Introduction

The prevalence of chronic kidney disease (CKD) varies across different regions and is influenced by a range of factors, including population demographics, access to healthcare, and lifestyle choices [1]. Recognized as a global health concern by organizations such as the World Health Organization (WHO) and the International Society of Nephrology, CKD has shown a consistent increase in prevalence in many parts of the world. Globally, CKD affects more than 10% of the population to varying degrees, but the specific prevalence rates can differ significantly from one country or region to another [2]. Notably, low- and middle-income countries may experience higher prevalence rates due to challenges like limited access to healthcare services, the presence of infectious diseases, and environmental factors [3]. In the United States alone, CKD is a significant public health issue. According to estimates provided by the National Kidney Foundation and the Centers for Disease Control and Prevention (CDC), approximately 37 million adults in the U.S. were living with CKD, representing around 15% of the adult population. Ageing, diabetes, obesity, and hypertension are considered the main causes of CKD, and place a considerable financial burden in most advanced healthcare systems [2, 3]. Moreover, the number of deaths due to poor access to renal replacement therapy in developing countries remains unacceptably high [4].

Nephron loss triggers hypertrophy and hyperfiltration in remaining nephrons, and even with successful management of diabetes and hypertension, CKD often advances to end-stage renal disease (ESRD) affecting an estimated 4.902–7.083 million individuals worldwide. CKD has become a significant global health concern, with a prevalence of approximately 13.4%, impacting morbidity and mortality on a global scale. Its growth is primarily driven by factors such as diabetes, hypertension, obesity, and an aging population, though specific causes may vary regionally, including infections and environmental factors [4, 5]. However, significant challenges exist in clinical practice regarding the capacity to predict the temporal progression of CKD, preventing the institution of suitable preventive and management strategies. CKD is characterized by chronic low-grade inflammation, endothelial dysfunction, and platelet activation [6, 7]. The inflammatory state in CKD is associated with the release of cytokines and the production and activation of adhesion molecules that lead to T-cell adhesion and migration into the interstitium, with consequent stimulation of pro-fibrotic factors. Inflammation in CKD increases the risk of cardiovascular events, all-cause mortality, and its progression by contributing to the development of vascular calcifications and endothelial dysfunction [6]. Although several inflammatory markers have been investigated in CKD, such as CRP, erythrocyte sedimentation rate, interleukin (IL)-6, IL-8, IL-12, tumor necrosis factor (TNF)-alpha, and IL-33, their ability to predict disease progression is unclear. In recent years, routinely measured blood cell indexes in several diseases [8,9,10,11,12] as the neutrophil/lymphocyte ratio (NLR), platelet/lymphocyte ratio (PLR), monocyte/lymphocyte ratio (MLR), derived NLR (dNLR), systemic inflammation index (SII), and aggregate index of systemic inflammation (AISI) have been investigated as markers of inflammation in renal disease [13] and acute kidney injury after major abdominal surgery [14, 15]. However, their clinical significance has primarily been investigated in cancer and cardiovascular disease [16,17,18,19,20]. In this study we investigated the ability of these blood cell indexes to predict (a) the further decline in renal function in CKD and (b) the transition of CKD stage 3–4 to ESRD.

Methods

We conducted a retrospective study of CKD stage 3–4 patients (defined as an e-GFR between 30–60 and 15–30 ml/min/1.73 m2, respectively); including patients with comorbidities such as type 2 diabetes, hypertension, coronary artery disease, and the elderly attending the nephrology outpatient clinic at the Hue Central Hospital in Viet Nam from October 2015 to October 2018. Sample size: the required sample sizes for estimating specificity concerning the pre-determined value of specificity ascertained by previously published data [21].

$$ n = \frac{{Z_{{\frac{\alpha }{2}}}^{2} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{P} (1 - \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{P} )}}{{d^{2} }} $$

P = 0.69, is a pre-determined value of specificity from published data of Ismail Kocyigit et al. [22]; α = 0.05; marginal error d = 5%. Sample size n = 329.

Exclusion criteria were follow-up time <1 year, eGFR > 60 ml/min/1.73 m2, acute inflammation (defined by a WBC count>15,000/mm3 and/or a C-reactive protein (CRP) concentration > 50 mg/dL), hematological diseases, other chronic inflammatory processes (such as rheumatoid arthritis, ankylosing spondylitis, and gout), and treatment with immunosuppressive therapies. We also excluded CKD patients exhibiting changes in eGFR to values compatible with stages less than 3 or 4 during follow-up, leaving 377 patients with CKD stages 3–4 suitable for analysis.

