Multicenter Insights into Peritoneal Dialysis for Autosomal Dominant Polycystic Kidney Disease

Abstract

Key Points

• This study demonstrates that peritoneal dialysis is a safe and feasible option for patients with autosomal dominant polycystic kidney disease, even with high cystic organ volumes.

• Larger organ volumes, including height-adjusted cumulative kidney and liver volume, were not associated with increased peritoneal dialysis–related complications.

Background

Autosomal dominant polycystic kidney disease (ADPKD) is the most prevalent genetic kidney disorder and the fourth leading cause of kidney failure. Peritoneal dialysis (PD), preferred for its home-based convenience and cost-effectiveness, is often underutilized in ADPKD because of concerns over enlarged kidneys and heightened risk of complications.

Methods

This retrospective cohort study used data from the Mayo Clinic Polycystic Kidney Disease Database to evaluate individuals with ADPKD undergoing PD. We analyzed demographics, clinical parameters, and PD-related parameters. Complications were correlated with kidney and liver volumes derived from pre-kidney failure imaging.

Results

A total of 155 individuals with ADPKD on PD were included, of whom 45.1% were male. The mean age at PD initiation was 54.3±12.8 years, and the mean body mass index was 28.0±7.0 kg/m². The median duration of PD was 24.3 months (interquartile range [IQR], 10.7–43.1), with 21.9% transitioning to hemodialysis. The most common complications were abdominal hernias (30.3%) and peritonitis (23.9%), with a peritonitis rate of 0.11 episodes per patient-year. Imaging analyses performed on a subset of 50 patients showed a median height-adjusted total kidney liver volume of 2731.9 ml/m (IQR, 2102.5–3131.1) and a median height-adjusted total kidney volume of 1303.5 ml/m (IQR, 733.1–1829.4). Kaplan–Meier analysis demonstrated no differences in complications rates on the basis of height-adjusted total kidney liver volume or height-adjusted total kidney volume (above versus below median values) or body mass index categories. Multivariate Cox regression analysis revealed that higher height-adjusted cumulative organ volume was associated with a lower risk of PD-related complications (hazard ratio=0.56, P = 0.026).

Conclusions

PD is a safe and feasible treatment option for patients with ADPKD, with no increased risk of PD-related complications associated with larger height-adjusted cumulative organ volumes. Infectious complication rates in this cohort were within International Society of Peritoneal Dialysis guideline thresholds, further supporting the safety of PD in this population.

Visual Abstract

Introduction

Autosomal dominant polycystic kidney disease (ADPKD) is the most prevalent genetic kidney disorder and the fourth leading cause of kidney failure (KF), with 50% of patients reaching KF by their sixth decade of life.^1,2 Peritoneal dialysis (PD) is a widely used modality for KRT in patients with KF. PD offers several advantages over hemodialysis, including preservation of residual kidney function, convenience of home-based treatment, and cost-effectiveness.^3–9 Among a total of 5848 patients with ADPKD who initiated KRT from the Renal Epidemiology and Information Network Registry, only 638 (10.9%) were on PD.¹⁰ Historically, this low prevalence can be attributed to clinicians' reluctance to recommend PD to patients with ADPKD for reasons including the theoretical risk that enlarged kidneys can impair a patient's ability to tolerate sufficient intraperitoneal volume to achieve adequate dialysis clearance. There is also concern for PD-related complications including abdominal wall hernia, peritonitis, hydrothorax, and pericatheter and abdominal leaks.¹¹ Despite these concerns, a recent meta-analysis and systematic review that included 14,673 patients on PD from 12 different cohort studies, 931 of which were patients with ADPKD, demonstrated that the presence of ADPKD was associated with decreased mortality risk and was not associated with an increased risk of technique failure or peritonitis compared with non-ADPKD patients.¹² Although a small cohort study suggested an association between height-adjusted total kidney volume (htTKV) and higher intraperitoneal pressure (IPP),¹³ another larger cohort of 60 patients with ADPKD on PD found that higher IPP was not associated with organ volumes.¹⁴ In a small cohort of 24 patients with ADPKD on PD, kidney size by length did not affect outcomes on PD when compared with hemodialysis.¹⁵ However, there is little evidence correlating the cumulative abdominal cystic organ volumes with the incidence of PD complications in patients with ADPKD.

