A total of 11 patients were included. The mean age was 77 years (range: 58–93 years), with seven men (63.6%) and four women (36.4%). The burden of comorbidities was significant, with a mean Charlson score of 8.

All patients had advanced heart failure, NYHA functional class III or IV, associated with stage 4 chronic kidney disease. At the start of peritoneal dialysis, the glomerular filtration rate ranged from 20 to 25 mL/min/1.73 m².

The main demographic and clinical characteristics of the study population are presented in Table I

The cohort had a severe cardiovascular profile (Table II). The main causes of heart failure were ischemic heart disease (n = 5), hypertension (n = 3), and cardiac amyloidosis (n = 2), while one patient had valvular heart disease. Atrial fibrillation was present in nine patients, and three patients had dilated cardiomyopathy.

Among the 10 patients with interpretable echocardiograms, six had an impaired left ventricular ejection fraction (LVEF), and four had a preserved LVEF.

Regarding medical treatment (Table III) for heart failure at enrollment, all patients were treated with a loop diuretic, and nine of them were receiving a beta-blocker. Three patients were treated with an angiotensin-converting enzyme inhibitor or an angiotensin II receptor antagonist (ACEI/ARA2), six with a mineralocorticoid receptor antagonist, and two with a sodium-glucose cotransporter 2 inhibitor (SGLT2i).

Peritoneal dialysis regimens were tailored to each patient based on the degree of congestion, residual renal function, and clinical course.

The majority of patients were initially treated with continuous ambulatory peritoneal dialysis (CAPD), following a regimen consisting of a daily exchange using an icodextrin bag (Extraneal®) left in place for 12 hours. If fluid and sodium overload persisted, the regimen could be intensified by adding a bag of hypertonic glucose solution to increase ultrafiltration.

When volume control remained inadequate, a bimodal peritoneal dialysis strategy could be implemented on an ad hoc basis for a few days in the outpatient unit of our department. The implementation of this strategy at home was limited by the lack of specific training for private practice nurses in this modality.

Automated peritoneal dialysis (APD), which combined short dwells of 30 to 45 minutes with a hypertonic solution to promote ultrafiltration and a long dwell with icodextrin, was used in two patients.

Finally, one patient underwent a hybrid dialysis strategy combining peritoneal dialysis with a weekly hemodialysis session.

During follow-up, several complications related to peritoneal dialysis were observed. Three patients experienced an episode of peritonitis. Two patients developed a parietal leak, and two experienced a hemorrhagic episode. A loss of ultrafiltration efficiency was observed in two patients, leading to an adjustment of the peritoneal dialysis prescription. Finally, one patient developed an inguinal hernia. The complications observed during follow-up are presented in Table IV.

During a mean follow-up of 15 months (range: 1–36 months), an improvement in congestive symptoms and/or NYHA functional class was observed in nine patients. Two patients were hospitalized for cardiac decompensation after initiation of peritoneal dialysis.

At the end of the peritoneal dialysis follow-up period, the GFR ranged from 10 to 20 mL/min/1.73 m² in the majority of patients. Two patients experienced a more marked decline, with a GFR as low as 8 mL/min/1.73 m², requiring an increase in treatment frequency to four exchanges per day. Residual urine output was maintained in ten patients; one female patient became oliguric. NT-proBNP was monitored longitudinally in only two patients, with a marked decrease in both individuals. In the other patients, the lack of repeated measurements—particularly due to the sometimes short follow-up period—prevented a reliable analysis of NT-proBNP trends.

At the end of the observation period, seven patients had died. None of the deaths were considered directly attributable to peritoneal dialysis. One patient was transferred to hemodialysis, and three patients were continuing treatment with continuous ambulatory peritoneal dialysis at the end of follow-up. The main findings regarding the clinical course and outcomes of the patients are presented in Table V.

The potential benefit of peritoneal dialysis in refractory heart failure stems primarily from the control of congestion, a central component of the cardiorenal syndrome. While a decrease in cardiac output and renal perfusion has long been considered the primary mechanism underlying impaired renal function in heart failure, the role of renal venous congestion is now also recognized. Increased renal venous pressure contributes to a decrease in the glomerular filtration gradient, promoting fluid and sodium retention and thereby perpetuating the vicious cycle of cardiac dysfunction, renal dysfunction, and congestion ⁹¹⁰.

