Is CAPD Feasible for Cirrhotic Patients with Refractory Ascites?

Introduction

The incidence of chronic liver disease (CLD) among patients with ESKD is rising due to overlapping risk factors such as diabetes, metabolic dysfunction-associated steatotic liver disease (MASLD), and improved ESKD survival.¹ While combined liver-kidney transplantation is the definitive treatment for this condition, organ scarcity, technical complexity, and high costs limit its feasibility. Hemodialysis (HD) is often challenging in patients with cirrhosis owing to hypotension, hypoalbuminemia, cirrhotic cardiomyopathy, and autonomic neuropathy,² with rapid fluid shifts potentially precipitating hepatic encephalopathy.^2,3 Altered coagulation further complicates anticoagulation strategies during HD.

Continuous ambulatory peritoneal dialysis (CAPD) provides gradual ultrafiltration without the need of anticoagulation,⁴ facilitating effective clearance of uremic toxins and ascites, while reducing the need for frequent paracentesis in refractory ascites.^5-9 Yet, low serum albumin and risk of spontaneous bacterial peritonitis (SBP) raise concerns regarding hypoalbuminemia and peritonitis.⁷ The primary objective of this study was to evaluate patient survival at 1 year; the secondary objective was to study the peritonitis rate, change in serum albumin, and technique survival amongst patients with ESKD and cirrhotic ascites.

Materials and Methods

We conducted a retrospective observational single-center study from 1^st January 2018 to 31^st March 2023, which included all patients with CLD and refractory ascites on maintenance HD who underwent CAPD catheterization in the department of nephrology at a tertiary care center in northern India. All patients were on maintenance HD prior to CAPD; none received CAPD as the first kidney replacement therapy (KRT) modality. Exclusion: Contradiction to CAPD like abdominal stoma or surgeries, malignancy, severe cardiopulmonary disease, patients planned for liver-kidney transplant within 3 months, abdominal tuberculosis, pregnancy. This work was conducted according to the principles of the Declaration of Helsinki and good clinical practice. As the study involved only chart review and the patient details were anonymized and de-identified, ethics committee clearance was not required.

CLD was diagnosed clinically (stigmata of the liver cirrhosis), via imaging (coarse echotexture of the liver on ultrasonography, median > 7kPa on elastography), or by histopathology. Liver stiffness was measured by transient elastography (FibroScan® 502 Touch; Echosens, France) by an experienced hepatologist; median values > 7 kPa were used to support the diagnosis. Refractory ascites was defined as the need for >5 therapeutic paracenteses in the last 3 months after optimal clinical, dietary, and fluid management. Baseline peritoneal infection refers to episodes of SBP in the 12 months preceding CAPD initiation.

Before insertion of the PD catheter, pre-insertion evaluation, including clinical assessment (stigmata of cirrhosis, abdominal exam), ultrasonography, and ascitic fluid analysis, including routine culture and cell count, was conducted to exclude active SBP. Laparoscopic surgical insertion was done under local anesthesia by the urologist team. The deep cuff was seated within the rectus sheath, and the superficial cuff was in the subcutaneous tissue. A plain abdominal X-ray was done to confirm the position post-placement.

For ascites drainage, PD was started after 48 hours of catheter placement. On the first day, 2-3 L of ascitic fluid was drained, followed by a 500 mL exchange every 30 minutes with two bags containing 1-1.5 L of ultrafiltrate each day, for the next 5-7 days, until all the ascitic fluid was drained. The 500 mL exchanges continued till the 14^th day. After 2 weeks, full-volume exchanges were continued with appropriate dwell time and dextrose concentration. Human albumin (20%) was supplemented if serum albumin dropped below 2.5 g/dL in the initial 7-10 days.

A renal dietitian provided an individualized diet chart for each patient. It was reviewed each month. The initial protein target was 1.2 g/kg/day per International Society of Peritoneal Dialysis nutritional guidelines; mean actual intake was 1.25 g/kg/day (monitored by monthly dietitian assessment).

A monthly patient visit was scheduled. If the patient was unable to come, teleconsultation was utilized. A PD technician visited every alternate day for 2 weeks. Thereafter, twice monthly for the first 2 months, then monthly, and, subsequently, as needed.

At baseline, complete blood count (CBC), liver function test, kidney function test, prothrombin time (PT),and international normalized ratio (INR) were recorded. CBC and serum albumin were repeated each month.

The study’s objectives included survival of cirrhotic patients with refractory ascites on CAPD, determination of the peritonitis rate, and comparison of the change in serum albumin from baseline to follow-up.