This study aims to evaluate outcomes and complications of PD in a large cohort of patients with ADPKD treated within a multicenter academic health care system. In addition, it seeks to investigate the impact of htTKV, height-adjusted total liver volume (htTLV), combined height-adjusted kidney and liver volume, and body mass index (BMI) on the risk of PD-related complications.

Methods

Patient Selection

We conducted a retrospective, longitudinal cohort study using data from the Mayo Clinic Polycystic Kidney Disease Database including all Mayo sites (Rochester, AZ, FL) from 1990 to 2024. PD therapy was identified using specific International Classification of Diseases, Tenth Revision, Clinical Modification codes for ADPKD, continuous ambulatory PD (CAPD), continuous cycling PD (CCPD), outpatient PD, and PD catheter-related procedures, followed by a manual chart review that was conducted to ensure accuracy. A total of 155 patients with ADPKD who received PD treatment were included in the analysis (Supplemental Figure 1).

Data Collection

Data were collected through a comprehensive review of electronic medical records by two investigators. Demographic details such as age, sex, and ethnicity were extracted using electronic medical records. Clinical parameters (such as weight, height, and comorbid conditions), data related to PD (including age at PD initiation, PD modality [CCPD versus CAPD], duration of PD), and reasons for discontinuation of PD (transplant, transition from PD to hemodialysis) were recorded. Complications arising from PD and their respective management were thoroughly reviewed, including leaks within 30 days of PD catheter placement, catheter site infection, hernia, hydrothorax, hemoperitoneum, and peritonitis rate (episodes per patient-year—calculated as the total number of peritonitis episodes divided by the number of patients-years on PD). Patients were followed throughout their PD duration up to discontinuation—if present—or last follow-up on April 1, 2024. Peritonitis, catheter site infections, leaks, hydrothorax, and hemoperitoneum were recorded only during the active PD period, whereas hernias were documented if they occurred during or after PD. In this study, overall complications are defined as occurrence of any of the PD related complications.

Analyses were restricted to the initial PD period for patients who temporarily discontinued and later resumed PD. Genetic testing was done via targeted next-generation panels, whole-exome sequencing, or whole-genome sequencing as described previously.^16–22 This retrospective cohort study was conducted in accordance with the guidelines of the Mayo Clinic Institutional Review Board and adhered to the principles outlined in the Declaration of Helsinki.

Kidney and Liver Volume Measurements

A subanalysis was conducted on patients with available imaging data to describe the effect of kidney and liver volume on the onset of different PD complications. Fifty patients with an available computed tomography scan with and without contrast or magnetic resonance imaging within 3 years before PD initiation were identified. A deep learning–based planimetry approach, recognized for its accuracy and reliability, was used to segment the kidneys and the liver from the imaging data of our patient cohort.²³ A blinded expert in medical imaging subsequently performed quality checks on the segmentation results using an in-house-developed software tool (polycystic kidney disease-graphical user interface). Both total kidney volume and total liver volume were then calculated programmatically by multiplying the total number of labeled voxels by the voxel volume using Python. Total kidney liver volume was then obtained by adding total kidney volume and total liver volume. These biomarkers were then adjusted for height, measured in meters.

To evaluate potential selection bias related to kidney size in dialysis modality decisions, we identified a comparison cohort of 115 patients with ADPKD who underwent hemodialysis and had imaging data available within 3 years before dialysis initiation, applying the same inclusion criteria used for the PD group. Demographics and htTKV were compared between PD and hemodialysis cohorts to assess potential differences.

Study Outcomes

The primary outcome of this study was to describe the frequency and types of complications associated with PD in patients with ADPKD. Secondary outcomes included evaluating the impact of BMI, as well as kidney and liver organ volumes, on the development of PD-related complications.

Statistical Analysis

Quantitative variables were reported as mean±SD or median (interquartile range [IQR], Q1–Q3), depending on data distribution. Categorical variables were reported as numbers (No.) and percentages (%). Comparisons between male and female patients were conducted using the independent samples t test for continuous variables and chi-squared test for categorical variables. To assess differences in htTKV between PD and hemodialysis cohorts, age-adjusted quantile regression analyses were conducted at the 25th, 50th, and 75th percentiles of the htTKV distribution, adjusting for age at dialysis initiation. A subanalysis focused on patients undergoing PD with imaging data (used to calculate TKLV) and follow-up data for overall complications, hernia, and peritonitis. The Kaplan–Meier (KM) method was used to estimate the freedom from overall complications, peritonitis, and hernia at different time points. To explore the relationships between htTKV, htTLV, height-adjusted total kidney liver volume (htTKLV), and each end point, volume variables were dichotomized at the median observed in the imaging cohort: <1300 and ≥1300 ml/m for htTKV, and <2730 and ≥2730 ml/m for htTKLV. These cutoffs allowed unbiased stratification of patients in KM survival and Cox regression analyses. The log-rank test was used to compare survival curves between groups stratified by volume and adjusted BMI. Univariable Cox regression models were conducted to estimate hazard ratios (HRs) for risk factors, and multivariable Cox regression models were used to evaluate the associations of htTKLV and htTKV with overall complications after adjusting for age and sex separately. A P value < 0.05 was considered statistically significant. Statistical analysis was conducted using JMP Pro software version 18.0 and R4.2.2.