Patients underwent clinical and biochemical profile assessments to evaluate CKD progression, including CRP, urea, and creatinine. NLR (neutrophil to lymphocyte ratio), PLR (platelet to lymphocyte ratio), MLR (monocyte to lymphocyte ratio), dNLR (derived NLR), RDW (Red Cell Distribution With), AISI (aggregate index of systemic inflammation, calculated by multiplying the counts of neutrophils, monocytes, and platelets and then dividing the product by the lymphocyte count), and SII (systemic inflammation index, calculated by (neutrophil counts × platelet counts)/lymphocyte counts) were assessed at baseline and then serially during outpatient visits. Smoking status, cause of CKD, type of antihypertensive drugs, and laboratory tests were also recorded during the first visit. The time point of the last outpatient visit within the 3 years was defined as the last assessment. ESRD is defined as an irreversible decline in kidney function, which requires dialysis or transplantation to improve the glomerular filtration rate (GFR) [23]. For patients who transitioned to ESRD before the last outpatient visit, the last assessment was defined as the time point of transition. Therefore, the time to transition to ESRD (months) was the time interval between the first visit and the time point of transition to ESRD.

Estimated glomerular filtration rate (eGFR) was calculated using the MDRD equation [24]. According to KDIGO (International Kidney Disease: Improving Global Outcomes) 2012 guideline classification CKD stage 3 and 4 were defined as an eGFR between 30–60 and 15–30 ml/min/1.73 m2, respectively [23]. eGFR decline, the primary endpoint of the study, was defined as rapid or slow if the eGFR reduction was either >= 5 mL/min/year or < 5 mL/min/year, respectively, according to the following formula: annual progression rate (mL/min/year) = baseline eGFR-last eGFR (ml/min/1.73 m2) ×12 months/number of months [25].

Transition to ESRD, or initiation of renal replacement therapy, was the secondary endpoint to categorize the patients. Patients with an eGFR less than 15 ml/min/1.73 m2 are considered to be in an advanced state of renal failure and are often in need of dialysis or kidney transplantation, whereas Patients with an eGFR between 15 and 60 ml/min/1.73 m2 may have a variety of renal conditions but are not yet considered to be in transition.

Statistical Analysis

We expressed categorical data as percentage and number. Non-normally distributed variables were presented as median (range) and normally distributed variables as mean ± SD, as appropriate. Statistical significance was concluded when p-value < 0.05. We used the Chi-square test to assess between-group comparisons for nominal variables and the Independent Samples T-test to compare means between groups. Spearman’s rank correlation was used to determine correlations between paired variables. Variables showing associations with p-values <0.05 were included in logistic regression to determine their ability to independently predict the decline in eGFR and the transition to ESRD. We measured the ROC curve to determine the sensitivity, specificity, AUC, and cutoff values of predictive factors.

Ethical Standards

Compliance with ethical standards and approved by Ethic Council in biomedical research of Hue University of Medicine and Pharmacy, Vietnam. Date of Approval: 6 January, 2018.

Results

A total of 377 patients were included in the study (Table 1). There were no statistically significant differences in age, gender, BMI, smoking status, baseline serum creatinine, eGFR, type of antihypertensive drugs, blood cells (M, RDW, and PDW), and CRP concentrations between CKD patients with slow vs. rapid progression. There were significant differences in the duration of follow-up, causes of CKD, white blood cells (W), ANC(N), platelets (P), mean platelet volume (MPV), NLR, PLR, MLR, dNLR, SII, and AISI.

Table 1 Distribution of baseline characteristics according to eGFR decline
Full size table

We run a correlation analysis between pre-specified variables and CKD progression rate. Variable showing associations with p-values < 0.05 were included in logistic regression to determine the independent predictive ability for the decline in eGFR and the transition to ESRD.

In univariate logistic regression analysis, the duration of follow-up, baseline eGFR, baseline WBC, baseline N, baseline L, baseline P, baseline MPV, baseline NLR, baseline PLR, baseline dNLR, baseline MLR, baseline SII, and baseline AISI were significantly associated with a rapid decline in eGFR. However, in multivariate analysis, only baseline N, baseline MPV, baseline NLR, baseline PLR, baseline dNLR, and baseline SII remained independent predictors of the rapid decline in eGFR (Table 2).

Table 2 Logistic regression analysis of rapid decline in eGFR
Full size table

The ability of N, MPV, NLR, PLR, dNLR, and SII to predict a rapid decline in eGFR was further assessed by ROC analysis (Fig. 1). The sensitivity and specificity of NLR and dNLR were higher than the other indexes, 88.6 and 81.7%; 85.2 and 75.8%, respectively, using a cutoff level of 3.36, and 2.45, respectively. AUC areas of NLR and dNLR were 0.877 (95% CI 0.837–0.918, p < 0.001) and 0.849 (95% CI 0.805–0.892, p < 0.001), respectively. The detailed data of other indexes are described in Table 3.

