Clinical Results and Complications of Peritoneal Dialysis in Acute Kidney Injury

Historical Perspective

The initial description of PD for the management of ARF is credited to Professor G. Ganter, a German clinical investigator who in 1923 used this technique to treat a woman with uremia and obstructive uropathy. His recommendations then formed the basis for what later became known as intermittent peritoneal dialysis (IPD).

In March 1946, Fine et al. reported the successful application of peritoneal dialysis in a case of antibiotic-induced ARF. This report also established the closed dialysis system as well as the constituents of the dialysis solution, use of which became the standard for peritoneal dialysis. A number of reports followed that reviewed the literature during that period and confirmed the usefulness of peritoneal dialysis in uremia.

By the 1970s, IPD was established as an effective form of renal replacement therapy. Subsequently, as interest in PD grew, the treatment underwent further improvement, with the development of better peritoneal access and the use of automated peritoneal dialysis machines.

Effect on Mortality

There is a paucity of data concerning the effects on mortality in patients on PD for AKI. Chionh et al. recently published a detailed systematic review to describe outcomes in AKI treated with PD and also compared PD with extracorporeal blood purification techniques such as intermittent hemodialysis or with continuous renal replacement therapy.

Thirteen studies described patients treated with PD only. Eleven studies compared PD with continuous or intermittent RRT, of which seven were cohort studies and four were randomized controlled trials. Overall there was no difference in mortality between PD and extracorporeal blood purification therapies in the observational studies (OR 0.96, 95% CI, 0.53–1.71) and the four randomized controlled trials (OR, 1.5: 95% CI, 0.46–4.86).

Among the four randomized controlled trials, two compared PD with continuous therapies, whereas the third compared PD with daily intermittent HD. The fourth RCT randomized to either PD or intermittent HD had only eight patients with AKI.

Phu et al. conducted an open randomized trial comparing continuous venovenous hemofiltration (CVVHDF) with peritoneal dialysis in patients with infection-associated AKI; 48 of the patients had falciparum malaria, and 22 were septic. The mortality was significantly higher in patients treated with peritoneal dialysis (47%) than in those patients treated with CVVH (15%; p < .005). The CVVH group had a significantly lower mortality (15%) in spite of a lower dose of replacement fluid (24 L/day) in relation to current practice. Some adverse factors in the peritoneal dialysis group such as the use of acetate as buffer, the use of rigid catheters, presence of a cloudy dialysate suggesting infection in 42% of patients, and other technical and specific factors could have attributed to the poorer outcome in the PD group. These factors must be considered before one can conclude that peritoneal dialysis is inappropriate for infection-associated AKI. In contrast to this study, Mishra et al., in retrospective study of patients with cerebral malaria and AKI, showed no difference in survival between patients treated with PD versus daily HD despite patients in the PD cohort having a higher number of patients with cerebral malaria.

George et al. performed a randomized study comparing CVVHDF and PD in critically ill patients looking at solute control and fluid overload. Although urea, creatinine clearance, and the control of fluid overload was better with CVVHDF, acidosis was better controlled with PD. The mortality rates between the two groups were similar.

Gabriel et al. performed a randomized controlled trial comparing HVPD (high-volume peritoneal dialysis) with daily intermittent HD. Both modalities achieved metabolic and acid-base control, and mortality did not differ significantly between the two groups ( Table 184.1 and Fig. 184.1 ).

Ponce et al. then compared the effects of HVPD with prolonged or extended HD (PHD) in a prospective study in patients with AKI. Although the delivered Kt/V and ultrafiltration was higher in the PHD group, there was no difference in mortality or in renal recovery.

Ponce et al. recently reported a prospective cohort study in which all AKI patients on PD were studied between January 2004 and January 2014. For comparison, patients were divided into two groups according to the year of treatment: 2004 to 2008 and 2009 to 2014. A total of 301 patients were included in the study. There was an improvement in patient survival and technique failure (TF) with a relative risk reduction (RR) of 0.86 (95% CI, 0.77–0.96) in patients treated during 2009 to 2014 compared with patients treated between 2004 and 2008. This improvement was thought to be related to better fluid control and improved management of PD-related infections.

Dialysis Adequacy

Traditionally, IPD has been held to be potentially inadequate to control azotemia, especially in hypercatabolic patients. This perception was reflected in a survey among nephrologists to determine modalities in the treatment of AKI; 90% of those surveyed believed that solute clearance with peritoneal dialysis was inadequate.

In contrast to these expectations, however, a number of early studies evaluating peritoneal dialysis in patients with AKI who were deemed hypercatabolic reported satisfactory control of fluid and metabolic derangements. These studies had major limitations. The majority of the study populations were small in numbers, were not randomized, and did not use appropriate measurements of dialysis adequacy and catabolic rate.

Chitalia et al. conducted a randomized, prospective, crossover trial comparing adequacies of tidal peritoneal dialysis (TPD) with continuous equilibration peritoneal dialysis (CEPD) in 87 patients with mild to moderate hypercatabolic AKI ( Table 184.2 ). Compared with CEPD, TPD produced higher solute clearance in a smaller dialysis volume. Comparing adequacy indices (Kt/V, normalized creatinine clearances, solute reduction indices), these investigators concluded that TPD and CEPD are reasonable options for mild to moderate catabolic AKI, even though CPD fell short of the adequacy standard. Because TPD offers better clearances at lower cost and time, developing countries that have access to PD cyclers should consider TPD for the treatment of hypercatabolic AKI. However, one of the limitations of the study was that the patient base was different from most of the studies dealing with hypercatabolic AKI; therefore the results may not be applicable to critically ill patients, especially in developed countries. The major limitation of the use of TDP is the high protein loss.