Results

Demographics

Our cohort included 155 patients with ADPKD who underwent PD. Among the included cohort, 45.1% were male and 83.2% were White, with a mean BMI of 28.0±7.0 and a mean age at PD initiation of 54.3±12.9 years. Most patients (76.1%) were hypertensive. Before PD initiation, 5.8% had undergone kidney transplantation and 7.7% had nephrectomy. Genetic testing was performed in 26 (16.8%) patients, with PKD1 pathogenic variants being the most common, followed by PKD2. One patient had a DNAJB11 pathogenic variant, and one patient had both PKD1 and COL3A4 pathogenic variants. Baseline demographics, comorbidities, and genetic characteristics are shown in Table 1.

Table 1.

Baseline demographics of patients with autosomal dominant polycystic kidney disease undergoing peritoneal dialysis

Variable	Total	Males	Females	P Value
No.	155	70	85
Age at PD initiation, yr, mean (±SD)	54.3 (12.9)	53.9 (13.3)	54.6 (12.6)	0.75
White, No. (%)	129 (83.2)	59 (84.3)	70 (82.4)	0.66
BMI, kg/m 2
No.	119	54	65	0.08
Mean (±SD)	28.0 (7.0)	29.2 (4.9)	26.9 (8.2)
Adjusted BMIa, kg/m 2
No.	48	23	25	0.95
Mean (±SD)	28.4 (5.4)	28.4 (5.9)	28.5 (4.9)
Comorbidities, No. (%)
Hypertension	118 (76.1)	54 (77.1)	64 (75.3)	0.79
History of KT	9 (5.8)	4 (5.7)	5 (5.9)	0.96
History of nephrectomy	12 (7.7)	8 (11.4)	4 (4.7)	0.12
PKD genotype, No. (%)	26 (16.8)	10 (14.3)	16 (18.8)	0.88
PKD1	22 (84.6)	8 (80.0)	14 (87.5)
PKD2	2 (7.7)	1 (10.0)	1 (6.2)
Otherb	2 (7.7)	1 (10.0)	1 (6.2)

BMI, body mass index; KT, kidney transplant; PD, peritoneal dialysis; PKD, polycystic kidney disease.

^aAdjusted body mass index was calculated by dividing the adjusted weight by the height squared. The adjusted weight was calculated by subtracting the kidney weight from the body weight, assuming a kidney tissue density equal to that of water.

^bOther variants include DNAJB11, PKD1-COL3A4.

PD Characteristics

Detailed characteristics of PD modality, duration, and reason for PD termination are shown in Table 2. The majority of patients (80.6%) initiated PD as their first KRT modality, whereas the remainder transitioned from hemodialysis. CCPD was more common than CAPD, with 53.5% of patients using CCPD compared with 16.8% using CAPD, whereas the PD modality was not specified in medical records in 29.7% of patients. The median duration on PD was 24.3 months (IQR, 10.7–43.1), with no significant differences observed between sexes. Overall, 21.9% of patients transitioned from PD to hemodialysis. Kidney transplantation was the most common reason for discontinuing PD (72.4%), followed by PD-related complications (16.3%). Three patients had inadequate clearance on PD while one patient discontinued PD because of discomfort associated with the procedure. The reasons for discontinuing PD were not significantly different between sexes.

Table 2.

