Water and Solute Transport through Different Types of Pores in Peritoneal Membrane in Capd Patients with Ultrafiltration Failure

Abstract

Free water transport, an estimate of aquaporin function, was evaluated in 7 continuous ambulatory peritoneal dialysis (CAPD) patients with permanent ultrafiltration failure. In 3 patients, peritoneal transport was studied also before the onset of ultrafiltration failure. Transcapillary ultrafiltration and fluid absorption rates were assessed using radiolabeled albumin, and free water transport by kinetics of sodium concentration in dialysis fluid. Diffusive and convective transport rates of small solutes were estimated using the modified Babb–Randerson–Farrell model. Increased diffusive transport of small solutes was found in 5 patients and increased peritoneal fluid absorption in 2 patients. The 3-pore model was fitted to these data. Overall, hydraulic conductivity and the fractional contributions of aquaporins to hydraulic conductivity were either decreased or normal. We conclude that the quantitative role of aquaporins in overall fluid transport may vary substantially in normal patients as well in patients with ultrafiltration failure.

Keywords

Peritoneal transport sieving coefficient mathematical modeling free water fraction sodium kinetics aquaporins

Permanent ultrafiltration failure (UFF) is a common complication in long-term peritoneal dialysis (PD) patients (1 -3). Its etiology is multifactorial and includes increased diffusivity of small solutes, loss of aquaporin function, increased peritoneal absorption of dialysis fluid, and different combinations of these factors. Recently, theoretical and clinical studies have enabled the investigation of aquaporin function via free water transport (FWT), the part of transcapillary ultrafiltration that passes through water-only permeable channels (aquaporins) and is therefore depleted from solutes. Information about kinetics of net fluid transport and sodium concentration in dialysis fluid is used for that purpose (2,4,5). The three-pore model has also been applied in analysis of possible causes of and involvement of aquaporin dysfunction in UFF (6).

These new clinical approaches are based on additional assumptions and approximations (2,4,5). A holistic analysis of fluid and sodium transport may, however, be performed using data from dwell studies with a volume marker to assess kinetics of dialysis fluid volume during a single dwell study and differentiate between transcapillary ultrafiltration and fluid absorption from the peritoneal cavity (5). Furthermore, detailed kinetics of sodium concentration not only during a short initial period of dwell time [as in the mini-peritoneal equilibration test (PET) (4)] but also during later phases of the dwell study allows for evaluation of the diffusive mass transport parameter (K_BD, MTAC) for sodium with no additional assumptions about this parameter [such as equality of K_BD values for sodium and uric acid; c.f. (2)].

An interesting attempt to provide a theoretical explanation for UFF in PD patients using the three-pore model was presented by Parikova et al. (7). Three groups of patients with UFF and different times on PD were investigated; the group with the short time on the treatment (<24 months) was compared to the matched control group without UFF. It was assumed that the fractional contribution of aquaporins to hydraulic conductance (LpS) is the same in all patients but the radius of small pores, LpS, and the diffusivity of small solutes were adjusted to the data on individual patients. This cross-sectional study showed no difference in the reflection coefficient for glucose between the three groups of UFF patients but did show decreased values of LpS, FWT, and peritoneal absorption in the group with the longest time on PD (>60 months). The dynamics of the changes in peritoneal transport characteristics when a patient looses ultrafiltration capacity and after the onset of UFF still needs to be investigated in detail.

In the present study we assessed FWT using sodium kinetics during 6-hour dwell studies with radioiodinated human serum albumin (RISA) as a volume marker in a group of 7 patients with UFF on continuous ambulatory peritoneal dialysis (CAPD). Three of these patients had also been investigated with the same methods before the onset of UFF and therefore, in these 3 patients, we were able to observe the changes in components of fluid transport from regular to impaired fluid removal. Two other patients had a relatively rare cause of UFF with increased absorption of fluid from the peritoneal cavity. Therefore, although some data about these patients were reported previously (1,8 -10), the new approach to the investigation of FWT enabled us to add this important information about the mechanisms for UFF in PD patients.

In addition to the direct assessment of FWT from clinical data, we fitted the three-pore model to these data to describe the values of transport parameters that cannot be estimated without mathematical modeling, such as hydraulic conductivity (LpS) and the fraction of transcellular pores (aquaporins). This approach made it possible to estimate to what extent impairment of aquaporin function contributes to UFF [c.f. (6)].