Fig. 1
figure 1

ROC analysis of high N, MPV, NLR, PLR, dNLR, SII as predictors of rapid decline in eGFR, N—neutrophil. MPV—mean platelet volume. NLR—neutrophil/lymphocyte ratio. PLR—platelet/lymphocyte ratio. dNLR—derived NLR. SII—systemic inflammation index. ROC—Receiver Operating Characteristics

Full size image
Table 3 ROC analysis of blood cell indexes
Full size table

After a 3-year follow-up, 19.1% (72/377) CKD stage 3–4 transitioned to ESRD. There were no statistically significant differences in age, BMI, smoking status, duration of follow-up, type of antihypertensive drugs, causes of CKD, blood cells (W, N, M, P, MPV, RDW, and PDW), PLR and CPR concentrations between CKD patients with or without ESRD (Table 4). However, there were statistically significant differences in sex, baseline creatinine, eGFR, L, NLR, MLR, dNLR, SII, and AISI. Our results showed that patients transitioning to ESRD were also those with rapid eGFR deterioration (Fig. 2).

Table 4 The distribution of baseline characteristics according to transition to ESRD
Full size table
Fig. 2
figure 2

ROC analysis predictors of L transition to ESRD. L—lymphocyte. ROC—Receiver Operating Characteristics. ESRD—end stage renal disease

Full size image

In univariate analysis, gender, baseline eGFR, kidney progression rate, baseline L, baseline P, baseline NLR, baseline NLR grade, baseline dNLR, baseline MLR, baseline SII, and baseline AISI were significantly associated with transition to ESRD. However, in multivariate analysis, only gender, baseline creatinine, kidney progression rate, and L had independent predictive roles in the transition to ESRD with p<0.001, p<0.001, p = 0.001 and p = 0.008, respectively) (Table 5).

Table 5 Logistic regression analysis of transition to ESRD
Full size table

Discussion

This study showed that N, MPV, NLR, PLR, dNLR, and SII significantly predicted a rapid decline in eGFR, with both the NLR and the dNLR exhibiting the highest predictive capacity based on AUC values. The decline in eGFR was faster in patients with high NLR (cutoff level > 3.36) and dNLR (cutoff level > 2.45) group, results similar to those of Kocyigit's et al. [22]. Neutrophils release chemotactic substances that promote neutrophil migration to the kidney, worsening glomerular injury. These changes contribute to the fast progression of CKD. NLR provide information regarding inflammation patients in CKD [26], predicts the severity and prognosis of several pathological conditions chronic [27, 28], NLR and dNLR predict survival in breast cancer [29], urological [30], colorectal, and gastrointestinal tumors [31], high pre-treatment NLR (≥4) predicted a poor prognosis in lung cancer [32]. NLR could be used to predict cardiovascular endpoints in CKD patients in moderate to severe conditions associated with endothelial dysfunction, predicts all-cause mortality in geriatric patients with CKD [33]. However, its predictive role in the progression of CKD has not been comprehensively investigated. Our study demonstrates that dNLR independently predicts rapid CKD progression. In CKD patients, inflammation is associated with pro-coagulant and platelet-activating factors contributing to endothelial dysfunction [34]. PLR was shown to be superior to NLR in terms of inflammation in ESRD patients [35] and could independently predict all-cause mortality in hemodialysis patients [36]. SII has been shown to be a useful indicator [37]. In our study, both predicted a rapid decline in eGFR with sensitivity and specificity were inferior to that of NLR and dNLR. Also, CRP has a significant role in the rapid decline in eGFR supporting the potential advantage of using NLR and dNLR as predictors of eGFR decline in CKD stage 3–4. The ability of blood count cells to predict the transition to ESRD also was assessed by logistic regression and ROC analysis. We did not identify any useful blood cell index to predict the transition to ESRD. Several studies are in agreement with our study, which has shown that the percentage of patients transitioning into ESRD was not different between high and low-baseline NLR groups [38] and revealed that NLR is associated with the risk of ESRD in Chinese patients with stage 4 CKD [39, 40]. More research is warranted to clarify the role of the NLR and other inflammatory indexes in the progression of CKD. The main limitation of our study is the retrospective design and the lack of all patient data on the inflammation marker CRP (Table 6).

Table 6 L towards transition to ESRD
Full size table

Conclusion

In conclusion, our results suggest that the NLR and the dNLR have the highest predictive capacity of a rapid decline in eGFR and may represent useful markers for monitoring in CKD stage 3–4 in clinical practice because they are simple, inexpensive and easy to access.