Peritoneal dialysis characteristics including duration, modality, and reason to discontinue peritoneal dialysis in patients with autosomal dominant polycystic kidney disease

Variable	Total	Males	Females	P Value
No.	155	70	85
PD as first modality started, No. (%)	125 (80.6)	56 (80.0)	69 (81.2)	0.85
Duration of PD treatment, mo, median (Q1–Q3)	24.3 (10.7–43.1)	24.3 (10.7–43.1)	19.8 (11.4–36.1)	0.42
PD modality, No. (%)				0.58
CCPD	83 (53.5)	35 (50.0)	48 (56.5)
CAPD	26 (16.8)	14 (20.0)	12 (14.1)
Not specified	46 (29.7)	21 (30.0)	25 (29.4)
Modality switch (from PD to hemodialysis), No. (%)	34 (21.9)	13 (18.6)	21 (24.7)	0.36
PD discontinuation, No. (%)	123	58	65	0.43
Transplant	89 (72.4)	45 (77.6)	44 (67.7)
Complicationsa	20 (16.3)	7 (12.1)	13 (20.0)
Othersb	14 (11.4)	6 (10.3)	8 (12.3)

CAPD, continuous ambulatory peritoneal dialysis; CCPD, continuous cycling peritoneal dialysis; PD, peritoneal dialysis.

^aComplications include hernia, peritonitis, hydrothorax, hemoperitoneum.

^bOthers include inadequate clearance, quality of life, and reasons not specified in the charts.

PD Complications

PD-related complications are summarized in Table 3. There was no significant difference between both sexes for PD-related complications. Abdominal wall hernia was the most common complication observed in 30.3% of patients at a median onset time of 20.3 months (IQR, 8.5–43.6) after PD initiation. Nearly half of patients (53.1%) who developed a hernia were treated with surgical repair. Peritonitis was the second most common complication (23.9%), followed by hemoperitoneum (3.2%), hydrothorax (1.3%), catheter site infection (1.3%), and pericatheter fluid leakage (0.6%).

Table 3.

Peritoneal dialysis–related complications in patients with autosomal dominant polycystic kidney disease

Variable	Total	Male	Female	P Value
No.	155	70	85
Hemoperitoneum, No. (%)	5 (3.2)	4 (5.7)	1 (1.2)	0.11
Hydrothorax, No. (%)	2 (1.3)	0 (0.0)	2 (2.4)	0.20
Abdominal wall hernia, No. (%)	47 (30.3)	20 (28.6)	27 (31.8)	0.67
Time to onset after PD initiation, mo, median (Q1–Q3)	20.3 (8.5–43.6)	15.5 (4.8–54.8)	24.0 (10.5–36.8)	0.38
Managed with surgical repair, No. (%)	25 (53.1)	11 (55)	14 (51.8)	0.79
At least one episode of peritonitis, No. (%)	37 (23.9)	13 (18.6)	24 (28.2)	0.16
No. of episodes per patient-year	0.11	0.08	0.13
PD fluid leak within 30 d of PD catheter insertion, No. (%)	1 (0.6)	0 (0.0)	1 (1.2)	0.36
Catheter site infection, No. (%)	2 (1.3)	1 (1.4)	1 (1.2)	1

PD, peritoneal dialysis.

Imaging Characteristics

Demographics and imaging data were available for 50 patients on PD and 115 patients on hemodialysis as shown in Table 4. The two groups were similar in terms of sex distribution (48.0% versus 55.7% male, P = 0.40) and age at dialysis initiation (55.9±13.7 versus 59.0±12.8 years, P = 0.14). The median time from imaging to KF was also comparable between groups (7.0 months in both cohorts).

Table 4.

Descriptive analysis of demographics and imaging characteristics in patients with autosomal dominant polycystic kidney disease undergoing peritoneal dialysis and hemodialysis

Variable	PD	Hemodialysis	P Value
No.	50	115
Male, No. (%)	24 (48.0)	64 (55.7)	0.40
White, No. (%)	43 (86.0)	103 (89.6)	0.38
Age at dialysis initiation, yr, mean (±SD)	55.9 (13.7)	59.0 (12.8)	0.14
Time of imaging before KF, mo, median (Q1–Q3)	7.0 (3.5–15.7)	7.0 (2.0–20.0)
htTKV, median (Q1–Q3)	1303.5 (733.1–1829.4)	1646.6 (194.2–8795.4)	0.048
htTLV, ml/m
No.	50	106
Median (Q1–Q3)	1086.0 (984.0–1425.9)	1164.6 (941.0–1723.0)	0.73
htTKLVa, mL/m, median (Q1–Q3)	2731.9 (2102.5–3131.1)	2919.8 (2071.1–4311.5)	0.13
MIC, No. (%)			0.037
1A	4 (8.0)	6 (5.4)
1B	7 (14.0)	17 (14.7)
1C	11 (22.0)	36 (31.3)
1D	20 (40.0)	28 (24.3)
1E	5 (10.0)	24 (20.9)
2B	2 (4.0)	0 (0)
Could not be calculatedb, No. (%)	1 (2.0)	4 (3.4)

htTKV, height-adjusted total kidney volume; htTKLV, height-adjusted total kidney liver volume; htTLV, height-adjusted total liver volume; KF, kidney failure; MIC, Mayo Imaging Classification; PD, peritoneal dialysis.

