Prognostic implications of predialysis patients’ symptoms in peritoneal dialysis patients

Abstract Background As kidney disease progresses, patients often experience a variety of symptoms. There are very few studies reporting spectrum of predialysis patients’ symptoms in peritoneal dialysis (PD) patients. Furthermore, the clinical significance of predialysis patients’ symptoms for PD patients’ prognosis remains unknown. Methods In this retrospective cohort study, patients who started PD during 1 January 2006 to 31 January 2018 were included. Patients’ predialysis symptoms and clinical parameters were obtained. Both the short- and long-term patients’ outcome were investigated by Cox regression and Kaplan–Meier’s survival analysis to identify the relationship between clinical symptoms and patients' mortality on PD. Results A total of 898 incident PD patients were included. The anorexia (58%) was the most common predialysis symptom in the present cohort, followed by insomnia (32.7%), fatigue (27.6%), syndromes of heart failure (27.6%), and nausea (20.5%). The only symptom significantly associated with both six-months and 12-months mortality on PD was nausea (HR 2.359, 95% CI 1.377–4.040, p=.002 and HR 1.791, 95% CI 1.176–2.729, p=.007, respectively). But in the long-term, anorexia (HR 1.392, 95% CI 1.070–1.811, p=.014) was the only symptom significantly associated with patient's all-cause mortality after adjusting for other confounding factors. Conclusions Our study demonstrated that nausea and anorexia were the most important predialysis symptoms, which was associated with patients’ short- and long-term mortality on PD treatment, respectively. The results indicated that predialysis evaluation and management of symptoms of nausea and anorexia may be a possible way to improve patients’ outcome on PD.


Background
As kidney disease progresses, patients often experience a variety of symptoms [1,2], such as anorexia, fatigue, cognitive impairment, depressive symptoms, pruritus, and sleep disturbances [3]. As showed by the IDEAL study [4] that 76% of the patients allotted to late dialysis had to start dialysis before the target of estimated glomerular filtration rate (eGFR) <7 ml/min due to uremic symptoms. These further supported the notion that the predialysis symptoms are crucial in clinic setting for clinician to decide the right timing to begin dialysis [4,5].
To identify the most meaningful symptoms and accordingly, to decide the right timing of dialysis initiation is still a challenge for the nephrologists. However, there are very few studies reporting spectrum of predialysis patients' symptoms in peritoneal dialysis (PD) patients and little is known about the burden of symptoms of predialysis patients especially PD patients [6]. In addition, unlike many predialysis indictors linking to the patients outcomes, such as GFR, serum albumin [7,8], calcium [9,10], body mass index [11], and comorbidities [8,12], the potential predictive value of patients' predialysis symptoms have not been fully elucidated.
In the present study, we investigated spectrum of predialysis symptoms and tried to identify the most important predialysis symptoms linking to patients' short-term and long-term outcomes.

Methods
In this retrospective cohort study, all incident patients who started PD therapy from 1 January 2006 to 31 January 2018 from Peking University Third Hospital were included. Patients were excluded if they had received any form of dialysis for more than 1 month, or if they had received previous kidney transplantation. Patients who started PD in other center were also excluded for we were unable to get their detailed predialysis symptom information. All the patients were followed by PD clinic of Peking University Third Hospital. Patients were followed until death, cessation of PD or end of study as of 15 November 2018. Patients' medical charts were reviewed to extract the information. In our center, predialysis patients need to be hospitalized to department of nephrology for performing PD catheter implantation. Therefore, patients' detailed predialysis clinic symptoms, predialysis manifestations, clinical history, demographics, lab assay were recorded when admitted. During the study period, patients' hospitalization medical charts were carefully reviewed to obtain the predialysis symptom data.
The study protocol was approved by the Peking University Third Hospital Medical Science Research Ethics Committee, IRB00006761-M2019290. Review of patients' records and use of data by this study were permitted by Peking University Third Hospital Medical Science Research Ethics Committee. Patients' eGFR was estimated by Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation [13]. Hypocalcemia were defined as fasting plasma calcium level lower than 2.0 mmol/l and hyperphosphatemia were defined as fasting plasma phosphorus level higher than 1.78 mmol/l.
The primary outcomes examined were all-cause mortality on PD (censored for loss to follow-up, renal transplantation, recovery of renal function, transferred to hemodialysis (HD) due to technical failure, transfer to other dialysis centers and end of study).

Statistical analysis
Results were expressed as frequencies and percentages for categorical variables, mean ± standard deviation for continuous normally distributed variables, and median (interquartile range, IQR) for continuous variables that were not normally distributed. Comparison for patients in different groups was performed using chi-square tests, two-tailed unpaired t-tests or Mann-Whitney's tests, depending on data distribution. Mortality risks (both 0.5, 1 year and long-term mortality) were analyzed by the Kaplan-Meier and multivariate Cox's proportional hazard model in which all the significant variables (p<.1) from the univariate analysis were included. Statistical analysis was performed using IBM SPSS software, version 22.0 (Armonk, NY). p Values less than .05 were considered statistically significant.

