Neutrophil-to-lymphocyte ratio as a predictor for early mortality in older patients requiring hemodialysis; insights for hemodialysis access planning

Abstract

Introduction

The 2019 Kidney Disease Outcome Quality Initiative guidelines emphasize the importance of selecting dialysis based on the life expectancy of the patient. However, it is difficult to predict the life expectancy of a patient during arteriovenous fistula creation. We investigated whether neutrophil-to-lymphocyte, monocyte-to-lymphocyte, and platelet-to-lymphocyte ratios measured before dialysis could predict mortality.

Materials and methods

Between January 2016 and December 2020, we retrospectively analyzed electronic medical records of 448 patients aged ≥ 70 years undergoing first-time arteriovenous access surgery at three tertiary care hospitals, all of whom had not received prior dialysis treatment. Only patients undergoing blood tests on the day before surgery were included in the analysis. Patients who died within one year after surgery were included in the non-survival group, while those who died after one year were included in the survival group.

Results

Patients were categorized into non-survival (n = 52) and survival (n = 396) groups. Multivariate analysis for one-year mortality revealed that the preoperative neutrophil-to-lymphocyte ratio demonstrated a 1.15 hazard ratio ( p < 0.001 ). Also, cancer (HR 2.50, p = 0.02) and peripheral arterial disease (HR 4.62, p < 0.001) were risk factor for one-year mortality. The preoperative platelet-to-lymphocyte and monocyte-to-lymphocyte ratios were not identified as one-year mortality risk factors. In the total mortality multivariate analysis, monocyte-to-lymphocyte ratios were one of the risk factors (HR 2.74, p < 0.007).

Conclusion

The neutrophil-to-lymphocyte ratio was a risk factor associated with one-year mortality in patients aged ≥ 70 years. However, further research is required to determine whether these can be used for predictive purposes.

Introduction

The global number of patients receiving renal replacement therapy exceeds 2.5 million and is projected to double by 2030 to 5.4 million, mirroring the increasing prevalence of chronic kidney disease [1]. Hemodialysis is the primary method for renal replacement therapy and depends on vascular access for maintenance [2]. The challenge lies in selecting optimal access, balancing reliability, achieving freedom from complications, and determining suitability for individual patients. Renal replacement therapy options include arteriovenous fistula, arteriovenous graft, or central venous catheter. The Kidney Disease Outcomes Quality Initiative Vascular Access Guidelines has evolved since 1997 and provide best practice guidance for managing hemodialysis access [3]. Recognizing the importance of maintaining adequate hemodialysis access is essential, as hemodialysis access dysfunction in patients requiring hemodialysis is associated with increased mortality and morbidity risks [4].

The 2019 Kidney Disease Outcome Quality Initiative guidelines herald a paradigm shift that highlights the pivotal role of hemodialysis access through a patient-centric prism. The updated guidelines depart from the “fistula first” approach by introducing the “End Stage Kidney Disease Life-Plan,” an individualized strategy extending beyond vascular access to encompass a comprehensive plan for lifelong dialysis modalities [3]. This shift from a singular hemodialysis access focus to a holistic patient-centered dialysis access strategy aims to achieve reliable, functioning, and complication-free access while preserving future options aligned with the enduring needs of the patient.

Efforts to enhance hemodialysis access planning in patients with chronic kidney disease have focused on investigating predictors associated with mortality. Yap et al. identified older age, peripheral artery disease, cardiomegaly, higher white blood cell count, and early arteriovenous fistula failure as overall mortality predictors in patients requiring hemodialysis, emphasizing the significance of inflammation [5]. Planning hemodialysis access based on patient condition, overall health, and life expectancy is crucial, particularly with the development of simple predictors reflecting mortality. These predictors provide substantial assistance in strategic hemodialysis access planning [3]. This study aims to explore serologic markers, including the neutrophil-to-lymphocyte ratio, as early mortality predictors in patients aged ≥ 70 years with suboptimal access patency and requiring hemodialysis. These findings will contribute to the ongoing endeavors to improve hemodialysis access planning.

Materials and methods

The three institutions involved in this study received institutional review board approval to conduct the research (Korea University ANAM Hospital 2023AN0159, Korea University GURO Hospital 2023GR0523, and Korea University ANSAN Hospital 2023AS0099).

Between January 2016 and December 2020, we conducted a retrospective analysis of electronic medical record data of patients aged ≥ 70 years undergoing first-time arteriovenous fistula or arteriovenous graft surgery at three tertiary care hospitals: Korea University Anam Hospital, Korea University Guro Hospital, and Korea University Ansan Hospital. We investigated patient demographics, including sex, age, and medical history, such as that pertaining to diabetes mellitus, chronic obstructive pulmonary disease, cancer (any type diagnosed within the past five years), peripheral artery disease, cerebrovascular disease, ischemic heart disease, liver disease, congestive heart failure, and valve disease. Additionally, we examined preoperative complete blood count parameters, including platelet-to-lymphocyte, neutrophil-to-lymphocyte, and monocyte-to-lymphocyte ratios, and the type of arteriovenous fistula/arteriovenous graft surgery. Additionally, this study explored the one-year and overall mortality outcomes.