^aTotal kidney liver volume; the sum of total kidney volume and total liver volume.

^bCould not be calculated because of age older than 80 years.

The median htTKV for the total PD cohort was 1303.5 ml/m (IQR, 733.1–1829.4), whereas the median htTLV for the total PD cohort was 1086.0 ml/m (IQR, 984.0–1425.9). The median htTKLV for the total cohort was 2731.9 ml/m (IQR, 2102.5–3131.1). The distribution of Mayo Imaging Classification (MIC) is shown in Table 4. Most PD patients were classified as MIC1D (40%), followed by MIC1C (22%) and MIC1B (14%). Interestingly, MIC1A was seen in four (8%) patients, three of whom had coexisting kidney disorders including focal segmental glomerular sclerosis, amyloidosis, or prior AKI.

Comparison of htTKV and htTLV between PD and Hemodialysis Groups

The median htTKV was significantly lower in the PD group (1303.5 ml/m [IQR, 733.1–1829.4]) compared with the hemodialysis group (1646.6 ml/m [IQR, 194.2–8795.4], P = 0.048). Age-adjusted quantile regression analyses, as shown in Table 5, showed no significant differences at the 25th percentile (estimate=−132.07; 95% confidence interval [CI], −467.01 to 19.6; P = 0.44) or 50th percentile (estimate=−278.84; 95% CI, −666.04 to −32; P = 0.17). However, at the 75th percentile, patients in the hemodialysis group had significantly higher htTKV (estimate=−723.87; 95% CI, −1054 to −319.62; P = 0.014). By contrast, there was no statistically significant difference in htTLV between the PD group (1086.0 ml/m [IQR, 984.0–1425.9]) and the hemodialysis group (1164.6 ml/m [IQR, 941.0–1723.0], P = 0.73).

Table 5.

Age-adjusted quantile regression analysis comparing height-adjusted total kidney volume between peritoneal dialysis and hemodialysis patients at the 25th, 50th, and 75th percentiles

Comparison Between PD and Hemodialysis Groups	Regression Analysis
	Estimate (95% CI)	P Value
Quantile regression at 25th percentile of htTKV	−132.07 (−467.01 to 19.6)	0.44
Quantile regression at 50th percentile of htTKV	−278.84 (−666.04 to −32)	0.17
Quantile regression at 75th percentile of htTKV	−723.87 (-1054 to −319.62)	0.014

CI, confidence interval; htTKV, height-adjusted total kidney volume; PD, peritoneal dialysis.

Survival Analysis and Cox Regression for Risk of Complications

Imaging-Based Predictors of PD Complications

A complication-free survival plot comparing patients on the basis of htTKV and htTKLV is shown in Figure 1. Patients with a htTKLV ≥2730 ml/m had a significantly lower risk of overall complications at any time point (P = 0.050). In the univariate Cox proportional hazards model, both a 1000 ml/m increase in htTKLV (HR=0.56; 95% CI, 0.34 to 0.93; P = 0.026) and htTKV (HR=0.53; 95% CI, 0.29 to 0.99; P = 0.047) were significantly associated with a reduced risk of complications. However, a 1000 ml/m increase in htTLV was not significantly associated with overall complications (P = 0.240). In multivariable models, the association between an increase in htTKLV and reduced risk of complications remained significant after adjustment for sex (HR=0.56; 95% CI, 0.34 to 0.94; P = 0.028) or age (HR=0.45; 95% CI, 0.25 to 0.83; P = 0.010). Similarly, the association between an increase in htTKV and reduced complications persisted after adjustment for age (HR=0.50; 95% CI, 0.26 to 0.95; P = 0.035) but was NS after adjustment for sex (HR=0.53; 95% CI, 0.29 to 1.00; P = 0.051). Univariable and multivariable Cox proportional hazards model results are summarized in Tables 6 and 7.

Figure 1.

