Association between serum total indoxyl sulfate, intraperitoneal inflammation, and peritoneal dialysis technique failure: a 3-year prospective cohort study

Study design

This prospective observational cohort study was conducted within the research project focused on “Determinants of Peritoneal Dialysis Technique Survival and Possibilities for Pharmacological Correction,” undertaken by the State Institution “Institute of Nephrology of the National Academy of Medical Sciences” in Kyiv, Ukraine, spanning from January 2017 to January 2022 (Domestic Trial Registration Identifier 0117U002122). The study protocol received approval from the Institute’s Ethics Committee (Protocol #7, dated September 12, 2016), and all participants provided informed consent before being included in the study.

Sample size

The sample size calculation for our study utilized MedCalc version 19.2.6 (Ostend, Belgium) and.

G*Power 3.1.9.4 Statistical Software, taking insights from prior research. Our study is the first to examine the relationship between serum tIS concentration and cytokine levels in PDE. Therefore, we referred to studies that investigated variations in PDE IL-6 concentrations and the association between serum tIS concentration and PD technique survival [25, 26]. Zhou et al. reported an effect size of 0.92 with a power of 0.95 and an alpha of 0.05, utilizing independent t-tests for log-transformed data to assess differences in IL-6 levels in PDE between two groups, each comprising 15 patients [26]. Lin et al., in their examination of the association between tIS concentration and PD technique survival, found a Hazard Ratio (HR) of 1.27 (95% CI 1.03–1.61) with survival rates of 0.9% and 0.58% in two groups, totaling 46 patients [25]. Based on these findings, we determined that a minimum of 33 participants per group is necessary to detect differences with a power of 0.80 and an alpha of 0.05, using either the Student’s t-test or the Mann-Whitney test. For survival analysis, at least 73 participants are required to achieve the same power and alpha level. To account for possible dropouts and ensure sufficient study power, we increased the sample size by 30% beyond the initial calculation, resulting in a total of 95 patients for inclusion.

Study cohort

The study focused on patients aged 18 years or older undergoing continuous ambulatory peritoneal dialysis (CAPD) with a technique survival of more than one year. To ensure a homogeneous study cohort, inclusion criteria encompassed patients with well-functioning peritoneal access, a target Kt/V of at least 1.7, and a peritoneal ultrafiltration (UF) capacity exceeding 400 mL over 4 h using a 3.86% glucose solution. The recruitment period spanned from 2017 to 2019, and the patients were followed up for three years. Treatment was provided at the Dialysis Medical Center LLC “Link-Medital” in Odesa, Ukraine, and the State Institution “Institute of Nephrology of the National Academy of Medical Sciences” in Kyiv, Ukraine.

The exclusion criteria comprised patients who had experienced peritonitis or hospitalization for any other reasons and had taken antibiotics or probiotics within the past three months. Additional exclusion criteria included a history of cardiovascular events, hemoglobin levels below 90 g/L, systemic disease, malnutrition, malignancy, acute inflammation, or undergoing immunosuppressive treatment. These criteria were precisely delineated to minimize the potential impact of other immune or inflammatory factors on the concentrations of tIS and cytokines under investigation.

All patients underwent their regularly prescribed dialysis treatment with a dwell time of 4–5 h during the day and 8–10 h at night (4–5 exchanges per day). They were administered commercially available glucose-based Dianeal PD solution (Baxter Healthcare SA, Castlebar, Ireland) with varying glucose concentrations of 1.36% and 2.27%, along with Icodextrine.

Study procedures and baseline evaluation

Baseline demographic and clinical examination data were gathered during the first patient visit after obtaining informed consent. This included acquiring demographic details such as age, gender, comorbid conditions, and information regarding medication use at the time of study enrollment.

The day preceding the first visit, patients were instructed to collect their urine and dialysis drainage over a 24-hour period. During the first visit, tests conducted on these samples included measurements of urea, creatinine, and total volume. Additionally, whole blood samples were collected from patients after an overnight fasting period and processed immediately. Routine biochemical parameters, comprising blood concentrations of urea and creatinine, serum albumin, total protein, C-reactive protein (CRP), glucose, electrolytes, intact parathormone (iPTH), lipid profile parameters, and hemoglobin, were measured during the first visit.

On the night preceding the second visit, scheduled 1 to 2 days after the initial appointment, patients executed their routine overnight dialysis exchange. The ensuing morning at the center, the overnight effluent was completely drained, and a standard peritoneal equilibration test (PET) was conducted, following the protocol outlined by Twardowski et al. [27]. Subsequent PETs were scheduled every 6 months or earlier if clinically necessary, such as within a month following an episode of peritonitis, unexplained fluid retention, or loss of ultrafiltration efficiency.

At the start of the PET, prior to the instillation of the dialysis solution, 5-mL blood samples were collected from each patient for cytokine and tIS measurements. Following the test, 5-mL samples of the overnight drained PDE were also collected from all patients for cytokine analysis. The patients fasted overnight and avoided strenuous activity for 24 h before blood collection. The blood samples underwent centrifugation at 1500 rpm for 10 min to separate the serum. From the serum, 0.5 mL was extracted immediately for tIS determination. The remaining serum from the blood samples and the PDE samples were promptly stored at − 20 °C to preserve the cytokines’ integrity for subsequent assays.

Biochemical marker analyses were conducted using the automated Flexor Junior analyzer (Vital Scientific, Spankeren, the Netherlands). Hematological parameters of blood were assessed using the ABX Micros-60 (Horiba Medical, Montpellier, France). Blood lipid profile parameters included triglyceride (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C). The atherogenic index of plasma (AIP) was computed from plasma triglyceride (TG) and HDL-C (log [TG/HDL-C]). Body mass index (BMI) was calculated as weight in kilograms divided by the square of the height in meters.