Results

Over 5 years, 34 patients underwent CAPD [Figure 1]. Baseline characteristics are presented in Table 1. Follow-up ranged from 8 to 15 months (mean 12 ± 2.1 months). Alcohol liver disease was the most common etiology for CLD (n=15, 44%), followed by viral (n=9, 26%). Diabetic kidney disease (n=15, 44%) was the most common etiology for ESKD, followed by chronic glomerulonephritis (CGN) (n=10, 29%) [Table 1]. Around 30 (88%) patients underwent PD insertion by the open surgical method. All patients had early start PD within 48 hours as a protocol. None of the patients in our cohort required temporary HD sessions during CAPD.

Close

Figure 1:

Flowchart showing screening, inclusion, exclusion, and final analysis.

Export to PPT

At the 1-year follow-up, 32 (94%) patients survived. One patient died from worsening CLD with hepatic encephalopathy and aspiration pneumonia. The other patient developed a myocardial infarction. Two patients (6%) experienced technique failure (defined as permanent transfer to HD) at months 9 and 11, both following ultrafiltration failure after peritonitis. Both patients later experienced recurrent ascites requiring therapeutic paracentesis. One patient eventually stabilized with standard ultrafiltration protocols and did not need further hospitalizations, while the other continued to have refractory ascites and ultimately underwent transjugular intrahepatic portosystemic shunt (TIPS).

The peritonitis rate was 1 episode/45 patient-months. The peritonitis episodes during HD in the last 12 months (SBP) were significantly less than those after PD initiation (3 vs. 9; p=0.04). All episodes were managed conservatively with antibiotics without the need for catheter removal.

PD-related complications were observed in five (15%) patients with one omental wrap at 13 weeks (required omentopexy), three pericatheter leaks (all within the initial 4 weeks), and one umbilical hernia at 3 weeks. All these were managed conservatively. The single umbilical hernia was managed non-operatively with an elastic abdominal binder, manual reduction as needed, and low-volume exchanges for 2 weeks, with no progression or need for surgical repair. For pericatheter leaks and hernias, dwell volumes were reduced to 1 L or less, and the exchange frequency was increased to shorten dwell time.

Laboratory values at baseline and 12 months did not show any significant difference [Table 2]. The mean value of monthly serum albumin is shown in Figure 1, which shows an initial 6% dip during the first 2 months. By 3 months, albumin values stabilized and were maintained throughout the study. Child-Pugh score and Model for End Stage Liver Disease (MELD) score did not show any significant change at the end of the study [Table 2].

The Kaplan-Meier estimated event-free survival was 97.1% (95% CI 91.4-100%) at 8 months and 93.5% (95% CI 84.6–100%) at 10 months, with the survival probability plateauing at 93.5% through 15 months. Median event-free survival was not reached, as fewer than 50% of subjects experienced the event [Figure 2].

Close

Figure 2:

Mean serum albumin levels (blue line) over a 12-month follow-up period.

Export to PPT

Discussion

This is one of the first studies to evaluate the suitability and tolerability of CAPD among ESKD patients with cirrhosis with refractory ascites from India. Our findings indicate that continuous treatment via PD not only offers significant benefits in managing CLD and ascites but also facilitates gradual and progressive drainage of ascitic fluid, mirroring the outcomes observed in previous studies that reported successful ascitic fluid drainage using PD catheters [Table 3].^5,6,8,10-14 However, it is important to review and monitor nutritional status, as many patients need a change in dietary prescription. A close and vigilant follow-up is essential.

The survival of patients with refractory ascites is dismal, at about 50% in most prior studies. In our cohort, it was >90, a possible reason being the selection of patients with relatively preserved albumin levels and mild to moderate Child-Turcotte-Pugh (CTP) score and MELD scores. The traditional definition of refractory ascites cannot be applied to patients with ESKD since they are unresponsive to diuretics due to renal failure, and diuretic therapy optimization by adding mineralocorticoid antagonists is usually avoided due to the risk of hyperkalaemia. Hence, despite a mild to moderate stage CLD, patients with End Stage Renal Disease can have difficult-to-manage ascites. Peritoneal dialysis can be considered in these patients as it can lead to a significant improvement in quality of life by reducing the need for repeated paracentesis.

