. 7 Flessner MF, Fenstermacher JD. Dedrick RL, Blasberg RG: Perito neal absorption of macromoleculcs studied by quantitative autoradiog raphy. Am J Phvsiol 1985:248:H26H32.
Baxter Healthcare. Renal Division Research. McGaw Park. 111.. USA
Pathways for Fluid Loss from the Peritoneal Cavity
Key Words
Abstract
Peritoneal cavity Lymphatics Starling forces Dialysis Fluid loss
During peritoneal dialysis, fluid is transported out of the peri toneal cavity by lymphatic and nonlymphatic pathways, thereby decreasing net ultrafiltration by 40-50% and reducing small solute clearance by 15-20%. The direct lymphatic path way consists of the diaphragmatic lymphatics, which directly connect the peritoneal cavity to the bloodstream. The intersti tial lymphatic and direct blood entry pathways convey fluid that has been driven into the interstitial space of the tissue surrounding the peritoneal cavity by the increased intraperito neal pressure, and return it to the bloodstream. Since flow through lymphatic pathways is only a portion of the flow through all pathways, total fluid loss is greater than lymph flow. The best technique for estimating lymph flow is direct measurement by cannulation of lymphatic vessels, a tech nique that is not clinically feasible. The tracer disappearance technique, which measures the rate at which macromolecules leave the peritoneal cavity, is an indirect measure of fluid loss. The tracer appearance technique, which measures the rate at which macromolecules reach the blood from the peritoneal cavity, slightly overestimates lymph flow because some tracer may enter the bloodstream directly from the tissues. Much of the previous controversy over the contribution of the lym phatic pathways to total fluid loss can be resolved by under standing the differences in what these techniques measure.
Introduction
In peritoneal dialysis (PD), fluid is instilled into the peritoneal cavity, is held there for a specified time (dwell time), and is subse
quently drained, taking with it uremic toxins that have entered the peritoneal cavity during the dwell. The use of hypertonic solutions, which draw fluid out of the body, into the peritoneal cavity, allows control of body fluid
Ty R. Shockley, ScD Renal Division Research. Mail Code MPR-DI Baxter Healthcare Corporation 1620 Waukegan Rd. McGaw Park IL 60085 (USA)
© 1992 S. Karger AG. Basel 0253-5068/92/ 0104-0115S2.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
T.R. Shockley Norma J. Ofsthun
116
Shockiev/Ofsthun
Pathways for Fluid Loss from the Peritoneal Cavity during PD
Experimental evidence suggests that fluid leaves the peritoneal cavity via three kinds of pathways: direct lymphatic, interstitial lym phatic and direct blood entry (fig. 1). The direct lymphatic pathway consists of the dia phragmatic lymphatics, which directly connect the peritoneal cavity to the bloodstream. The interstitial lymphatic and direct blood entry pathways take fluid that is transported into the interstitial space of the tissue surrounding the peritoneal cavity and return it to the blood stream, thereby reducing net ultrafiltration. In the normal state, extravasated plasma is directly returned to the blood as a result of Starling's forces (i.e., sum of local oncotic and hydrostatic pressure differences) and indi rectly through local interstitial lymphatics; the result is maintenance of minimal levels of interstitial fluid. These two pathways can be expected to transport peritoneal fluid from the interstitial space to the blood during PD since, at the local level, the local interstitial lymphatics and the capillaries cannot distin guish this fluid from that extravasated from the local blood supply. Thus, once dialysis fluid from the peritoneal cavity has entered the interstitial space, it can either be trans ported directly back into capillaries as a result of Starling’s forces or be indirectly returned to the blood via local interstitial lymphatics. Studies have shown that intraperitoneal (i.p.) pressure increases as a function of in stilled volume [4], indicating that the support structure of the peritoneal membrane. al though somewhat compliant, is unable to completely ‘stretch’ to maintain i.p. pressure in the normal range. Cannulation experiments in sheep have shown that flow rates through lymphatic vessels that drain the peritoneal cavity increase in response to an increased instilled fluid volume [5]. This increase in
Fluid Loss from the Peritoneal Cavity
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