Dialysis adequacy was assessed through several measures. The weekly peritoneal creatinine clearance (CrCl), normalized to 1.73 m² of body surface area (BSA), and Kt/V using the Watson formula for body water were measured [28]. Both peritoneal Kt/V and renal Kt/V were estimated separately. The dialysate/plasma creatinine (D/P) ratio was calculated from concentrations in 4-hour dialysate and plasma creatinine in the PET test. Residual kidney function (RKF) was determined as 24-hour urine volume and weekly renal urea clearance (Kt/V) [29]. Anuria was defined as a 24-hour urine volume < 100 mL. Total fluid removal was calculated as the sum of urinary volume and peritoneal UF over 24 h.

Determining tIS serum concentration

The serum tIS concentration was determined using indoxyl sulfate potassium salt (Sigma-Aldrich, St. Louis, MO, USA) with a purity of ≥ 98% via a modified Obermeyer’s reagent method [13]. Briefly, 0.5 mL of serum was mixed with 0.5 mL of a 20% trichloroacetic acid solution, followed by centrifugation at 3000× g for 10 min. The resulting supernatant was combined with a few drops of an alcoholic thymol solution and 1 mL of Obermayer’s reagent. After a 20-minute incubation, 2 mL of chloroform was added, and the absorbance was measured spectrophotometrically at 450 nm. tIS levels were measured in duplicates and expressed in µmol/L.

Cytokine’s measurements

Serum and PDE IL-6, MCP-1, and TNF-α testing were performed using the “SunRise TouchScreen” enzyme and commercially available ELISA kits from IBL International GmbH, Hamburg, Germany. The cytokine analysis followed the manufacturer’s protocol, with samples run in duplicate. The minimum detectable dose for IL-6 quantification was 0.2 pg/mL in serum and 0.7 pg/mL in PDE, with standards ranging from 0 to 300 pg/mL. For MCP-1, the minimum detectable dose was 50 pg/mL in serum and 60 pg/mL in PDE, with standards ranging from 0 to 2000 pg/mL. TNF-α quantification had a minimum detectable dose of 0.2 pg/mL in serum and 0.3 pg/mL in PDE, with standards ranging from 0 to 250 pg/mL. The Tecan SunriseTM Absorbance Microplate Reader was used for ELISA measurements, and the overall inter-assay coefficient of variation was 7.1% for IL-6, 8.5% for MCP-1, and 9.2% for TNF-α.

Study outcome

Throughout a prospective 3-year follow-up period, patients were monitored until they either.

transitioned to HD, underwent kidney transplantation, were lost to follow-up, died, or the study concluded on January 31, 2022. The outcome measure in our study was PD technique failure, defined as the inability to continue PD effectively, leading to a switch to HD. Incidents of technique failure, regardless of the cause, were meticulously documented from patient enrollment to the study endpoint. Patients with transplantation or death during the follow-up period were excluded from the analysis.

UF insufficiency was assessed according to the guidelines set by the International Society for Peritoneal Dialysis: specifically, net ultrafiltration from a 4-hour PET test falling below 100 mL, utilizing a 2.27% glucose/2.5% dextrose solution, or below 400 mL when using 3.86% glucose/4.25% dextrose solution [30]. Before classifying an event as technique failure in patients with PD catheter dysfunctions, each case underwent a thorough examination. This involved consistent and appropriate application of conservative or surgical interventions, ensuring a meticulous evaluation before definitively categorizing it as PD technique failure.

Statistical analysis

Statistical analysis was conducted using MedCalc Statistical Software version 19.2.6 (Ostend, Belgium). Due to the skewed distribution of the majority of variables, quantitative data were expressed as the median (Me) and interquartile ranges (Q25-Q75), and group comparisons were made using the Mann-Whitney test (U-test). Categorical variables were presented as proportions, and the Chi-square (χ2) test was utilized for between-group comparisons. Spearman’s correlation test assessed associations between serum tIS, PDE cytokine levels, and other markers.

Multivariate logistic regression analysis was employed to adjust for the potential confounding factors affecting tIS concentration. Receiver operating characteristic (ROC) curve analysis evaluated the overall discriminative ability of baseline serum tIS concentration for predicting PD technique failure events.

The Kaplan-Meier method estimated cumulative survival rates for time to PD technique failure in PD patients with serum tIS concentrations above and below the best cutoff point, and comparisons were made using the log-rank test.

Cox proportional hazard regression models explored the association between baseline serum tIS concentration and the occurrence of PD technique failure. The analysis was conducted in two stages: initially, an unadjusted model was applied, followed by an adjusted model that included variables significantly associated with elevated serum tIS levels from the multivariate logistic regression analysis.

To validate the robustness and reliability of our findings, a set of sensitivity analyses was conducted. First, we applied a Box-Cox transformation to account for the non-parametric distribution of tIS and cytokine levels data. Subsequently, a two-way ANOVA, complemented by Tukey’s multiple comparisons test, was employed. This analysis aimed to explore the effects of log-transformed tIS, categorized by its median level, on the log-transformed levels of PDE cytokines. Additionally, we examined any potential interaction with the most influential factor identified in the logistic regression analysis. Second, we created a new cohort by excluding patients with diabetes, and Cox regression analysis was subsequently repeated on this subset. Finally, we removed outliers (n = 3) from the tIS variable and reran the Cox regression analysis, treating tIS as a continuous variable within the models. No missing data were encountered in this study due to its prospective nature, ensuring completeness and accuracy in the statistical analyses conducted.