In a recent meta-analysis, there was no difference between PD with cirrhosis and without cirrhosis in ESKD when compared to HD, projecting PD as a viable alternative option for the patients.¹⁵

Our study adds to the evidence on the use of PD as a KRT modality in patients with cirrhosis. Potential issues with PD catheter insertion in patients with ascites can be effectively reduced or managed with meticulous care. In our center, the catheter insertion is done by an experienced team, and a post-procedure X-ray is conducted as per protocol. This has likely contributed to 100% technique survival over 1 year in our cohort. Each case with mechanical complications was identified at the earliest owing to regular follow-up and fast corrective measures, such as exit site resuturing or omentoplexy. Local suturing of the exit-sites and PD for 48 hours is done for early leaks (< 4 weeks), then low-volume exchanges (500 mL dwell) are resumed for 1 week before escalating. We have noted that this deviates from the ISPD recommendation for purely conservative management, and we now follow ISPD guidelines.

The rate of peritonitis significantly increased in the CAPD patients when compared to spontaneous bacterial peritonitis (SBP) before CAPD initiation. However, the peritonitis rate among patients on PD with cirrhosis was comparable to that of previous data from the same institute on patients without cirrhosis on PD (1 episode/45 patient months vs 1 episode/29 patient months) of 12 months. Most infections were due to Gram-negative bacteria, which is consistent with the microbiological spectrum of our center. The results of the present study closely match with De Vecchi’s study, which showed a peritonitis rate of 1 episode/39 patient months.⁶ Few other studies have reported a much higher peritonitis rate than ours.^5,11 As available studies on similar subjects were very old, this may reflect the changing practices, CAPD connectology improvement, patient training and re-training, hygiene, and infection control in the present time.

Another major concern about PD is the protein loss due to exchanges. Now, this issue is important in CLD patients who are mostly hypoalbuminemic. In the initial phase of the study, most patients had a significant drop in serum albumin levels during the first month. In the following months, all the patients achieved their baseline serum albumin levels and were able to maintain them throughout the study [Figure 3]. Although protocol allowed albumin 20 % infusion for < 2.5 g/dL, no patient’s serum albumin fell below this threshold; hence, albumin infusions were not administered. In the study by Selgas et al. on eight patients with CLD, there was a significant loss in the initial period (>30 g/day) with a gradual decline in subsequent months (<10 g/day).⁸ Similarly, De Vecchi et al. reported a decline in protein loss with the passage of time.⁶ In the study by Marcus et al., only one patient among nine developed hypoalbuminemia.⁵ It is hypothesized that elevated intra-abdominal pressure provides counterpressure against portal hypertension, which in turn reduces ascites formation, thereby diminishing protein loss.¹⁶ It is also possible that the marked enhancement of appetite, attributable to the superior tolerability of PD, contributes to the maintenance of serum albumin concentrations. Nevertheless, the importance of systematic reinforcement of nutritional guidance and dietary counselling should not be underestimated.

Various protocols have been published to drain the ascites while initiating PD. Bajo et al. suggested draining 20% extra fluid over the infused volume in each exchange.¹¹ Chaudhary et al., reported their practice of draining 1 L of peritoneal fluid before starting PD, infusing 1 L of 1.5% dextrose PD solution, and draining 1.5 L with a dwell time of 1-2 hours.¹⁶ This is continued till the peritoneal fluid is drained, preferably performed as an inpatient procedure. Our protocol was similar, but we drained a fixed volume of 1-1.5 L for 7-10 days till all ascitic fluid was drained. The ascitic drainage was one of the main reasons for switching from HD in the study. PD obviated the need for repeated ascitic fluid tapping in all patients.

There are several limitations to the present study, like retrospective nature, small sample size, and short follow-up. Patients did not have very low serum albumin and were on a rigorous follow-up with hepatology and nephrology services. We were unable to retrieve peritoneal membrane characteristics and clearance data. Another shortcoming was that this study lacked a PD first cohort for comparison, limiting assessment of first-line CAPD in ESKD with CLD. However, this is the first report from India to analyze the suitability of CAPD among CLD with refractory ascites and ESKD.

In conclusion, the survival of carefully selected CLD patients with ascites and ESKD on CAPD over 1 year is ∼90% with no patient requiring a switch back to HD. The risk of peritonitis is increased when compared to the frequency of SBP in CLD patients on HD; however, it is comparable to peritonitis among PD patients and can be managed conservatively. There was no additional risk of worsening of albumin levels beyond 2 months in ESKD patients with CLD and refractory ascites. However, active surveillance for malnutrition and a multidisciplinary approach is key. Larger studies with longer follow-ups are needed to confirm and validate our findings.