Methods

Patients and Clinical Methods

Seven patients on CAPD underwent repeat dwell studies that were carried out before and/or after the onset of UFF [Table 1; see also (1,8,9)]. The patients who were classified into the group with loss of ultrafiltration capacity had net ultrafiltration (calculated as the difference between dialysis fluid volume after 4 hours and the sum of the infused and residual volumes) less than 400 mL with glucose 3.86% solution (11). The patients also showed clinical symptoms of UFF as described in Refs. (1) and (9). The identified cause of UFF was increased diffusive transport of small solutes in Patients 1 – 5 and increased peritoneal absorption of fluid in Patients 6 and 7 [Patient 6 also had increased diffusivity of small solutes; c.f. (1,9)].

Table 1

Fluid Transport Components After Sixty Minutes of Dwell Time in Patients with Ultrafiltration Failure

Patient	Dwell	SPWT (mL)	FWT (mL)	CUF (mL)	FWF
1	1	397	191	588	0.33
	2^a	253	185	438	0.42
2	3	143	226	369	0.61
	4^a	141	35	177	0.20
3	5^a	236	71	306	0.23
4	6	354	253	607	0.42
	7^a	250	78	328	0.24
	8^a	169	24	193	0.12
5	9^a	381	85	465	0.18
	10^a	49	4	53	0.08
6	11^a	584	21	605	0.04
	12^a	477	100	577	0.17
7	13^a	309	166	476	0.35

SPWT = small-pore water transport; FWT = free water transport; CUF = cumulative transcapillary ultrafiltration; FWF = free water fraction.

Dwell study after the onset of loss of ultrafiltration capacity.

Dialysis fluid (2 L) with 3.86% glucose concentration was applied during 6-hour dwell studies. Dialysate samples were collected at 0 (before infusion) 3, 15, 30, 60, 90, 120, 180, 240, and 360 minutes and blood samples at 0, 180, and 360 minutes of dwell time. Dialysate volume and fluid absorption rate were measured using RISA. Sodium concentration in plasma and dialysis fluid was measured by flame photometry; plasma concentrations were corrected for plasma lipid and protein volumes and the Donnan factor (1,9). A detailed description of the methods may be found in Ref. (12).

Calculations

Cumulative transcapillary ultrafiltration at dwell time t [CUF(t)] was calculated as the difference between dialysis fluid volume at 3 minutes and time t, plus cumulative fluid absorption during the same time period. The values of sieving coefficient (S_Na) and diffusive mass transport coefficient (K_BD) for sodium were estimated using two-parameter linear regression of the integral form of the modified Babb–Randerson–Farrell model (12).

Small-pore water transport (SPWT) was calculated as SPWT(t) = S_Na · CUF(t) (5). Free water transport was defined as the difference between total CUF and cumulative fluid volume that passes through small pores (5). Free water fraction (FWF) was defined as the fraction of fluid that passes through ultrasmall pores and was calculated as FWF = 1 – S_Na (5). In particular, CUF, SPWT, and FWT are reported for t = 60 minutes according to the standard of a mini-PET (4) (c.f. Table 1), although our method of estimation is different and valid for any dwell time.

Mathematical Modeling

The basic version of the three-pore model as used in the present study was taken from Rippe et al. (13) [c.f. (14)]. The relative contribution of each pore type to total LpS was α_TP, α_SP, and α_LP for transcellular, small, and large pores, respectively, with α_LP = 0.08 and α_TP and α_SP = 1 – α_LP – α_TP selected by fitting the model to the data together with total LpS and peritoneal fluid absorption rate (Q_a). The values of diffusive mass transport coefficients (PS) and LpS were multiplied by the function f(t) = 1 + 0.6875·exp(–t/50), where t is dwell time, to take into account the effect of initial vasodilation on fluid and solute transport (14). Diffusive mass transport coefficient was estimated separately for glucose, urea, creatinine, and sodium by fitting the three-pore model predictions to the data on concentration kinetics of these solutes in dialysis fluid (14). The fitting was performed for the whole dwell period (3 – 360 minutes) except for sodium in dwell 2, where the fit and the estimation of sodium transport parameters were performed for 3 – 180 minutes because the dialysate-to-plasma ratio for sodium at 360 minutes was >1 for unclear reasons.

