**3.3 Fluid management in PPPHD**

130 Progress in Hemodialysis – From Emergent Biotechnology to Clinical Practice

closed and in the final phase (θ=270°~360°), p1 is open, and p2 and p3 remain closed in preparedness for the next filling phase. These time-delayed tube openings and closures constitute one cycle of pulse generation. In the same manner, effluent pulsations were also generated through the effluent tube, although in this case, the actions of actuators 1 and 3 were reversed, and the pulsatile flow pattern was 180O out of phase with that in the

Theoretically, forward and backward filtration rates during one cycle of PPPHD are identical to effluent and dialysate flow rates, respectively. The moment when pure dialysate is driven to the dialyzer (i.e., during p2 squeezing), the effluent dialysate path is closed at p6. At the same time, p1 is also closed, and thus, the pure dialysate pushed into dialyzer should move into the blood stream (backfiltration), because the whole dialysate compartment is fixed and closed. Immediately after the backfiltration is completed, the effluent tubing (p5) begins to expand (i.e., p5 expansion), and since the dialysate and effluent pathways are still closed at p1 and p6, respectively, dialysate pressures in the hemodialyzer drop steeply and ultrafiltration takes place at a rate determined by effluent

During animal experiments using the PPPHD, in which we used an acute canine renal failure model (achieved by ligating renal arteries and veins), the animals remained stable without any procedurally related complications. Molecular removals were satisfactory, and total protein levels, albumin concentrations, and glucose levels were preserved uniformly throughout PPPHD sessions (Table 2). Furthermore, TMPs clearly cycled positive and negative due to huge fluctuations in hydraulic dialysate pressures (Fig. 8). In addition, despite the use of a peristaltic roller pump for blood, the blood pressures acquired during PPPHD showed an obvious fluctuation which was perfectly synchronized with dialysate pressure pulsation. Generally, peristaltic roller pumps create small fluctuations in flow and pressure, because of the way they squeeze tubing. However, the blood pressure fluctuations acquired during PPPHD were much larger than that observed for conventional HD, which

provides evidence of dialysate flux to the blood stream or vice versa (Fig. 8).

Fig. 8. Pressure Profiles during PPPHD treatment (upper), and Changes in Blood Pressures for PPPHD and CHD (below). (MDP, mean dialysate pressure; MBP, mean blood pressure;

dialysate tube.

stroke volume.

CHD, conventional HD)

Recently, the dual pulsatile pump integrated into the dialysate stream has been remarkably ameliorated to achieve a substantial increase in the accuracy of volume control. Maintaining pre-determined flow rates and precise volume control are pre-requisites of extracorporeal renal replacement treatments for ESRD patients, particularly when using membranes with high-water permeability. Therefore, the dual pulsation system acting on the PPPHD dialysate compartment was replaced with a dual piston pump, as shown in Fig. 9. This modification allows pulse generation and push/pull to be achieved, not only by the novel design of the piston pump, but also by the unique control of piston movements offered (Fig. 10). As the dialysate piston compresses the cylinder, pure dialysate is forced into the dialyzer, but at this time, the effluent stream is functionally closed at the effluent piston pump, and thus, dialysate compartment pressures increase rapidly and backfiltration occurs (Phase 1).The effluent piston then begins to expand and dialysate moves into the effluent cylinder, while the dialysate supply line is still closed at the dialysate pump. Because of effluent suction, dialysate compartment pressures fall sharply and water flux from blood lumen to dialysate occurs (Phase 2). During the final step, pure dialysate fills the dialysate cylinder, and simultaneously used dialysate is drained (Phase 3).

In an *in vitro* test of PPPHD with the dual piston pump, in which bovine blood was circulated through the blood lumen of the hemodialyzer at 200 ml/min and isotonic saline solution was used as dialysate at the rate of 400 ml/min, the phenomena of push (backfiltration) and pull (ultrafiltration) were well sustained throughout sessions, and their levels perfectly matched those of stroke volumes of the dialysate and effluent pumps, respectively. In addition, as was expected, dialysate and effluent piston pumps served as a flow equalizer, and controlled isovolumetic dialysate flow rates upstream and downstream of the dialyzer. Hydrostatic dialysate pressures were maintained at 520~700 mmHg during the backfiltration phase (Phase 1) and at -400~-540 mmHg during the ultrafiltration phase (Phase 2).