volume as well as clearance of uremic toxins. The difference between the volume of fluid drained from the peritoneal cavity (drain vol ume) and that instilled into it (instilled vol ume) is net ultrafiltration. Net ultrafiltration is positive if fluid is removed from the patient and negative if fluid is gained by the patient. Because small uremic toxins equilibrate dur ing the dwell time, the clearance of these tox ins is equal to the drain volume. During the dwell, fluid is transported out of the peritoneal cavity by lymphatic and non lymphatic pathways, thereby decreasing both net ultrafiltration and solute clearances. There is a general consensus that the overall magnitude of this fluid loss averages 1.0-1.5 ml/min [1,2] ( 1.4-2.1 liters/day), with a re ported range (mean ± 2 SD) of 0.3-2.7 ml/ min [3]. This mean value for the patient in overall fluid balance is approximately equal to the daily net ultrafiltration. In other words, in the average patient, fluid loss from the peri toneal cavity adversely affects the efficacy of peritoneal dialysis by reducing the potential for net fluid removal on average by 40-50%, and by reducing small solute clearance by 1520%. If fluid loss from the peritoneal cavity during PD could be reduced, smaller instilled volumes would be required to achieve equiva lent therapeutic results. Although it is generally agreed that fluid loss from the peritoneal cavity is clinically rel evant, there has been controversy in the litera ture regarding the pathways for its exit. This controversy has resulted from unorthodox use of terminology and differences in experimen tal techniques. In this paper, we describe the pathways for fluid loss from the peritoneal cavity and discuss the techniques for estimat ing flow rates through lymphatic and nonlym phatic pathways.
Fig. 1. Pathways for fluid loss from the peritoneal cavity during PD.
investigators have explained this paradox by assuming that any tracer found in the intersti tium is transported there by diffusion, while the majority of tracer and fluid leaves the peritoneal cavity together via the lymphatic system [9], However, protein transport stud ies provide strong evidence for significant convection of fluid (and tracer) through the interstitial pathway [10]. An alternative ex planation for this paradox is the nonuniform nature of the peritoneal ‘membrane’ (e.g., vis ceral vs. parietal, diaphragmatic vs. mesenter ic). Local convection through a nonuniform membrane could be either into or out of the peritoneal cavity at different locations since net fluid transport depends on the gradients of hydrostatic and osmotic pressure.
Techniques for Measuring or Estimating Flowthrough the Various Pathways
A variety of techniques can be used to esti mate total fluid loss from the peritoneal cav ity or flow rates through individual pathways. These techniques can be grouped into direct methods such as cannulation of lymphatic vessels and indirect methods such as those that employ macromolecular tracers.
II7
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
lymphatic How rate could simply be a passive effect of the increased hydrostatic pressure. Alternatively, it could result from an active biological response (e.g.. increased lymphatic pumping) to the increased pressure [6], In addition to causing increased flow rates through lymphatic vessels, an increase in i.p. pressure as a result of fluid instillation leads to a gradient in hydrostatic pressure from the peritoneal cavity to the interstitial space of the tissues surrounding the peritoneal cavity. Transport studies in rodents have shown that i.p. administered macromolecules move into the tissues adjacent to the peritoneal cavity [7. 8], Although the exact mechanism for this macromolecular transport has not been delin eated, the transport is most likely due to con vection of fluid (and entrained solutes) which results from the increased pressure driving force. Macromolecular solute concentration would remain unchanged during such fluid movement. During PD with hypertonic dialysate. fluid passes out of the blood, through the interstitium, and into the peritoneal cavity. As dis cussed above, tracer studies suggest that fluid moves from the peritoneal cavity into the interstitium. How can fluid be convected si multaneously in opposite directions? Some
Indirect Methods Since direct cannulation of lymphatic ves sels is ethically unacceptable in humans and cannot be used for measuring flow rates through nonlymphatic pathways (direct blood entry), clinical investigators have relied upon various indirect methods. These techniques use macromolecules as tracers and assume that fluid and macromolecular tracer move together, thus, movement of the tracer serves as a surrogate for fluid flow. Investigators have measured both the disappearance from the peritoneal cavity and the appearance in the blood of intraperitoneallv instilled macro molecules.