Only patients undergoing a complete blood count analysis on the day before surgery were included in the analysis. To account for potential granulocytopathy, we excluded patients with a white blood cell count < 1500/mL, lymphocyte count < 5%, monocyte count > 20%, and platelet count < 50,000/mL. Additionally, patients who had previously undergone peritoneal dialysis or hemodialysis were excluded from the study. The non-survival group comprised patients who died within one year after surgery; however, the survival group included patients who died after one year.

Blood samples were collected at the beginning of the hemodialysis session from patients requiring hemodialysis. The neutrophil-to-lymphocyte, platelet-to-lymphocyte, and monocyte-to-lymphocyte ratios were calculated by dividing the absolute neutrophil, platelet, and monocyte counts by the absolute lymphocyte count, respectively. The primary outcome was one-year mortality, and comorbidities that may have affected mortality were identified. Vascular access operations were performed in consultation with a nephrologist. Patients who had already undergone dialysis and those in a preemptive state were included. The inclusion criteria generally specified arteries and veins larger than 2 mm without stenosis, as determined by ultrasonography.

Continuous variables were compared using a t-test, whereas categorical variables were assessed using the chi-square or Fisher’s exact tests. Logistic regression with a forward conditional approach was used for multivariate analysis to identify risk factors for 1-year mortality, and Cox regression was used for total mortality. The p-value threshold for the multiple regression analysis was set at 0.2. The receiver operating characteristic curve analysis was conducted to determine the optimal cutoff value, sensitivity, and specificity. A p-value less than 0.05 was considered statistically significant.

Results

From the initial 1805 patients undergoing first-time arteriovenous fistula/arteriovenous graft surgeries across three hospitals, we included 448 patients aged ≥ 70 years in the analysis, excluding individuals based on the criteria outlined in the methods. Among these patients, 52 were non-survivors and died within one year after surgery, while 396 were survivors. Table 1 describes the baseline characteristics of the two groups.

Table 1 Baseline characteristics

Table 2 Multivariate analysis of one-year mortality in patients aged ≥ 70 years

Notably, patients in the non-survival group exhibited a significantly older age than those in the survival group (82.5 ± 6.5 vs. 79.4 ± 5.7 years, p < 0.001). Moreover, chronic obstructive pulmonary disease, cancer, and peripheral artery disease were more prevalent in the non-survival group, whereas the survival group had a higher proportion of patients with diabetes mellitus. The other variables did not show significant differences.

As described in Table 2, we conducted a logistic regression analysis with one-year mortality as the outcome. Univariate analysis was used to examine the variables mentioned in the baseline risk factor characteristics. Among these variables, diabetes mellitus, chronic obstructive pulmonary disease, cancer, peripheral artery disease, the preoperative monocyte-to-lymphocyte ratio, and the preoperative neutrophil-to-lymphocyte ratio were one-year mortality risk factors. These variables were subjected to multivariate analysis, confirming that the preoperative neutrophil-to-lymphocyte ratio (HR 1.15, p < 0.01), diabetes mellitus (HR 0.34, p < 0.01), cancer (HR 2.53, p = 0.02), and peripheral artery disease (HR 4.62, p < 0.01) were definitive one-year mortality risk factors. However, the preoperative platelet-to-lymphocyte and monocyte-to-lymphocyte ratios were not identified as one-year mortality risk factors. We generated a receiver operating characteristic curve to assess whether the preoperative neutrophil-to-lymphocyte ratio could predict one-year mortality. Among patients aged ≥ 70 years, the receiver operating characteristic curve for neutrophil-to-lymphocyte ratio yielded a 0.592 area under the curve (Fig. 1).

Fig. 1

One-year mortality receiver operating characteristic curve for the neutrophil-to-lymphocyte ratio

In addition to one-year mortality, we conducted Cox regression analysis with overall mortality as the outcome, with a mean follow-up of 917.0 ± 631.2 days. The preoperative neutrophil-to-lymphocyte and monocyte-to-lymphocyte ratios, chronic obstructive pulmonary disease, cancer, peripheral artery disease, and cerebrovascular disease were overall mortality risk factors in the univariate analysis. In the multivariate analysis, cancer (HR 1.92, p = 0.046), diabetes mellitus (HR 0.46, p = 0.013), peripheral artery disease (HR 2.08, p = 0.007), cerebrovascular disease (HR 1.89, p = 0.01), and monocyte-to-lymphocyte ratios (HR2.74, p = 0.007) emerged as the final risk factor for overall mortality. The preoperative neutrophil-to-lymphocyte and platelet-to-lymphocyte were not associated with overall mortality (Table 3).