Results

A drop in net ultrafiltration after the onset of UFF was evident in Patients 1, 2, and 4 but the extent of the change was different, with a relatively small drop in Patient 1 and pronounced drops in Patients 2 and 4. Patients 6 and 7 had much lower final volumes than Patients 1 – 5. The three-pore model was able to describe the kinetics of dialysis fluid volume with high accuracy. The initial decrease in sodium concentration in dialysis fluid (sodium dip) was pronounced in dwell studies before the onset of UFF. In dwell studies in patients with UFF, a slight sodium dip was also sometimes observed (as in Patients 1, 2, and 4) or some initial delay in the increase in sodium concentration in dialysis fluid could be noted (as in Patients 4 and 6). The three-pore model described these sodium profiles with high accuracy.

Cumulative transcapillary ultrafiltration was much reduced after the onset of UFF in Patients 1 – 5 but not in Patients 6 and 7 (Table 1). Free water fraction and FWT were also markedly reduced in Patients 2 – 6 but normal in Patients 1 and 7 (Table 1). The values of LpS and osmotic conductance (a_G) decreased in Patients 1, 2, and 3 after the onset of UFF, and LpS was low in Patient 2 even before the onset of UFF (Table 2). A decrease in LpS with time after the onset of UFF was found in Patient 5. The fractional contribution of transcellular pores (aquaporins) to LpS (α_TP) was >0.02 in Patients 2 and 4 before the onset of UFF; the value of α_TP = 0.02 is usually assumed by the three-pore model for PD patients. In Patients 2 and 4, α_TP decreased after the onset of UFF but no decrease was found in Patient 1 (Table 2). A normal value of α_TP after the onset of UFF was estimated also for Patients 3 and 7 but α_TP was low after the onset of UFF in Patients 4, 5, and 6. The net reflection coefficient for glucose (σ_G) varied between 0.030 and 0.088 (Table 2). The values of a_G were high in Patient 6 despite the fact that the contribution of aquaporins to water transport was very low (Table 2).

Table 2

Sodium Diffusivity (PS), Peritoneal Fluid Absorption Rate (Q^a), Hydraulic Permeability (LpS), the Fractional Contribution to Hydraulic Conductance by Transcellular Pores (α_TP) and Small Pores (α_SP), the Net Reflection Coefficient for Glucose (σ_G), and Osmotic Conductance (a_G) Obtained Using the Three-Pore Model

Patient	Dwell	PS (mL/min)	Q^a (mL/min)	LpS (mL/min·mmHg)	α_TP	α_SP	σ_G	a_G (mL/min/mOsm/L)
1	1	8.0	1.90	0.103	0.020	0.900	0.049	0.089
	₂ ^a	38.0	2.00	0.058	0.030	0.890	0.059	0.060
2	3	6.5	1.00	0.048	0.060	0.860	0.088	0.074
	₄ ^a	12.5	1.50	0.024	0.030	0.890	0.059	0.025
3	₅ ^a	22.5	1.90	0.061	0.020	0.900	0.049	0.053
4	6	8.5	1.60	0.076	0.035	0.885	0.063	0.086
	7^a	15.0	1.82	0.056	0.020	0.900	0.049	0.048
	8^a	7.0	1.23	0.039	0.010	0.910	0.039	0.027
5	9^a	11.0	1.80	0.115	0.010	0.910	0.039	0.080
	10^a	0.0	0.70	0.028	0.000	0.920	0.030	0.015
6	11^a	2.0	4.10	0.162	0.000	0.920	0.030	0.084
	12^a	13.0	5.30	0.131	0.010	0.910	0.039	0.091
7	13^a	10.0	4.80	0.082	0.020	0.900	0.049	0.071

Dwell study after the onset of loss of ultrafiltration capacity.

Sodium diffusivity was much increased in Patient 1 (dwell 2) and in Patient 3 and moderately increased in Patient 4 (dwell 7), Patient 5 (dwell 9), Patient 6 (dwell 12), and Patient 7. Abnormally low values of sodium diffusive mass transport parameters were found in Patient 5 (dwell 10) and Patient 6 (dwell 11) and a very high value was found in Patient 1 after the onset of UFF (dwell 2). The reason for such unusual values, which were not confirmed in other dwell studies in the same patients, is not clear but the values of PS describe well the sodium profiles during these three dwell studies. In Patient 4, after a marked increase in diffusivity of sodium observed after the onset of UFF (dwell 7), some return to the normal range occurred in the consecutive dwell study (dwell 8).