118
Shocklev/Ofsthun
Tracer Disappearance Technique Tracer disappearance from the peritoneal cavity can be used as an indirect measure of total fluid loss from the peritoneal cavity. This technique assumes that fluid and tracer always leave the cavity together, regardless of the pathway. To perform this technique, the investigator infuses into the peritoneal cavity a known volume of solution containing a tracer and immediately draws a sample to determine the initial i.p. volume (instilled volume plus residual volume). At the end of dialysis, fluid is drained from the peritoneal cavity, the drain volume is recorded, and the tracer concentration is measured. The final residual volume is determined by tracer dilu tion after addition of a known quantity of fresh (tracer-free) dialysate; the final i.p. vol ume is calculated from the sum of the drain volume and the final residual volume. Given initial and final values for i.p. volume and tracer concentration, the mass of tracer lost from the peritoneal cavity during dialysis is calculated. Then the total volume of fluid lost during the dwell is computed by dividing the mass of tracer lost by its average concentra tion during the dwell (geometric mean of the initial and final values). Finally, the rate of fluid loss through all pathways is calculated from the total volume lost by assuming that the rate is constant throughout the dwell. The assumption that fluid and tracer al ways leave the peritoneal cavity together may be incorrect. Struijk et al. [13] have shown that the concentration of dextran in hypoosmotic i.p. fluid increases with time. This sug gests that fluid leaves the peritoneal cavity at a somewhat higher rate than tracer, as would be expected if macromolecular transport is hindered by the mesothelium. In the past, some investigators have pre sumed that all fluid exits the peritoneal cavity via lymphatic pathways, and have used the tracer disappearance technique to estimate
Fluid Loss from the Peritoneal Cavity
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
Direct Methods Currently, there is no direct method for measuring total fluid loss from the peritoneal cavity, and the only individual pathway whose flow rate can be measured directly is uptake through lymphatic vessels. In large an imals like sheep, lymph How rates have been measured by cannulation [5, 11]. Unfortu nately, each small lymphatic vessel that drains the peritoneal cavity cannot be cannulated, so instead, investigators cannulate larger vessels further downstream from the peritoneal cavity. However, since the lym phatics that drain the peritoneal cavity are fed not only from the peritoneal space but also from other tissues and organs, the actual lymph flow rate from the peritoneal cavity is estimated using i.p. macromolecular tracers (by monitoring the mass of i.p. tracer that en ters a given lymph compartment and dividing by the concentration of the tracer in the peri toneal cavity) [11], One drawback with this technique is that the cannulation process may affect the flow rate [ 12], Moreover, since can nulation is not ethically feasible in humans, use of this technique to estimate lymph flow during PD requires extrapolation from ani mal data.
dure for the tracer appearance technique is similar to that for the tracer disappearance technique, but in this case both blood and dialysate samples are collected for determina tion of tracer concentration. Over a given time interval, the apparent lymph flow rate is estimated by dividing the change in mass of tracer in the plasma by the average i.p. tracer concentration during that time interval. As suming that the tracer appearance rate is con stant, the change in mass of tracer over the course of an entire experiment is simply cal culated by multiplying plasma volume by the final concentration of tracer in the plasma. Thus, lymph flow rate can be estimated by multiplying the final concentration of tracer in the plasma by the plasma volume, and di viding the result by the average i.p. tracer con centration (geometric mean of the initial and final values). Although not discussed here, some investigators correct for tracer that is lost from the blood by distribution into other body compartments during the course of the experiment [15], The assumption that fluid and tracer ap pear in the blood only as a result of movement through lymphatic pathways may be incorrect because tracer can enter the bloodstream through the direct blood entry pathway [ 16], The magnitude of this error is small because the capillary wall significantly hinders the passage of macromolecules. Therefore, the tracer appearance technique slightly overesti mates lymph flow.