KM curves illustrating freedom from overall complications among patients with ADPKD on PD. (A) Stratification by height-adjusted cumulative volume (<2730 versus ≥2730 ml/m). (B) Stratification by height-adjusted total kidney volume (<1300 versus ≥1300 ml/m). ADPKD, autosomal dominant polycystic kidney disease; KM, Kaplan–Meier; PD, peritoneal dialysis.

Table 6.

Univariable Cox regression models predicting overall complications in patients with autosomal dominant polycystic kidney disease undergoing peritoneal dialysis

Variables	Univariable
	No.	HR (95% CI)	P Value
Sex	50
Male	24	Ref	Ref
Female	26	1.28 (0.54 to 3.05)	0.572
Age at PD initiation	50	0.98 (0.95 to 1.01)	0.21
htTKV per (1000) ml/m increase	50	0.53 (0.29 to 0.99)	0.047
htTLV per (1000) ml/m increase	50	0.51 (0.17 to 1.56)	0.240
htTKLV per 1000 ml/m increase	50	0.56 (0.34 to 0.93)	0.026
Adjusted BMI	48
≥18.5 and <25	17	Ref	Ref
≥25 and <30	16	1.09 (0.35 to 3.37)	0.88
≥30	15	0.76 (0.23 to 2.54)	0.66

BMI, body mass index; CI, confidence interval; HR, hazard ratio; htTKLV, height-adjusted total kidney liver volume; htTKV, height-adjusted total kidney volume; htTLV, height-adjusted total liver volume; PD, peritoneal dialysis.

Table 7.

Multivariable Cox regression models predicting overall complications in patients with autosomal dominant polycystic kidney disease undergoing peritoneal dialysis

Variables	Multivariable Model 1		Multivariable Model 2
	HR (95% CI)	P Value	HR (95% CI)	P Value
htTKV per 1000 ml/m increase	0.53 (0.29 to 1.00)	0.051	0.50 (0.26 to 0.95)	0.035
htTKLV per 1000 ml/m increase	0.56 (0.34 to 0.94)	0.028	0.45 (0.25 to 0.83)	0.010

Multivariable model 1: sex was used as a covariate. Multivariable model 2: age at peritoneal dialysis initiation was used as a covariate. CI, confidence interval; HR, hazard ratio; htTKLV, height-adjusted total kidney liver volume; htTKV, height-adjusted total kidney volume.

KM analysis of peritonitis-free or hernia-free survival in patients with imaging data is shown in Figure 2. As compared with patients with lower volumes, htTKLV ≥2730 ml/m or htTKV ≥1300 ml/m were not significantly associated with a lower likelihood of peritonitis-free survival (P = 0.057 and 0.061, respectively; Figure 2, A and B) or hernia-free survival (P = 0.950 and 0.928, respectively; Figure 2, C and D).

Figure 2.

KM curves illustrating freedom from peritonitis and hernia among patients with ADPKD on PD. (A) KM curve illustrates freedom from peritonitis stratified by height-adjusted cumulative volume (<2730 versus ≥2730 ml/m). (B) KM curve illustrates freedom from peritonitis stratified by height-adjusted total kidney volume (<1300 versus ≥1300 ml/m). (C) KM curve illustrates freedom from hernia stratified by height-adjusted cumulative volume (<2730 versus ≥2730 ml/m). (D) KM curve illustrates freedom from hernia stratified by height-adjusted total kidney volume (<1300 versus ≥1300 ml/m).

BMI and Risk of PD Complications

The univariate Cox proportional hazards model showed no significant associations between adjusted BMI categories and the risk of complications (P = 0.88 for BMI 25–29.9; P = 0.65 for BMI ≥30, compared with patients with BMI 18.5–24.9), as shown in Table 6. Similarly, KM survival analysis demonstrated no statistically significant differences in freedom from complications, peritonitis, or hernia across BMI categories (P = 0.793, 0.367, and 0.201, respectively). Detailed KM survival rates for BMI categories are shown in Figure 3.

Figure 3.

KM curves illustrating freedom from overall complications, peritonitis, and hernia among patients with ADPKD on PD stratified by BMI. (A) KM curve illustrates freedom from overall complications stratified by BMI categories. (B) KM curve illustrates freedom from peritonitis stratified by BMI categories. (C) KM curve illustrates freedom from hernia stratified by BMI categories. BMI categories: ≥18.5 and <25 (red), 25–30 (green), and ≥30 (blue). BMI, body mass index.