Discussion

This analysis shows that UFF may have different patterns of change in specific fluid transport components. In Patient 1, UFF was related to a marked reduction in LpS without any decrease in the fractional contribution of transcellular pores (an increase in α_TP was found; Table 2). This drop in LpS occurred despite a marked increase in the diffusive transport of small solutes [Table 2 and Ref. (9)] and is an example of a decoupling of the diffusive permeability for small solutes and LpS, which was predicted by Rippe et al. (6).

In Patients 2, 4, and 5, a substantial reduction in LpS was found together with decreased α_TP concomitantly with increased diffusive mass transport parameters for small solutes, although this increase in diffusive permeability was not always seen in PS values for sodium [Table 2 and Ref. (9)]. Patient 3, with peritoneal transport investigated only after the onset of UFF, had a low LpS, but it was within the expected range, and normal α_TP; thus, the only reason for UFF in this patient was an increase in diffusive transport of small solutes [Table 2 and Ref. (9)]. Patients 6 and 7 had very high rates of peritoneal absorption and only slightly increased diffusive transport of small solutes [Table 2 and Ref. (9)]; however, their fluid transport parameters differed: in Patient 6, LpS was high but α_TP low, whereas in Patient 7 these two parameters were approximately normal (Table 2).

Individual estimations of LpS and α_TP were performed in the present study, in contrast to previous publications where only osmotic conductance was reported or a constant α_TP was assumed for all patients (2,9,10). The separation of these two parameters, which contribute to the lumped parameter osmotic conductance, using the data on fluid and sodium transport rates is a consequence of the application of the three-pore model and the fact that water transport depends on osmotic conductance whereas, for a given total ultrafiltration, sodium dip depends substantially on α_TP.

The variability in UFF patterns, as found for the fluid transport parameters in this small group of patients, is in agreement with previously reported values in a much larger patient population (2,7). Frequently, a few different abnormalities are found, such as an increase in diffusive transport of small solutes, a decrease in the hydraulic permeability of peritoneal membrane, a decrease in the contribution of aquaporins to fluid flow, and/or an increase in the peritoneal absorption of fluid, which may occur concomitantly in various combinations. It is also worth noting that changes in diffusive mass transport parameters (PS) and hydraulic permeability (LpS) are generally not correlated and these parameters often change in opposite directions [c.f. Ref. (6)] despite the prediction from the three-pore model that all these parameters should be proportional to the total pore surface area (13). This observation suggests that other structures and phenomena not described by the three-pore model may be involved in peritoneal transport, such as, for example, the interstitium (15 -19). The changes in the interstitium (thickening, fibrosis) and separately in the capillaries (angiogenesis, functional loss of aquaporins) with time on PD may contribute in various ways to UFF. A mathematical description of these processes needs a more sophisticated modeling, such as, for example, a multilayer model (17) or a distributed model (18,19).

In general, the estimated values of PS for sodium correlated well with the PS values for other small solutes and the respective K_BD values estimated using the membrane model [c.f. Ref. (9)]. However, some extreme values of sodium PS had to be fitted to the sodium concentration profile in some dwell studies, such as the high PS value in dwell 2 and the low PS values in dwells 10 and 11 (Table 2). These extreme values were found during the dwell studies performed after the onset of UFF when transport is, in general, abnormal in patients with clinically evident substantial overhydration. Furthermore, sodium transport, even on average, is different from that expected, for example, by the three-pore model (13,20 -23). However, such extreme deviations in sodium profiles are not frequent and no reason for these observations may be given based on the current study.

The values of osmotic conductance to glucose (a_G) estimated in the present study using the three-pore model were similar to those reported by Smit et al. (2) and Parikova et al. (7), who also used a three-pore model, and to a_G values estimated using a phenomenological model (9,10). These studies reported, on average, lower a_G in patients with UFF than in patients with regular ultrafiltration, although with some overlap of a_G ranges between these two groups of patients. The same overlap is seen in the data presented in Table 2. Osmotic conductance to glucose, which is a lumped parameter, may be high despite a low contribution by aquaporins (as in dwells 9 and 11 – 13) or low despite high contribution by aquaporins (as in dwell 4).

