Introduction

Peritoneal dialysis (PD) is a well-established home-based renal replacement therapy. Compared to in-center hemodialysis (HD), it has several advantages including better preservation of residual renal function, a more gradual and continuous solute and fluid removal, minimal cardiac stress, lower medical costs, treatment flexibility, better quality of life and similar survival [1–3]. In spite of these benefits, risks of technique failure and peritonitis are higher in patients undergoing PD than in those undergoing HD patients [4, 5]. Both are major challenges in caring for PD patients.

The conventional PD solution is glucose-based dialysate, which is cheap, safe, easily available, and an effective osmotic agent. However, the solution contains lactate, high concentrations of glucose, and glucose degradation products (GDPs) and has a low pH value and high osmolality [6]. These non-physiological components in this dialysate are not only harmful to the local peritoneal membrane but also have detrimental systemic effects on PD patients [6]. The higher glucose load in the glucose-based solution is associated with increased risks of technique failure and death [7, 8].

Icodextrin is an iso-osmolar mixture of corn starch-derived high molecular weight glucose polymers, containing lactate and no glucose, with a low pH value and a low concentration of GDPs. This solution is more slowly absorbed than glucose by the peritoneal cavity, mainly through lymphatics, maintaining the generated colloid osmotic pressure for a longer dwell (8–16 hours) [6]. Compared to glucose-based dialysates, icodextrin offers benefits, such as more biocompatibility, favorable metabolic effects and better fluid management [9–12].

However, the results of studies about the impact of icodextrin on outcomes of PD patients are inconsistent [13–16]. The aim of this study was to investigate the impact of icodextrin on the risks of mortality, technique failure, and peritonitis in a cohort of incident Asian PD patients. The risks between patients with and without icodextrin treatment identified from medical records of patients cared at China Medical University Hospital, a tertiary medical center in central Taiwan, were compared.

Methods

From medical records retrieved from January 1, 2007 to December 31, 2018, incident end-stage kidney disease (ESKD) patients aged more than 18 years on PD for at least 90 days were identified. The data were assessed for research purposes from July 9, 2019 to July 8, 2020. The records of all eligible patients were retrospectively reviewed for information on demographic data, medical history, underlying comorbid conditions, laboratory data and treatment measures. Cardiovascular comorbidity was defined as a history of coronary artery disease, congestive heart failure, or stroke. Icodextrin users were defined as patients utilizing this dialysate for ≥ 50% of their PD duration [14]. According to the regulations of the Taiwan’s National Health Insurance, icodextrin (Extraneal; Baxter Healthcare Corporation) could be prescribed once daily in patients with an HbA1c > 7.0%, required 2.5% or 4.25% dextrose solution in more than half of the daily exchanges, or were in high or high-average peritoneal membrane transporter status. However, not all patients meeting these criteria had received icodextrin. The prescription of icodextrin was up to medical staff’s discretion for enhancing ultrafiltration and improving glycemic control. Medical records were reviewed until transfer from PD to HD, renal transplantation, transfer to another hospital, death, or December 31, 2018, whichever came first. During the review period, events of death, technique failure, and the first episode of peritonitis were identified. This study was performed in compliance with guidelines of the Declaration of Helsinki. This retrospective observational study was approved by the Research Ethics Committee of China Medical University Hospital [CMUH103-REC2-070 (CR5)]. All data were de-identified and analyzed anonymously. Authors had no access to information that could identify individual participants during or after data collection.

Results

A total of 725 incident PD patients were identified from the medical records, consisting of 190 icodextrin users and 535 nonusers (Table 1). There were more men with lower mean age in the users than non-users without significant difference. Hypertension and diabetes were more prevalent in icodextrin users than non-users. The high or high average transporter status was also higher in users. In addition, icodextrin users had lower renal Kt/V, nPNA, and serum albumin, and higher peritoneal Kt/V and glycated hemoglobin (HbA1c), and were more likely to initiate PD in the year of 2007–2010, compared to the non-users. The mean follow-up period was longer in icodextrin users than in non-users (3.9 ± 2.8 vs. 3.2 ± 2.6 years; p = 0.002). During the follow-up period, 217 patients were transferred to HD: 61 (32.1%) of the icodextrin group, and 156 (29.2%) of the non-icodextrin group. 172 patients died during the follow-up period: 56 (28.5%) of the icodextrin group, and 116 (21.7%) of the non-icodextrin group. 12 of 56 (21.4%) patients in the icodextrin group and 21 of 116 (18.1%) in the non-icodextrin group died within 90 days after switching to HD. In addition, 51 patients received renal transplantation: 5 (2.6%) of the icodextrin group, and 46 (8.6%) of the non-icodextrin group. 40 patients were transferred to another center during the follow-up period: 10 (5.3%) of the icodextrin group, and 30 (5.6%) of the non-icodextrin group.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. The baseline characteristics of patients on peritoneal dialysis compared between cohorts of icodextrin users and non-users.

https://doi.org/10.1371/journal.pone.0297688.t001

Table 2 shows that risks of mortality, technique failure and the first peritonitis episode were all significantly lower in icodextrin users than in non-users, with the corresponding adjusted HRs of 0.62 (95% CI = 0.42–0.91), 0.61 (95% CI = 0.47–0.81), and 0.22 (95% CI = 0.14–0.35), respectively, for the user cohort compared to the non-users. The results based on various cut-offs of PD duration for defining icodextrin users were in general consistent with that of the primary analysis with defining icodextrin users as utilization of this solution for >50% of PD duration. Moreover, the higher cut-off of PD duration for defining icodextrin users was associated with additional risk reduction in all three outcomes (S1 Table). After excluding 51 patients who underwent renal transplantation, the results also indicated a beneficial effect of icodextrin on the risk of the first episode of peritonitis (S2 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Mortality, technique failure and peritonitis compared between icodextrin users and non-users.