Tracer Appearance Technique Tracer appearance in the plasma has been used as an indirect technique for estimating lymph flow from the peritoneal cavity during a single PD dwell [8, 12, 14. 15]. This tech nique assumes that fluid and tracer move out of the peritoneal cavity together and reach the bloodstream simultaneously only through lymphatic pathways. The experimental proce
Comparison oj Tracer Disappearance and Appearance Techniques One might expect that rates of tracer ap pearance would equal rates of tracer disap pearance. However, as discussed above, the tracer disappearance technique estimates to tal fluid loss, while the tracer appearance tech nique estimates (or slightly overestimates) lymph flow. Experiments which measure
119
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
‘lymph flow’ from the peritoneal cavity [2, 3]. However, as discussed above, other investiga tors have shown that both fluid and tracer macromolecules can pass into the interstitial tissues surrounding the peritoneal cavity, and can reenter the bloodstream at different rates [7, 8], Therefore, tracer disappearance from the peritoneal cavity overestimates lymph flow, but is a valid technique for estimating total fluid loss. Because tracer disappearance rates have been shown to remain constant following ap parent saturation of the interstitial space of peritoneal tissues by extended administration of i.p. tracer [9], one might conclude that the tracer disappearance technique does not mea sure the contribution of the interstitial path ways. and therefore is a valid measure of flow through the direct lymphatic pathway. This would be true if diffusion were the only mech anism by which tracer could be transported into and through the interstitium, and if fluid could not reenter the bloodstream (indepen dent of tracer) as a result of Starling’s forces. However, saturation of a tissue with a tracer does not prevent convection of fluid and tracer into and through the tissue. With con vection, some tracer and fluid would be re turned to the blood via interstitial lymphatics, but other fluid could enter blood capillaries directly, leaving behind its tracer. Thus, even with apparent saturation of the interstitium, tracer disappearance is not a measure of lymph flow.
120
Shockley/Ofsthun
in tracer appearance and disappearance rates can be accounted for by distribution of tracer in the interstitial tissues surrounding the peri toneal cavity. Particulate Tracers While the above discussion has focused on the use of macromolecular tracers, particulate tracers (e.g., labeled RBC) have also been used [8. 15]. The rates of disappearance and appearance of particulate tracers arc similar in value because these tracers cannot move into and through the interstitial pathway dis cussed above. The small observed differences (disappearance rates approximately 30% higher) can be explained by hindered passage of particulates through the lymphatic system [8], Since particulates leave the peritoneal cavity primarily via the direct lymphatic pathway, the disappearance rate of particu late tracers provides an estimate for flow through the direct lymphatic pathway, not total fluid loss. Modeling Various mathematical models have been developed for simulating fluid and solute transport in PD patients [19. 20]. These mod els assume that membrane hydraulic perme ability and lymph flow rate vary from patient to patient, but assume that other parameters (e.g.. hydrostatic pressure) are uniform among all patients. By fitting experimentally determined i.p. volume profiles, estimates for membrane hydraulic permeability and lymph flow rate are obtained. A major limitation of these models is the assumption that blood vessels are juxtaposed to the peritoneal cavi ty. with no interstitial tissue between them. Therefore, these models cannot predict fluid flow in the tissue space between the peritoneal cavity and the blood.
Fluid Loss from the Peritoneal Cavitv
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
rates of both tracer appearance and tracer dis appearance have shown that macromolecular tracers disappear from the peritoneal cavity at least twice as fast as they appear in the blood [8. 12, 15]. This difference in rates can be explained by the dependence of the results on the control volume, i.e., the ‘compartment' for analysis. In the tracer disappearance tech nique. the control volume is the peritoneal cavity, and therefore this technique does not differentiate among the possible pathways for tracer exit from the peritoneal cavity. On the other hand, in the tracer appearance tech nique, the control volume is the blood space, and therefore this technique only accounts for pathways through which tracer enters the blood. The difference between control vol umes using tracer disappearance and appear ance techniques is the volume of the intersti tial tissues that surround the peritoneal cavity. Can the difference in control volumes ac count for observed differences in tracer ap pearance and disappearance rates? In other words, is the missing tracer located within some fraction of the interstitial tissue volume surrounding the peritoneal cavity? Assuming a difference of 1 ml/min, 240 ml of fluid-con taining tracer at its i.p. concentration must be accounted for over the course of a 4-hour dwell. If this fluid (and its accompanying trac er) were distributed in interstitial tissue hav ing an average volume fraction of 30% (upper estimate) available to protein [17], it would occupy a tissue volume of 800 ml. Dividing by a peritoneal surface area of 1.75 m2 in humans [17, 18] yields an average tracer pene tration depth of approximately 450 pm. This calculated depth is similar to that observed in animal experiments [7] measuring tracer pen etration into the tissues surrounding the peri toneal cavity, as expected given the similari ties in structure among mammalian tissues. Thus, it appears that the observed differences
Conclusion
In this paper, we have described the path ways that transport fluid out of the peritoneal cavity during PD. The direct lymphatic path way consists of the diaphragmatic lymphatics, which directly connect the peritoneal cavity to the bloodstream. The interstitial lymphatic and direct blood entry pathways convey fluid that has been driven into the interstitial space of the tissue surrounding the peritoneal cavity
by the increased i.p. pressure, and return it to the bloodstream. Since flow through lym phatic pathways is only a portion of the flow through all pathways, total tluid loss is greater than lymph flow. The net effect of fluid loss through all pathways is a reduction in net ultrafiltration and small solute clearances. Strategies that reduce fluid loss (through any pathway) would therefore reduce required in stilled volume and improve PD as a therapy for end-stage renal disease.