The role of aquaporins in the peritoneal transport of water and the phenomenon of sodium dip has been well documented in animal studies by demonstrating that the lack of aquaporin-1 (in knockout mice) or blocking aquaporin-1 (by mercurials in rats) results in decreased ultrafiltration during PD and disappearance of the sodium dip (24 -26). A clinical observation in an UFF patient with impaired ultrafiltration and no sodium dip but normal expression of aquaporin-1 in submesothelial capillaries suggests that the low contribution by aquaporins to peritoneal transport may be an effect of structural alterations in this protein rather than its disappearance from the endothelial cell membrane (27).

The variability in the peritoneal pore system may include differences in the contribution of different types of pores to hydraulic conductance and the reflection coefficient (α parameters) or the radii of the pores. The radii of the small pores were previously individualized in accordance with the values of diffusive mass transport parameters for patients with UFF (7). Recently, both types of parameters were modified — compared to the standard version of the three-pore model — to get a consistent description of fluid and solute transport with three different osmotic agents (glucose, icodextrin, and a mixture of glucose and icodextrin) (28). In the present study, it was sufficient to modify only the α parameters, with no individual variations in the radii of small and large pores.

In summary, our results demonstrate high variability in the free water fraction and the fractional contribution of ultrasmall pores to transcapillary ultrafiltration, both before and after UFF. Fluid transport parameters are often decreased after the onset of UFF concomitantly with the increased diffusivity of small solutes; however, other combinations are also possible. Ultrafiltration failure due to high peritoneal absorption may be associated with a normal or a decreased fractional contribution by transcellular pores to hydraulic conductivity.

Footnotes

J. Waniewski and M. Debowska report no financial support. B. Lindholm is affiliated with Baxter Healthcare Inc.

References

Heimburger

, Waniewski

, Werynski

, Tranaeus

, Lindholm

Peritoneal transport in CAPD patients with permanent loss of ultrafiltration capacity.

Kidney Int 1990; 38: 495–506.

Smit

, Schouten

, van den Berg

, Langedijk

M.J.

, Struijk

D.G.

, Krediet

R.T.

Analysis of the prevalence and causes of ultrafiltration failure during long-term peritoneal dialysis: a cross-sectional study.

Perit Dial Int 2004; 24: 562–70.

Kawaguchi

, Hasegawa

, Nakayama

, Kubo

, Shigematu

Issues affecting the longevity of the continuous peritoneal dialysis therapy.

Kidney Int Suppl 1997; 62: S105–7.

La Milia

, Di Filippo

, Crepaldi

, Del Vecchio

, Dell'Oro

, Andrulli

Mini-peritoneal equilibration test: a simple and fast method to assess free water and small solute transport across the peritoneal membrane.

Kidney Int 2005; 68: 840–6.

Waniewski

, Debowska

, Lindholm

Ultrafiltration and absorption in evaluating aquaporin function from peritoneal transport of sodium.

Perit Dial Int 2007; 27: 687–90.

Rippe

, de Arteaga

, Venturoli

Aquaporins are unlikely to be affected in marked ultrafiltration failure: results from a computer simulation.

Perit Dial Int 2001; 21(Suppl 3): S30–4.

Parikova

, Smit

, Struijk

D.G.

, Krediet

R.T.

Analysis of fluid transport pathways and their determinants in peritoneal dialysis patients with ultrafiltration failure.

Kidney Int 2006; 70: 1988–94.

Heimburger

, Waniewski

, Werynski

, Park

M.S.

, Lindholm

Lymphatic absorption in CAPD patients with loss of ultrafiltration capacity.

Blood Purif 1995; 13: 327–39.

Waniewski

, Sobiecka

, Debowska

, Heimburger

, Werynski

, Lindholm

Fluid and solute transport in CAPD patients before and after permanent loss of ultrafiltration capacity.

Int J Artif Organs 2005; 28: 976–86.

10.

Waniewski

, Heimburger

, Werynski

, Lindholm

Osmotic conductance of the peritoneum in CAPD patients with permanent loss of ultrafiltration capacity.

Perit Dial Int 1996; 16: 488–96.

11.