Detailed KM survival rates for peritonitis, hernia, and other complications across various time points are summarized in Supplemental Table 1.

Discussion

This study evaluates the safety and outcomes of PD in patients with ADPKD treated across multiple Mayo Clinic sites, offering new insights into the relationship between cystic organ volumes and PD-related complications. Notably, our findings demonstrate that larger organ volumes, including htTKLV and htTKV, are not associated with an increased risk of PD-related complications. These results underscore that PD is a safe and viable treatment option for patients with ADPKD, even in the context of high cystic organ volumes.

Interestingly, larger kidney and combined kidney and liver volumes were associated with a reduced risk of PD-related complications (HR=0.56, P = 0.02). This finding may be attributed to the lack of a significant correlation between intra-abdominal organ volumes and IPP, which is a known risk factor for complications such as peritonitis, abdominal wall hernias, and other PD-related complications.^14,24 Although previous studies have investigated kidney size in relation to PD outcomes, detailed analyses directly linking organ volume to PD complications have been limited. A retrospective study by Courivaud et al.¹⁵ found no significant differences in kidney length, a suboptimal measure of TKV, between ADPKD patients on PD with and without complications. Another small cohort of 15 patients with ADPKD found that only one patient, who had the largest organ volume of 8912 ml, developed dialysate leakage, leading to discontinuation of PD.²⁵

To address the potential for selection bias—specifically, the effect of larger kidney size on dialysis modality selection—we compared htTKV between PD and hemodialysis cohorts. Although the median htTKV was significantly higher in hemodialysis patients on univariate analysis, this difference should be interpreted with caution. The hemodialysis cohort was slightly older on average and because of the relatively smaller number of PD patients with imaging. In addition, the age-adjusted quantile regression analyses showed no significant differences in htTKV between groups at the 25th and 50th percentiles. A statistically significant difference was observed only at the 75th percentile, where patients in the hemodialysis group had substantially higher htTKV. The sex distribution between the two groups was similar (48.0% versus 55.7% male, P = 0.40), and therefore sex was not included as a covariate in the regression model. These findings suggest that dialysis modality selection may have been influenced by extreme kidney volumes in a minority subset of patients, rather than reflecting a systematic bias across the entire volume distribution.

Furthermore, our analysis revealed no significant association between BMI and the overall rates of PD complications, including abdominal wall hernias and peritonitis. Although Sigogne et al.¹⁴ demonstrated a strong correlation between BMI and IPP (β=0.372, P < 0.01) in a multicenter cohort of 60 patients with ADPKD with a mean BMI of 26±4.2 kg/m², our findings suggest that BMI alone does not predict PD-related complications.

Transition from PD to hemodialysis occurred in 21.9% of patients in our cohort, a rate comparable with the 23.2% observed in a French national cohort study after 3 years of follow-up.¹⁰ Kidney transplantation was the predominant reason for PD discontinuation, accounting for 72.4% of cases, consistent with the French study, where patients with ADPKD who received kidney transplants were more likely to have been treated with PD than hemodialysis (59.7% versus 49.9%; P < 0.01).¹⁰

The most frequent PD complications in our cohort were abdominal wall hernias (30.3%) and peritonitis (23.9%). The hernia rate aligns with previous studies, such as Boyer et al.,²⁶ who reported rates of 7%–27% in PD patients, and Del Peso et al.,²⁷ who found a prevalence of 37%. A meta-analysis of 14,673 PD patients, including 931 with ADPKD, further established that ADPKD is a significant risk factor for hernias, with an odds ratio of 2.11 (95% CI, 1.34 to 3.33).¹²

Peritonitis, the second most common complication, occurred at a rate of 0.11 episodes per patient-year, substantially lower than rates reported in prior studies.^28,29 For example, a prospective cohort comparing 212 non-ADPKD patients and 106 patients with ADPKD found a peritonitis rate of 0.54 episodes per patient-year in patients with ADPKD.²⁸ Similarly, Kumar et al.²⁹ reported one peritonitis episode per 26 months among 56 patients with ADPKD. The notably lower peritonitis rate in our cohort likely reflects the rigorous patient training, consistent use of exit-site topical antibiotics, and comprehensive care provided at Mayo Clinic. In addition, favorable social determinants of health among our patient population may have contributed to these outcomes. These findings align with the International Society of Peritoneal Dialysis recommendation to maintain peritonitis rates below 0.4 episodes per patient-year, underscoring the success of our infection prevention protocols.³⁰