Table 3 Multivariate analysis for total mortality in patients aged ≥ 70 years

Discussion

Among patients with chronic kidney disease aged ≥ 70 years undergoing first-time arteriovenous fistula/arteriovenous graft surgery in this study, the risk factors for one-year mortality were the preoperative neutrophil-to-lymphocyte ratio, diabetes mellitus, chronic obstructive pulmonary disease, and peripheral artery disease. Cancer, peripheral artery disease, cerebrovascular disease and monocyte-to-lymphocyte ratio were the risk factors for overall mortality. While the neutrophil-to-lymphocyte ratio demonstrated low diagnostic value with an area under the curve of 0.59 in the ROC analysis, it was identified as one of the important risk factors for 1-year mortality in the multivariate analysis. The authors posit that the neutrophil-to-lymphocyte ratio could be a useful reference for this specific purpose as a one-year mortality risk factor, despite its limited predictability considering the objective of the study aligned with the Kidney Disease Outcome Quality Initiative 2019 guidelines, which was to determine valuable hemodialysis access selection predictors, especially in fragile patients with chronic kidney disease patients such as older patients.

Diabetes mellitus was associated with a hazard ratio of 0.34 for 1-year mortality. Given that over 80% of patients in our study had diabetes, the severity of the disease, rather than its mere presence, may be a more relevant predictor of mortality. Factors such as insulin use, oral medication, and disease duration could provide clearer insights. As our study classified diabetes based solely on diagnostic codes, the lack of these detailed factors is a limitation.

Neutrophils are the most abundant immune cells in the human body and the primary defenders against infection as well as tissue damage. These cells play a crucial role in the inflammatory response [6]. The response of these cells causes various lymphocyte effects, including apoptosis, differentiation, downregulated proliferation, and decreased immune regulation owing to neurohumoral activation. [7, 8] The neutrophil-to-lymphocyte ratio is accessible through routine blood tests and has gained attention for its cost-effectiveness and widespread availability and has emerged as a prognostic marker in various chronic diseases [9]. The neutrophil-to-lymphocyte ratio is implicated in various solid organ tumors, including those in the stomach and pancreas, as well as cerebrovascular and cardiovascular diseases. This ratio functions as a surrogate inflammatory marker and correlates with treatment responsiveness, disease severity, and adverse events [10].

Chronic inflammation is prevalent in patients with chronic kidney disease and is closely linked to increased all-cause mortality and adverse cardiovascular outcomes [8]. Inflammation plays a crucial role in tubulointerstitial fibrosis and is a pivotal initiator significantly contributing to chronic kidney disease progression. The inflammatory state in patients with chronic kidney disease predisposes the vascular endothelium to atherosclerotic plaque formation and calcification, heightening the risk of cardiovascular events and mortality and involves neutrophils mediating an inflammatory response in kidney injury through various biochemical mechanisms, causing further tissue damage. Inflammation in chronic kidney disease is routinely used as a prognostic factor owing to its multifaceted impact on the disease trajectory and cardiovascular outcomes [11,12,13].

Recent investigations have highlighted the correlation between early arteriovenous fistula failure and higher white blood cell count in patients with chronic kidney disease [5]. Other studies have underscored the association between arteriovenous fistula patency and overall mortality in patients undergoing hemodialysis [14]. The Kidney Disease Outcome Quality Initiative 2019 guidelines advocate for a patient-centric approach in hemodialysis access selection, emphasizing the importance of considering individual patient factors over mechanical choices [4]. In this study, the risk factors associated with overall mortality one year after arteriovenous fistula/arteriovenous graft surgery were identified, including the neutrophil-to-lymphocyte ratio, diabetes mellitus, cancer, and peripheral artery disease. Similar to the categorization of patients based on one-year expected survival in the Kidney Disease Outcome Quality Initiative 2019, hemodialysis access planning should be carefully considered in patients with chronic kidney disease that have the aforementioned risk factors. Prioritizing hemodialysis access longevity and paying more attention to its patency are essential.

This study had several limitations. Old age poses a risk factor to diminish arteriovenous fistula patency [15], a notion the Kidney Disease Outcome Quality Initiative underscores, categorizing individuals as fragile with an expected survival of one year or less. The authors exclusively included individuals aged ≥ 70 years in this study. Nevertheless, further studies are essential in exploring the relationship between inflammatory mediators and hemodialysis access in different age groups. Another limitation of this study is the inability to control for additional confounding factors, including infection, previous lymphoproliferative disorders, and other surgeries. These confounders may have influenced the neutrophil-to-lymphocyte, platelet-to-lymphocyte, and monocyte-to-lymphocyte ratios. Finally, we could not investigate predictors that have a more long-term impact owing to the short follow-up duration.

In conclusion, the neutrophil-to-lymphocyte ratio is a potential early mortality predictor in older patients with chronic kidney disease undergoing hemodialysis access planning, offering possible utility in strategic decision-making. Despite the limitations of a short follow-up period, the findings underscore the significance of adopting a patient-centric approach in selecting hemodialysis access, aligned with the Kidney Disease Outcome Quality Initiative 2019 guidelines. Further investigations of diverse inflammatory markers are required to enhance the predictive accuracy of hemodialysis access planning.