Results
Clinical and laboratory characteristics at the initiation of PD are depicted in Table 1. Patients' uremic manifestations and syndromes at the initiation of PD are shown in Figure 1. The anorexia (58%) was the most common symptom in the present cohort, followed by insomnia (32.7%), fatigue (27.6%), syndromes of heart failure (27.6%), and nausea (20.5%), respectively.

Patients' symptoms and short-term outcome on PD
Both the six-months and 12-month mortality were evaluated in the present study for patients' short-term outcome. Univariate Cox regression analysis showed that the only symptom significantly associated with both six-months and 12-months mortality on PD was nausea (HR 2.359, 95% CI 1.377-4.040, p¼.002 and HR 1.791, 95% CI 1.176-2.729, p¼.007, respectively) ( Table 2).
Both the six-months and 12-month survival time of patients with symptom of nausea were significantly lower compared to patients without (log rank test, both p<.001) (Figure 2 Patients' symptoms and long-term mortality on PD Univariate Cox regression analysis showed that the predialysis symptoms significantly associated with patients' all-cause mortality were dyspnea (p¼.001), edema (p¼.016), nausea (p¼.033), anorexia (p<.001), and insomnia (p¼.008) ( Table 3). Interestingly, multivariate Cox regression analysis showed that anorexia (HR 1.392, 95% CI 1.070-1.811, p¼.014) was the only symptom significantly associated with patients' all-cause mortality after adjusting for other confounding factors ( Table 3). The survival time of patients with symptom of anorexia was significantly lower compared to patients who without anorexia (log rank test, p<.001) (Figure 3).

Discussion
In this retrospective cohort study, we found that the most common predialysis symptom was anorexia. Most importantly, we demonstrated that predialysis anorexia   was a strong and significant predictor for PD patients' long-term mortality. In addition, presenting with symptom of nausea was associated with high short-term PD mortality after adjusting for all the other confounding factors. Anorexia (58%) was the most common symptom in the present cohort, followed by insomnia (32.7%), fatigue (27.6%), syndromes of heart failure (27.6%), and nausea (20.5%). These results are different from the previous study that the most common symptoms were fatigue (44%), nausea (24%), and anorexia (22%) at the dialysis initiation [6]. Nausea and anorexia are major uremic symptoms and a frequent indication for starting dialysis [14]. The higher present of anorexia and nauseas may reflect a relatively late start of dialysis in the present predialysis patients. Interestingly, we found that nausea is a more powerful predictor for patients' short-term (6 and 12 months) mortality than anorexia after adjusting other confounding factors, which indicated that for patients with symptoms of nausea, more intensive treatment may be needed to ameliorate the symptom and thus to improve patients shortterm outcome.
Previous study in PD patients showed that selfreported appetite (anorexia) was a predictor of clinical characteristics and outcome for patients receiving PD [15] and HD patients [16]. Different from this study which anorexia were evaluated on PD at least 12 weeks or on HD for at least 120 days [16], symptom of anorexia was evaluated just before PD initiation in the present study. The symptoms of anorexia in the present study reflect predialysis patients' status rather than after dialysis, which may indicate that in advance chronic kidney disease (CKD), the symptoms of anorexia could be an indicator of long-term PD outcomes in spite that standard dialysis prescription was delivered. This result is consistent with the study from Thiane et al. who showed that in CKD5 non-dialysis patients, there was a significant increase in mortality risk in PD patients with poor appetite [16]. The underlining causes for anorexia could be accumulation of unidentified anorexigenic compounds, inflammatory cytokines, and alterations in appetite regulation [17,18]. This further contributes to the high prevalence of protein-energy wasting (PEW) and could have a direct adverse effect on morbidity and mortality of these patients [13].
Our study suggest that a single evaluation of predialysis symptoms of anorexia can help identify patients with high risk after PD, which suggested performing appetite evaluation periodically is crucial for patients with advance CKD and may help to decide the right timing for commencing maintenance dialysis treatment. Therefore, more intensive intervention and management either before or after PD treatment would be necessary for these patients. Further study is needed to elucidate the possible effect of appetite improvement by PD as well as its effect on clinical outcomes.
Our study has some limitations. First, patients included in this study were recruited from a single tertiary academic hospital in China, thereby raising the possibility of ascertainment bias. Second, due to the retrospective design, some detailed symptoms severity could not be obtained. Finally, although we attempted to adjust for a range of demographic, clinical, and laboratory characteristics, some detailed comorbidity index (e.g., Charlson comorbidity index and nutritional status assessment like subjective global nutrition assessment, SGA) was not collected, thus residual confounding remains possible.

Conclusions
Our study demonstrated that predialysis nausea and anorexia were the most important predialysis symptoms, which was associated with patients' short-term and long-term mortality on PD treatment, respectively. These results indicated that predialysis evaluation and management of symptoms of nausea and anorexia may be a possible way to improve patients' outcome on PD.