The prognostic value of monocyte-to-lymphocyte ratio in peritoneal dialysis patients

Chronic kidney disease (CKD) is an important public health problem cause of high incidence, high mortality and heavy economic and social burden. Peritoneal dialysis (PD) is a fundamental modality of renal replacement treatment for end-stage renal disease (ESRD) patients. Previous study reported the estimated systemic inflammation prevalence ranged from 12 to 65% in PD patients, depending on the different cut-off values of C-reactive protein (CRP) [1]. A variety of complex factors contribute to inflammation in PD patients, including accumulation of uremic toxins, oxidative stress, increased endotoxin level, volume overload, decreased clearance of proinflammatory cytokines, PD catheter and so on [1, 2]. Protein-energy malnutrition (PEM) and inflammation are highly prevalent and often occur concomitantly in patients with ESRD. Hence, these two conditions have been regarded together as ‘malnutrition-inflammation complex syndrome’ (MICS) [3, 4]. As is known to all, MICS is strongly correlated with higher incidence of cardiovascular disease (CVD) and mortality in dialysis patients [4‐6].

Traditional inflammatory mediators, including tumor necrosis factor (TNF), CRP, and interleukin (IL) have been studied in dialysis patients. Due to high test cost and complex technological process, they have not been widely implemented in clinical practice. The monocyte-to-lymphocyte ratio (MLR), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) are reproducible, inexpensive markers of inflammatory response. They can be easily calculated from routine blood test under simple laboratory conditions [7‐9]. Recent studies have demonstrated that NLR and PLR were strongly associated with poor outcome in PD patients [8, 9]. However, few researches were focused on the prognostic value of the MLR in dialysis patients. The MLR have been confirmed as a prognostic predictor in malignancies [10, 11]. Xiang F et al. first found that higher MLR was an independent risk factor of all-cause and CVD mortality in hemodialysis (HD) patients and overwhelmed NLR [12]. Subsequently, a multicenter retrospective cohort study suggested that the highest MLR tertile was independently correlated with increased CVD mortality in PD patients [7]. At present, no studies have reported the correlation between MLR and all-cause mortality in PD patients. Therefore, this study was aimed to investigate the prognostic value of MLR for all-cause mortality and new-onset CVD events in PD patients.

The main finding of this study was that baseline high MLR level (> 0.2168) was an independent predictor for all-cause mortality and CVD events, even after adjustment of traditional risk factors, including age, diabetes mellitus, CVD history, smoking, hyperlipidemia.

The PD favored policy in China has made contribution to increased utilization of PD. In recent decades, there was a significant reduction of mortality risk in PD patients [14]. Nowadays, most studies reported similar survival rate in PD and in-center HD [14‐16]. Despite all this, the mortality of PD is still higher than age-matched general population [2]. In our study, the 5-year survival rate was lower in high MLR group than in low MLR group (70.5% vs. 89.5%). We first demonstrated that high MLR was significantly and independently associated with a HR for all-cause mortality of 2.518 (95%CI = 1.020–6.214, P = 0.045) after adjusting recognized risk factors. It is worth noting that inflammation plays an important role in the pathogenesis and progression of CKD [17]. Enhancement of inflammation or oxidative stress, generation of proinflammatory cytokines and immune system change can result in nephron loss and kidney injury [18]. Naicker SD et al. compared leucocyte and monocyte populations in patients with CKD stage 1–5 and healthy controls [19]. They confirmed dysregulation of neutrophil and monocyte subset in CKD and found a distinct subpopulation of intermediate monocytes was significantly associated with estimated glomerular filtration rate (eGFR) [19]. A large longitudinal cohort study enrolled 11,280 participants and detected increased MLR was associated with new-onset CKD [20]. The predictive value of MLR for all-cause mortality has been substantiated in HD patients [12]. Our results suggested this conclusion could be applicable to PD patients.