Persistent congestion despite increased doses of diuretics poses a major challenge in patients with advanced heart failure. Resistance to diuretics results from multiple mechanisms, including pharmacokinetic and pharmacodynamic alterations and a compensatory increase in distal tubular sodium reabsorption during prolonged exposure to loop diuretics ¹¹. When optimized diuretic strategies no longer achieve satisfactory decongestion, the use of an ultrafiltration strategy may then be considered in selected patients.

In this context, peritoneal dialysis has several characteristics that are particularly well suited for patients with cardio-renal syndrome. It allows for slow and gradual ultrafiltration with minimal hemodynamic impact and promotes steady removal of water and sodium. The use of icodextrin specifically during prolonged stasis enables sustained ultrafiltration and sodium removal, explaining its value for this indication ⁴. The ability to gradually adjust the number and composition of exchanges based on congestion and renal function also allows for individualized treatment.

These physiological properties form the rationale for the strategy used in our cohort, which initially relied on a prolonged daily icodextrin exchange, followed by a gradual intensification of the regimen when congestion control became inadequate.

Selecting patients likely to benefit from peritoneal dialysis is a critical step. This strategy should not be considered systematically for all patients with heart failure associated with chronic kidney disease. Rather, it should be discussed on a case-by-case basis as part of a multidisciplinary consultation.

Patients with advanced heart failure (NYHA Class III–IV) despite optimized medical therapy, repeated episodes of congestive decompensation, resistance to diuretics, or a cardio-renal syndrome with persistent fluid and sodium overload are the primary potential candidates for this approach ⁴.

At our center, this approach has led to the implementation of a cardio-nephro-geriatric care pathway designed to identify patients both in the outpatient setting and during hospitalization. In the outpatient setting, referral to this pathway is discussed specifically for patients who remain symptomatic despite optimized medical therapy and who experience repeated episodes of cardiac decompensation. For patients hospitalized for heart failure refractory to diuretics, the indication for an ultrafiltration strategy is discussed in a multidisciplinary setting.

When the clinical situation requires rapid decongestion before peritoneal dialysis is actually initiated, temporary extracorporeal ultrafiltration may be used as a transitional strategy. However, the goal remains, whenever possible, to transition to a peritoneal technique that allows for long-term management at home.

This local care pathway is summarized in Figure 1.

One of the distinctive features of our approach is the implementation of a graduated treatment strategy, aimed at adapting the intensity of peritoneal dialysis to changes in fluid overload and renal function while prioritizing, whenever possible, the continuation of home treatment.

Treatment was first initiated with CAPD consisting of a daily exchange of icodextrin (Extraneal®), with a 12-hour dwell time. In the absence of persistent congestion, this regimen was continued with regular clinical reassessment.

If congestion persisted, the regimen could be intensified by adding a bag of hypertonic glucose solution. In cases of impaired renal function requiring increased clearance and/or ultrafiltration, the number of exchanges could be increased to four per day, with adjustments to volumes and solutions. In our cohort, this intensification to four daily exchanges was necessary in two patients.

In cases of inadequate volume control, a bimodal peritoneal dialysis strategy could be proposed. This approach combines, within the peritoneal prescription, a glucose-containing crystalloid solution and an icodextrin colloid solution. It aims to optimize volume control by utilizing the complementary ultrafiltration profiles of the two types of solutions while limiting exposure to glucose ¹².

Finally, when CAPD remains inadequate despite optimization of the treatment regimen, two strategies could be considered depending on the patient’s profile: switching to APD, with an individualized regimen combining short dwells with a hypertonic solution and a long dwell with icodextrin, or a hybrid dialysis strategy combining CAPD with a weekly hemodialysis session. APD was used in two patients, without any further improvement in ultrafiltration compared to the previous CAPD regimen.

This stepwise approach allows for individualized treatment adjustment based on clinical progression, congestion control, renal function, and tolerance. It constitutes a pragmatic strategy developed at our center and, in the absence of a prospective evaluation, cannot be considered a generalizable protocol.

This therapeutic strategy is summarized in Figure 2.

Tolerance to peritoneal dialysis is an important factor to consider in this elderly population with significant comorbidities. In our cohort, the main complications observed were peritonitis (n = 3), parietal leaks (n = 2), bleeding episodes (n = 2), ultrafiltration losses (n = 2), and an inguinal hernia.