Mujais

, Nolph

, Gokal

, Blake

, Burkart

, Coles

Evaluation and management of ultrafiltration problems in peritoneal dialysis. International Society for Peritoneal Dialysis Ad Hoc Committee on Ultrafiltration Management in Peritoneal Dialysis.

Perit Dial Int 2000; 20(Suppl 4): S5–21.

12.

Waniewski

, Heimburger

, Werynski

, Park

M.S.

, Lindholm

Diffusive and convective solute transport in peritoneal dialysis with glucose as an osmotic agent.

Artif Organs 1995; 19: 295–306.

13.

Rippe

, Venturoli

, Simonsen

, de Arteaga

Fluid and electrolyte transport across the peritoneal membrane during CAPD according to the three-pore model.

Perit Dial Int 2004; 24: 10–27.

14.

Waniewski

, Debowska

, Lindholm

How accurate is the description of transport kinetics in peritoneal dialysis according to different versions of the three-pore model?

Perit Dial Int 2008; 28: 53–60.

15.

Flessner

M.F.

The role of extracellular matrix in transperitoneal transport of water and solutes.

Perit Dial Int 2001; 21(Suppl 3): S24–9.

16.

Flessner

M.F.

The transport barrier in intraperitoneal therapy.

Am J Physiol Renal Physiol 2005; 288: F433–F442.

17.

Rippe

, Venturoli

Simulations of osmotic ultrafiltration failure in CAPD using a serial three-pore membrane/fiber matrix model.

Am J Physiol Renal Physiol 2007; 292: F1035–F1043.

18.

Stachowska-Pietka

, Waniewski

, Flessner

M.F.

, Lindholm

Distributed model of peritoneal fluid absorption.

Am J Physiol Heart Circ Physiol 2006; 291: H1862–H1874.

19.

Waniewski

, Dutka

, Stachowska-Pietka

, Cherniha

Distributed modeling of glucose-induced osmotic flow.

Adv Perit Dial 2007; 23: 2–6.

20.

Coester

A.M.

, Struijk

D.G.

, Smit

, de Waart

D.R.

, Krediet

R.T.

The cellular contribution to effluent potassium and its relation to free water transport during peritoneal dialysis.

Nephrol Dial Transplant 2007; 22: 3593–600.

21.

Freida

, Galach

, Divino Filho

J.C.

, Werynski

, Lindholm

Combination of crystalloid (glucose) and colloid (icodextrin) osmotic agents markedly enhances peritoneal fluid and solute transport during the long PD dwell.

Perit Dial Int 2007; 27: 267–76.

22.

Waniewski

, Heimburger

, Werynski

, Lindholm

Paradoxes in peritoneal transport of small solutes.

Perit Dial Int 1996; 16(Suppl 1): S63–9.

23.

Waniewski

, Heimburger

, Werynski

, Lindholm

Diffusive mass transport coefficients are not constant during a single exchange in continuous ambulatory peritoneal dialysis.

ASAIO J 1996; 42: M518–23.

24.

Carlsson

, Nielsen

, Zakaria el

, Rippe

In vivo inhibition of transcellular water channels (aquaporin-1) during acute peritoneal dialysis in rats.

Am J Physiol 1996; 271(6 pt 2): H2254–H2262.

25.

, Verbavatz

J.M.

, Rippe

, Boisde

, Moulin

, Rippe

Aquaporin-1 plays an essential role in water permeability and ultrafiltration during peritoneal dialysis.

Kidney Int 2006; 69: 1518–25.

26.

Yang

, Folkesson

H.G.

, Yang

, Matthay

M.A.

, Ma

, Verkman

A.S.

Reduced osmotic water permeability of the peritoneal barrier in aquaporin-1 knockout mice.

Am J Physiol 1999; 276(1 pt 1): C76–C81.

27.

Goffin

, Combet

, Jamar

, Cosyns

J.P.

, Devuyst

Expression of aquaporin-1 in a long-term peritoneal dialysis patient with impaired transcellular water transport.

Am J Kidney Dis 1999; 33: 383–8.

28.

Galach

, Werynski

, Waniewski

, Freida

, Lindholm

Kinetic analysis of peritoneal fluid and solute transport with combination of glucose and icodextrin as osmotic agents.

Perit Dial Int 2009; 29: 72–80.