CVD is a severe complication and leading cause of death, occupying nearly 60% of all-cause mortality in PD [7]. The baseline MLR has been found to predict CVD mortality in dialysis patients in previous studies [7, 12]. A retrospective study with 543 patients undergoing coronary angiography observed that MLR was an independent risk factor of coronary artery disease and had better predictive value of lesion severity than NLR [21]. Increased lymphocytes apoptosis causes lymphocytopenia that can reflect a poorly regulated immune response and impaired coronary microcirculation [22]. Monocytes involve in the formation and rupture of atherosclerotic plaque by recruiting to the artery wall, differentiating into macrophages and activating the production of matrix metalloproteinases, pro-inflammatory cytokines secretion, and reactive oxidative species [23]. Both the components of MLR were confirmed to be independent predictors of CVD [24]. Therefore, we speculate that MLR can be recognized as a better indicator than monocytes or lymphocytes separately. In our PD center, more than half of the patients occurred CVD and high MLR was associated with increased incidence of CVD (HR = 1.815, 95%CI = 1.157–2.849, P = 0.010).

Consistent with previous study, elevated CRP was found to predict all-cause death in PD [25]. We detected every 1 mg/L increase in CRP was independently related to higher all-cause mortality (HR = 1.016, 95%CI = 1.002–1.030, P = 0.021). In addition, the MLR was superior to CRP in predicting mortality according to ROC curves. We also detected increased CRP was associated with CVD, but no statistical significance was found in multivariate Cox analysis. Taken together, our study provided convinced evidences that MLR as a novel marker had a better kinetic pattern than CRP.

Anemia is not only a common complication of CKD, but also a risk factor for CVD and mortality [26]. It was an interesting finding that hemoglobin exhibited an incremental association with the mortality and CVD risk in univariate Cox analysis. Noteworthily, we only investigated the hemoglobin level before PD, further studies should integrate the treatment of hemoglobin and its variability. Dysregulation of lipid metabolism participates in CKD-associated inflammation and oxidative stress which promote CVD [27]. Experimental study demonstrated that reduced cholesterol and LDL-C could limit CVD and mortality in the general population [28]. Our study revealed that increased LDL-C was independently related to all-cause mortality (HR = 1.620, 95%CI = 1.157–2.269, P = 0.005), while elevated cholesterol was an independent risk factor for CVD (HR = 1.325, 95%CI = 1.146–1.531, P < 0.001). A prospective cohort study with 1616 PD patients showed that the highest LDL-C to HDL-C ratio tertile was significantly associated with all-cause and CVD mortality [29]. Another study analyzed data from Taiwan Renal Registry Data System and indicated that risk of death rose with a higher cholesterol level [30]. Stepanova N et al. conducted a cross-sectional pilot study and first demonstrated there was really a correlation between dyslipidemia and intraperitoneal inflammation in PD patients [31]. It seemed that lipid profiles had prognostic value in risk prediction. However, lipid-lowering therapy should be considered to determine the significance of dyslipidemia in future researches.

There are still several limitations in this study. First, because this is a single-center retrospective study conducted in China with a relatively small sample, its potential selection bias and center-specific effect may limit external validity and statistical power. A second issue is lack of comparison with traditional inflammation biomarkers, such as IL-6, TNF-α. Although prognostic values of them have been proved in numerous studies [2, 32], they are not measured routinely because of expensive and inconvenient assays. Besides, the optimal cut-off value of the MLR needs further validation in other ethnic cohorts. Because of competing risk such as HD transfer, kidney transplantation, the incidence of all-cause mortality and CVD risk may potentially be inflated by Kaplan–Meier method. Finally, we only investigated the effects of baseline variables, the influences of longitudinal changes have been ignored. Future studies should pay more attention to time-varying MLR and determine the causal relationship between MLR and clinical outcomes in PD patients.

This study first demonstrated that high MLR was significantly and independently correlated with all-cause mortality and CVD events in PD patients despite adjustments for multiple confounders. We suggested a possible potential predictive value for MLR was superior to monocytes, lymphocytes, and CRP. The MLR is an inexpensive and straightforward indicator to reflect systemic inflammation status. Future researches integrating baseline and changes over time are needed to identify the clinical significance of MLR and help clinicians improve PD management.