https://doi.org/10.1371/journal.pone.0297688.t002

The further subgroup analyses examined whether the 3 study outcomes associated with the efficacy of icodextrin therapy varied by selected covariates (Tables 3–5). The beneficial effects on patient and technique survival were greater for male and younger users, and patients on APD or with high average/high transporter status. The adjusted HRs of developing the first episode of peritonitis were extremely lower for patients with diabetes (0.07, 95% CI = 0.03–0.16) and cardiovascular (0.06, 95% CI = 0.02–0.22) (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Mortality in patients on PD compared between icodextrin users and non-users by subgroup analyses.

https://doi.org/10.1371/journal.pone.0297688.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Technique failure in patients on PD compared between icodextrin users and non-users by subgroup analyses.

https://doi.org/10.1371/journal.pone.0297688.t004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Peritonitis in patients on PD compared between icodextrin users and non-users by subgroup analyses.

https://doi.org/10.1371/journal.pone.0297688.t005

The causes of death are shown in S3 Table. Cardiovascular disease and infection were the major causes of death. The causes of technique failure are shown in S4 Table. Death and peritonitis were the major causes of technique failure. There were no significant differences in the causes of death and technique failure between the two groups (P = 0.658 and 0.297, respectively.)

Discussion

Our study demonstrated that PD patients using icodextrin were at lower risks of mortality, technique failure, and peritonitis. The subgroup analyses showed that icodextrin could confer additional advantages for lowering the risk of the first episode of peritonitis in those with diabetes and cardiovascular disease.

Icodextrin was introduced in the early 1990s as a glucose-sparing solution for long dwells [19]. Although icodextrin has several clinical benefits on PD care, studies on its impact on the technique and patient survival are limited. Several previous studies showed that icodextrin, as salvage therapy for fluid overload or ultrafiltration failure, could extend technique survival [20, 21]. A Korean retrospective observational study revealed that utilization of icodextrin was associated with lower risks of all-cause mortality (adjusted HR = 0.69, 95% CI = 0.53–0.90) and technique failure (adjusted HR = 0.60, 95% CI = 0.40–0.92) [14]. In this study, icodextrin users were defined as those who had utilized the solution for ≥ 50% of their PD duration [14]. However, information on important clinical data including membrane transport type, residual renal function, and adequacy data was unavailable. A randomized controlled trial among diabetic PD patients in Japan found that the technique survival rate was significantly better in icodextrin users than in non-users (71.4% versus 45.0%) [22]. However, this study was limited by a small sample size (41 patients). In a previous retrospective observational study, we reported that the icodextrin use could decrease technique failure and improve patient survival in incident PD patients, initiating dialysis from 2007 to 2011 [16]. An Australia and New Zealand study using multicenter registry data also found a lower risk of technique failure due to social reasons such as burnout in patients cared at centers with higher icodextrin usage [23]. A recent meta-analysis showed that icodextrin could increase ultrafiltration, reduce episodes of fluid overload, and probably minimize mortality risk. But it had no significant effects on technique failure, peritonitis, or residual renal function decline [13]. The study was limited by lack of long-term assessment, low mortality events (only 32 events) and variable trial qualities. These limitations factors may cause insufficient power to draw any reliable concluding remarks. Our study analyzed data from a large cohort of incident PD patients and could provide further evidence to support the benefits of icodextrin in both technique survival and patient survival.

Studies about the impact of icodextrin on the risk of peritonitis are limited. An Austria multicenter, longitudinal, prospective, observational study found no significant effects of icodextrin on peritonitis rates in incident and prevalent PD patients [15]. In the Peritoneal Dialysis Outcomes and Practice Patterns analysis, the likelihood of cure after a peritonitis episode increased with greater icodextrin use [24]. Our study demonstrated that icodextrin was associated with a reduced risk of the first episode of peritonitis in incident PD patients. The protective effect was more prominent for those with diabetes or cardiovascular disease.

The use of icodextrin has local peritoneal and systemic protective effects by minimizing glucose exposure and optimizing volume status. Icodextrin may preserve the peritoneal membrane function, have a positive influence on peritoneal host defense, and extend the PD duration because it contains low glucose, has a low GDP content and is iso-osmolar [6, 10, 25–27]. It can also enhance ultrafiltration in long dwells and mitigate fluid overload, especially in patients with high and high-average peritoneal transports [10, 11, 13]. The use of icodextrin therefore results in improved fluid balance and blood pressure control and reduction in left ventricle mass [28, 29]. The clearance of small solutes such as sodium and creatinine could be increased by icodextrin through enhanced convection [11]. In addition, icodextrin confers favorable metabolic control. The use of icodextrin could reduce weight gain and HbA1c level and improve dyslipidemia and insulin resistance [9, 12, 30–32]. Poor glycemic control is associated with subsequent risk of exit site, tunnel infection, technique failure, and mortality [33, 34].

The strength of this study is the use of a well-organized database of medical records collected in the recent decade with a sample size large enough to evaluate outcomes after a long follow-up period. There are limitations in this study. The number of exchanges that might influence the rate of peritonitis and medication use were not included in data analysis. Furthermore, this study was observational and retrospective in design. In addition, the prescription of icodextrin was up to physicians’ discretion, which might introduce selection and indication biases. However, multivariate analyses were performed to reduce these biases. This observational study contains valuable information that could be conveyed to the nephrology community to optimize the care of PD patients.

In conclusion, our study highlights the benefits of icodextrin over glucose-based dialysates. The use of icodextrin was associated with lower risks of mortality, technique failure, and the first episode of peritonitis. Further large-scale long-duration randomized controlled studies are necessary to validate our findings.

Supporting information