Peritonitis remains one of the main infectious complications of peritoneal dialysis. In studies focusing on PD in patients with refractory heart failure, the incidence of peritonitis varies. A systematic review of 21 studies and 673 patients reported an average annual peritonitis incidence of 14.5% ¹³. A more recent meta-analysis also highlights the significant heterogeneity in the rates reported across studies ¹⁴. These data make a direct comparison with our series difficult, especially since our dataset does not allow us to express the incidence of peritonitis as episodes per patient-year.

The recommendations of the International Society for Peritoneal Dialysis (ISPD) emphasize the importance of standardized monitoring of infectious episodes. Additionally, the ISPD specifically recommends tracking the overall peritonitis rate expressed as episodes per patient-year ¹⁵. This standardization should be incorporated into the prospective evaluation of our care pathway.

The mechanical complications observed in our cohort—notably, parietal leaks, bleeding, and inguinal hernias—are known complications of the technique. The ultrafiltration losses observed in two patients led to an adjustment of the treatment regimen. Regular assessment of ultrafiltration efficacy is particularly important in this population, since the primary goal of PD is precisely to maintain satisfactory volume control.

Despite these adverse events, no deaths were directly attributed to peritoneal dialysis in our series. Advanced age, severe heart failure, chronic kidney disease, and a high burden of comorbidities must be taken into account when interpreting the long-term outcomes of these patients.

Our study has several limitations. It is a retrospective, descriptive, single-center study involving a small sample size of 11 patients, with no control group. These characteristics do not allow us to establish a causal relationship between the initiation of peritoneal dialysis and the observed clinical outcomes.

Furthermore, although a small number of hospitalizations for cardiac decompensation were observed after the initiation of peritoneal dialysis, we do not have a standardized comparison of the number of hospital days or the hospitalization rate before and after the start of the treatment. Clinical improvement was based primarily on changes in signs of congestion and NYHA functional class, without a standardized assessment of quality of life.

The retrospective data collection also did not allow for the evaluation of all complications using standardized indicators. In particular, episodes of peritonitis were not expressed as episodes per patient-year, limiting comparison with published data and international guidelines. Furthermore, urine output was not quantified in a standardized manner, and NT-proBNP was monitored longitudinally in only two patients; these laboratory data can therefore be interpreted only descriptively.

The high mortality rate observed during follow-up must be interpreted in the context of a particularly frail population, characterized by advanced age, severe heart failure, advanced chronic kidney disease, and a high burden of comorbidities. No deaths were considered directly attributable to peritoneal dialysis.

Despite these limitations, this single-center study demonstrates the feasibility of structured peritoneal dialysis management in patients with refractory heart failure and cardio-renal syndrome. It also led our team to formalize a cardio-renal-geriatric care pathway and a graduated treatment strategy that allows for the adaptation of ultrafiltration parameters to patients’ clinical and renal progression.

A prospective evaluation of this care pathway now appears desirable. It could include a standardized comparison of hospitalizations before and after the initiation of peritoneal dialysis; changes in NYHA functional class, renal function, and signs of congestion; a validated measure of quality of life; and standardized data collection on complications. A multicenter study would also allow for an assessment of the reproducibility of this strategy and help clarify the role of peritoneal dialysis in the current management of refractory heart failure.

IG: study design, data collection and analysis, manuscript drafting. NF and VC: patient care and data collection. MA and ZB: supervision and critical review. All authors read and approved the final version of the manuscript.

This study was conducted in accordance with the principles of the Declaration of Helsinki. Patient identities were anonymized. The retrospective data were analyzed in aggregate form, without the possibility of individual identification. In accordance with the local regulations applicable to this retrospective study using anonymized data, review by an ethics committee was not required.

These findings were described retrospectively based on the hospital records of patients following their discharge or death. The data were processed in a strictly anonymized manner. No specific written consent for this retrospective study was obtained.

The authors declare that they received no direct or indirect funding for this study.

The authors declare that they have no conflicts of interest related to this work.

This study is based on data from patient records maintained by the institution. Local legislation does not permit the disclosure of individual records.

Originality of Figures. Figures 1 and 2 are original creations by the authors and were not reproduced from another article.

A generative artificial intelligence tool was used as a writing aid for linguistic reformulation, structuring the manuscript, and improving its readability. The scientific content, data analysis, interpretation of the data, verification of references, and final validation of the manuscript are the sole responsibility of the authors.

Ismahane Guenifi: 0009 0006 8287 866X

Nathalie FRANCH :0009 0003 5396 6861

Virginie CORNU: 0009 0004 8426 7438

Amirou Mustapha: 0009 0008 7500 116X

Zaid BENABID: 0009 0004 4107 6287