This study is the first to explore the relationship between organ volumes and PD-related complications in patients with ADPKD using a novel deep-learning planimetry segmentation model applied to a large cohort. Although we compared our cohort with a group of patients with ADPKD undergoing hemodialysis for evaluating selection bias, the lack of non-ADPKD patients on PD limits the broader applicability of these findings and prevents direct comparisons across populations. In addition, the potential for selection bias is significant because this study included only patients who opted for PD and were deemed suitable by their physicians. Those with significant abdominal discomfort or other contraindications to PD were excluded, potentially skewing the interpretation of the relationship between organ volumes and complications. Another limitation is the unavailability of consistent data on PD prescriptions—including fill volume, dwell time, and total daily volume delivered—because of variations in documentation practices over the long study period, especially in earlier records. This prevented a systematic assessment of whether prescription characteristics influenced complication rates. Moreover, the relatively small sample size and the low number of PD-related complication events restricted the statistical power to perform multivariable analyses. This limitation may have constrained the ability to fully evaluate the interplay of organ volumes, other risk factors, and outcomes in this cohort.

Future studies should aim to address these gaps. A prospective, multicenter cohort study comparing patients with ADPKD on PD with non-ADPKD patients on PD would provide more generalizable findings and allow for direct comparisons across different populations and modalities. In addition, longitudinal studies tracking patients with ADPKD from PD initiation through transitions to hemodialysis or kidney transplantation would offer critical insights into long-term technique survival, complication risks, and factors influencing treatment outcomes. Incorporating more diverse patient populations and expanding sample sizes will further strengthen the understanding of PD's role in managing ADPKD.

In conclusion, this study demonstrates that PD is a safe and feasible option for patients with ADPKD, even with high cystic organ volumes. Larger organ volumes, including height-adjusted cumulative kidney and liver volume, were not associated with increased PD-related complications. Complication rates, including peritonitis and abdominal wall hernias, were low, reflecting the impact of rigorous infection prevention and patient education. Although limited by its retrospective design and lack of control groups, these findings support the feasibility of PD in ADPKD and highlight the need for future prospective studies to further refine treatment strategies.

Supplementary Material

kidney360-6-1918-s001.pdf ^{(1.4MB, pdf)}

kidney360-6-1918-s002.pdf ^{(232.5KB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/B166.

Author Contributions

Conceptualization: Nabeel Aslam, Abdul Hamid Borghol, Fouad T. Chebib, Ahmad Ghanem, Fadi George Munairdjy Debeh.

Data curation: Bassel Alkhatib, Abdul Hamid Borghol, Marie Therese Bou Antoun, Ahmad Ghanem, Dana Hanna, Fadi George Munairdjy Debeh, Nay Nader, Stefan Paul, Vineetha Rangarajan, Levon Souvalian.

Formal analysis: Ahmad Ghanem, Zhuo Li, Fadi George Munairdjy Debeh.

Investigation: Abdul Hamid Borghol, Ahmad Ghanem, Fadi George Munairdjy Debeh, Vineetha Rangarajan.

Methodology: Adriana Gregory, Timothy Kline, Zhuo Li.

Project administration: Nabeel Aslam, Fouad T. Chebib, Ahmad Ghanem, Fadi George Munairdjy Debeh.

Software: Adriana Gregory, Timothy Kline.

Supervision: Nabeel Aslam.

Validation: Ahmad Ghanem, Zhuo Li, Fadi George Munairdjy Debeh.

Visualization: Ahmad Ghanem, Fadi George Munairdjy Debeh.

Writing – original draft: Ahmad Ghanem, Fadi George Munairdjy Debeh.

Writing – review & editing: Nabeel Aslam, Lyle W. Baker, Fouad T. Chebib, John J. Dillon, Dana Hanna, LaTonya J. Hickson, Andrea Kattah, Sayeed Khalillullah, Sandhya Manohar, Michael M. Mao, Ivan E. Porter, Levon Souvalian, Christopher L. Trautman.

Funding

None.

Data Availability Statements

Partial restrictions to the data and/or materials apply. Deidentified data are available upon reasonable request to the corresponding author.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/KN9/B167.

Supplemental Table 1. KM estimates of freedom from overall complication, peritonitis, and hernia in the cohort with imaging.

Supplemental Figure 1. Study flow chart and cohort selection shows patients' selection, exclusions, and subgroup analyses, including assessments of BMI and complications, as well as kidney and liver volumes in relation to complications.