The study by Phu et al. mentioned earlier demonstrated that the rate of resolution of acidosis as well as the decline in serum creatinine was more than twice as high in the hemofiltration group as in the peritoneal dialysis group ( p < .005). However, solute clearance and dialysis adequacy were not reported in the two study groups.

To overcome some of the limitations of PD use in AKI such as low solute clearance, especially in hypercatabolic patients and the unpredictable fluid removal, Gabriel et al. proposed the use of cyclers, flexible catheters, and high volumes of dialysis fluids. In 2007 these investigators assessed the efficacy of high-volume PD (HVPD) in a prospective study of 30 patients, of whom 66% were in the intensive care with AKI. PD was performed using a flexible Tenchkoff catheter with an automated PD cycler with a prescribed Kt/V of 0.65 per session (24 hours). HVPD was found to be effective in correcting uremia, metabolic acidosis, and fluid overload. The weekly delivered Kt/V was 3.8 ± 0.6, and the mortality rate was 57%. The investigators concluded that HVPD was a viable alternative to other forms of RRT for AKI.

In 2012 the same investigators performed another prospective study on 204 patients with AKI treated with HVPD (prescribed Kt/V 0.6/session). Sepsis was the main cause of AKI (54.7%), and 70% of the patients were in the intensive care unit. Urea and creatinine levels stabilized after four sessions, and the delivered weekly Kt/V was 3.5±0.68. With respect to AKI outcomes, the mortality rate was 57.3%, and 23% had renal recovery. They concluded that in selected patients, HVPD was effective in relation to metabolic and fluid control.

Dialysis Dose

There are also limited data on the appropriate of dose of PD for AKI. Gaiao et al.'s survey regarding the use of PD in AKI among delegates at three dialysis congresses found that 70% of the respondents were not certain of the appropriate dose for AKI in ICU and 66% in the wards.

In the study by Chitalia et al. mentioned earlier, TPD was more efficient than CEPD, achieving a significant higher weekly Kt/V of 2.43-± 0.87 versus 1.80 ± 0.32, respectively, with excellent outcome.

Gabriel et al. conducted a prospective randomized controlled trial to compare the effect of high-volume peritoneal dialysis (HVPD) with daily dialysis (DHD) on patients with AKI. A total of 120 patients were assigned to either HVPD or DHD. The weekly delivered Kt/V was 3.6 ± 0.6 in HVPD arm and 4.7 ± 0.6 in DHD arm ( p < 0.01), and they reported comparable outcomes.

Ponce et al. then followed up with a prospective randomized trial of 61 critically ill patients with sepsis and AKI comparing two levels of intensity of HVPD. They were randomized to receive a higher dose (n = 31) versus lower dose (n = 30) of PD therapy (prescribed Kt/V 0.8/session versus 0.5/session).

The lower-dose group achieved a Kt/V of 3.43, whereas the higher-dose group achieved a Kt/V of 4.13. However, the mortality rate of both groups was similar after 30 days (55% vs. 53%, p = 0.83), thus showing no added benefit using higher dose of therapy.

Chionh et al. in a recent review of PD dose in AKI recommended that continuous forms of PD should be prescribed with a minimum standardized Kt/V urea of at least 2.1 per week. However, the optimal dose for PD in AKI is still unclear ( Table 184.3 ).

Renal Recovery

Recovery of renal function after an episode of AKI is an important determinant of morbidity. In many of the older studies comparing peritoneal dialysis with hemodialysis for AKI, the reason for improved survival in the peritoneal dialysis group was related to a higher rate of renal recovery. Several published reports suggest that patients with AKI secondary to atheroembolic disease may have a better chance of renal recovery with peritoneal dialysis than with hemodialysis. The reasons for the apparent benefit were attributed to less hemodynamic fluctuation and the absence of anticoagulation during peritoneal dialysis. It also has been suggested that there is a more rapid recovery of renal function in AKI patients treated with PD. However, some of the published reports show conflicting results. One Brazilian study showed shorter time to renal recovery with PD compared to daily HD ; however, studies from India and Vietnam noted patients on PD required more or longer dialysis sessions.

Katz et al., who conducted a retrospective study in patients with AKI secondary to malignant hypertension, reported that 55% of patients undergoing peritoneal dialysis recovered renal function, compared with none undergoing hemodialysis. This finding suggests that peritoneal dialysis may be beneficial in patients whose AKI is due to malignant hypertension. Unfortunately, the rate of renal recovery is not reported in many of the other studies.

Lactate Versus Bicarbonate-Buffered Solutions

There is one randomized controlled trial from the Cochrane database that compared the effectiveness of bicarbonate versus lactate-buffered PD solutions in 20 AKI patients and found no difference between bicarbonate and lactate with respect to mortality and other adverse events. However, for patients in shock, a more rapid increase in serum bicarbonate was seen using bicarbonate-buffered solutions (21.2. ± 1.8 mmol/L vs. 13.4 ± 1.3 mmol/L) compared with lactate-based solutions. These results suggest that patients with AKI associated with shock should be managed using bicarbonate-buffered solutions rather than lactate.