In addition, PPPHD is also versatile and can be easily converted to conventional high-flux HD. Time-controlled piston operations perform the push and pull operations, but when the two piston movements are synchronized alternately (that is, dialysate piston compression and effluent piston expansion or dialysate piston expansion and effluent piston compression

Pulse Push/Pull Hemodialysis: Convective Renal Replacement Therapy 133

occur simultaneously), dialysate passes through the hemodialyzer without significant flow into blood lumen. In this situation, the two piston pumps serve as a flow equalizer and

The PPPHD unit presented was developed recently, and thus, it requires further investigation. Convective volume attained during PPPHD was equal to the accumulated total dialysate volume, and consequently, this unit delivered the maximally permissible level of total volume exchange. This encourages us to speculate on the capability of this unit in terms of removing mid-sized uremic toxins. Another issue regarding the enormous fluid exchange is the quantification of the contribution made by convection to dialytic efficiency. Backfiltration and ultrafiltration repeat in a relatively short time, and despite a large amount of filtration, the probability that some ultrafiltrate comes directly from dialysate backfiltered during a previous phase cannot be excluded, because that portion of ultrafiltrate does not contribute to depurative efficiency. In addition, forward filtration and backfiltration rates exceed the blood flow rate, which implies a reduction in solute concentrations due to dilution. As is the case for pre-dilution HDF, this repeated dilution may be expressed by an efficiency drop. Finally, although convection commonly inhibits diffusion during HDF, this inhibition is expected to be small for PPPHD due to repetitive backfiltration. Although an *in vitro* or *in vivo* setup revealed that alternate backfiltration has a positive influence on inhibiting concentration polarization and permeability reduction, it is believed that optimizations, in terms of pulse frequencies and stroke volumes, will further benefit the optimal use of membrane convective capacities

Much evidence shows that HDF delivers better dialysis outcomes than high-flux HD; for example, HDF has been shown to improve middle-to-large size molecular removal, allow better EPO control, reduce oxidative stress and inflammation (Lornoy et al., 2000, Vaslaki et al., 2006, Ward et al., 2000), and even to positively influence patient mortality (Canaud et al., 2006, Jirka et al., 2006). These benefits have been attributed to the higher convective doses permitted during HDF. Furthermore, ultrapure dialysate, required due to the large amount

In this chapter, we review HDF techniques that do not require exogenous substitution infusion. These techniques must be accompanied by spontaneous fluid restoration, such as, backfiltration or ultrafiltrate regeneration (Table 3). A simpler way might be to increase forward and backward filtration rates during HD sessions, although this can only be done to a limited extent. Much higher efficiencies can be achieved by the two-chamber techniques, that is, double high-flux HDF and HFR, which were developed in an effort to increase solute removal and shorten treatment times, by separating ultrafiltration and backfiltration, or convection and diffusion domains. However, these modalities appear to unavoidably increase overall system complexity. Push/pull HDF, which uses a single hemodialyzer, was derived by considering phases, rather than physical regions, for forward and backward filtration. The pulse push/pull HD described here is also based on the phase-separated use of forward filtration and backfiltration using a single high-flux dialyzer. This strategy was devised as a result of efforts to modulate flow patterns for extracorporeal dialysis treatment, and thus, a unique design for managing dual pulsation through the dialysate compartment

of substitution infusion, further inhibits the inflammation risk (Lonnemann, 2000).

allows the whole unit to be as simple as the conventional HD unit.

dialysis is largely achieved by diffusive mass transfer.

throughout PPPHD treatments.

**4. Conclusion** 

Fig. 9. Schematic Diagram of PPPHD with Dual Piston Pump. (D, dialysate pump; E, effluent pump)

Fig. 10. Pulse Generation and Push/Pull during PPPHD with Dual Piston Pump. (D, dialysate pump; E, effluent pump)

occur simultaneously), dialysate passes through the hemodialyzer without significant flow into blood lumen. In this situation, the two piston pumps serve as a flow equalizer and dialysis is largely achieved by diffusive mass transfer.

The PPPHD unit presented was developed recently, and thus, it requires further investigation. Convective volume attained during PPPHD was equal to the accumulated total dialysate volume, and consequently, this unit delivered the maximally permissible level of total volume exchange. This encourages us to speculate on the capability of this unit in terms of removing mid-sized uremic toxins. Another issue regarding the enormous fluid exchange is the quantification of the contribution made by convection to dialytic efficiency. Backfiltration and ultrafiltration repeat in a relatively short time, and despite a large amount of filtration, the probability that some ultrafiltrate comes directly from dialysate backfiltered during a previous phase cannot be excluded, because that portion of ultrafiltrate does not contribute to depurative efficiency. In addition, forward filtration and backfiltration rates exceed the blood flow rate, which implies a reduction in solute concentrations due to dilution. As is the case for pre-dilution HDF, this repeated dilution may be expressed by an efficiency drop. Finally, although convection commonly inhibits diffusion during HDF, this inhibition is expected to be small for PPPHD due to repetitive backfiltration. Although an *in vitro* or *in vivo* setup revealed that alternate backfiltration has a positive influence on inhibiting concentration polarization and permeability reduction, it is believed that optimizations, in terms of pulse frequencies and stroke volumes, will further benefit the optimal use of membrane convective capacities throughout PPPHD treatments.