References 8 Nagy JA: Lymphatic and non-lym phatic pathways of peritoneal ab sorption in mice: Physiology versus pathology. Blood Purif 1992:10: 148-1621 9 Struijk DG. Koomen GCM. Krediet RT, Arisz L: Indirect measurement of lymphatic absorption in CAPD patients is not influenced by trap ping. Kidney lnt 1992:41:16681775. 10 Flessner MF: Net ultrafiltration in peritoneal dialysis: Role of direct fluid absorption into peritoneal tis sue. Blood Purif 1992:10:136-147. 11 Tran LP. Rodela H. Abernethy NJ. Yuan Z-Y. Hay JB. Oreopoulos D. Johnston MG: Lymphatic drainage of hypertonic dialysis solution from the peritoneal cavity: Comparison between anesthetized and conscious sheep. Am J Physiol, in press. 12 Johnston MG: Studies on lymphatic drainage of the peritoneal cavity in sheep. Blood Purif 1992:10:122131. 13 Struijk DG. Krediet RT. Koomen GCM, Arisz L: Fluid kinetics in CAPD patients during dialysis with a hvpo-osmotic solution. Periton Dial Int 1992:12:164.
14 Rippe B. Stelin G. Ahlmen J: Lymph flow from the peritoneal cavity in CAPD patients; in Maher JF, Winchester JF (eds): Frontiers in Peritoneal Dialysis. New York, Field, Rich, 1986,pp 24-30. 15 Flessner MF. Parker RJ. Sieber SM: Peritoneal lymphatic uptake of fi brinogen and erythrocytes in the rat. Am J Physiol I983:244:H89-H96. 16 Bent-Hansen L, Svendsen JH: Tis sue to plasma capillary permeability of ,5ll-albumin in the perfused rab bit ear. Microvasc Res 1991:41: 141-148. 17 Bell DR. Watson PD. Renkin EM: Exclusion of plasma proteins in interstitium of tissues from the dog hind paw. Am J Physiol 1980:239: H532-H538. 18 Henderson LW: The problem of peritoneal membrane area and per meability. Kidney lnt 1973:3:409—
410. 19 Rippe B. Zakaria ER: Lymphatic vs. non-lymphatic fluid absorption from the peritoneal cavity' as related to the peritoneal ultrafiltration ca pacity and sieving properties. Blood Purif 1992:10:189-202. 20 Vonesh EF. Rippe B: Net fluid ab sorption under membrane transport models of peritoneal dialysis. Blood Purif 1992:10:209-226.
121
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 2/19/2018 4:31:08 PM
1 Stelin G. Rippe B: A phenomeno logical interpretation of the varia tion in dialysate volume with dwell time in CAPD. Kidney Int 1990:38: 465-472. 2 Krediet RT, Struijk DG, Koomen GCM, Arisz L: Peritoneal fluid ki netics during CAPD measured with intraperitoneal dextran 70. ASAIO Trans 1991;37:662-667. 3 Mactier RA. Khanna R, Twardowski Z. Moore H. Nolph KD: Contribution of lymphatic absorp tion to loss of ultrafiltration and sol ute clearances in continuous ambu latory peritoneal dialysis. J Clin In vest 1987,80:1311-1316. 4 Gotloib L. Mines M. Garmizo L. Varka I: Hemodynamic eflcctsofincreasing intra-abdominal pressure in peritoneal dialysis. Periton Dialy sis Bull 1981:1:41-43. 5 Drake RE. Gabel JC: Abdominal lymph flow response to intraperito neal fluid in awake sheep. Lymphology 1991:24:77-81. 6 Johnston MG. Elias R: The regula tion of lymphatic pumping. Lymphologv 1987:20:215-218. 7 Flessner MF, Fenstermacher JD. Dedrick RL, Blasberg RG: Perito neal absorption of macromoleculcs studied by quantitative autoradiog raphy. Am J Phvsiol 1985:248:H26H32.
