Abstract
Objective: To study the association between a high preoperative neutrophil lymphocyte ratio (NLR) and red cell distribution width (RDW) with arteriovenous fistula (AVF) failure, as well as to determine the cut-off values in a South Asian population.
Materials and Methods: A total of 150 consecutive patients with a failed fistula who presented in the Department of Vascular Surgery between January 2014 and January 2016. Patients fulfilling the inclusion criteria were selected as defined as Case. They were compared with 150 patients who had matured fistulae (Control).
Results: A significant difference was found between the Case and Control groups in mean preoperative NLR (3.3±0.5 versus 2.2±0.9, P value=0.011) and RDW (15.9±2.9 versus 13.6±1.1%, P value of 0.02), respectively. Multivariate analysis revealed that NLR (Odds Ratios (OR) 1.39; 95% Confidence Intervals (CI) 1.02 to 2.08; P<0.001) and RDW (OR 1.39; 95%CI 1.11 to 1.69; P<0.001) were strong independent predictors of AVF failure. A receiver operating characteristic curve analysis showed a cut-off value of 2.65 (specificity 80%, sensitivity 98%) and 15.1 (specificity 79%, sensitivity 98%) for NLR and RDW, respectively.
Conclusion: Increased preoperative NLR and RDW were associated with a high rate of AVF failure in a South Asian population.
Keywords: arteriovenous fistula, neutrophil lymphocyte ratio, red cell distribution width
Introduction
The most common cause of access failure is stenosis, and the development of thrombosis in the immediate post-stenotic site.1) The stenosis results from neointimal hyperplasia and a vast array of inflammatory factors.1–4) Various inflammatory predictive parameters to assess association of arteriovenous fistula (AVF) failure and inflammation have been studied, such as neutrophil lymphocyte ratio (NLR), red cell distribution width (RDW), white blood cells (WBC) count, Interlukin-6, C-reactive protein (CRP), and erythrocyte sedimentation rate.3–5)
NLR is a very simple indicator to assess systemic inflammation. NLR has been used as an indicator to assess coronary stenosis and to predict survival in patients with acute coronary syndrome.6) Furthermore, it is a strong predictor in stent failure due to restenosis.7) Studies have shown an association of high preoperative NLR with failure of a saphenous vein graft due to stenosis in coronary bypass grafting.8) Since NLR is a strong predictor of stenosis secondary to inflammation, and inflammation plays an important role in AVF stenosis and failure, the association of NLR with AVF stenosis is an area of interest.
RDW is another parameter, which has strong association with inflammation and subsequently AVF failure.9) Studies claim RDW to be a better marker than CRP and WBC count.10)
Considering the ease and availability of RDW and NLR in a simple blood test, this prospective cohort study was designed to test whether raised NLR and RDW are independent predictors of AVF failure. Additionally, what are the likely preoperative cut-off values that can be clinically useful to predict AVF failure in the South Asian population.
Patients and Methods
All patients reporting to Vascular Surgery Clinic in Combined Military Hospital Lahore between January 2014 and January 2016 for AVF creation, who fulfilled the inclusion criteria, were included in this study. All patients underwent a Duplex Scan with 5 MHz linear probe by a consultant radiologist to assess the superficial veins of the upper limb for thrombosis, stenosis, or aneurysmal dilation. All such patients were excluded from the study. Patients who had previous surgery on the arm, advanced cardiac failure [New York Heart Association (NYHA) functional classification stage 3 or 4], hematological malignancy, bleeding disorder, or using anticoagulants were also excluded from the study.
Those who fulfilled the inclusion criteria had their blood samples taken on the day of the surgery prior to the procedure. Blood samples were tested for the parameters described in Table 1. All patients underwent an AVF procedure by a consultant vascular surgeon, who had a minimum of five years of experience of doing such procedures under local anesthesia. Patients were given oral antibiotics and analgesia on discharge. They were provided written instructions about hand and forearm exercises and general care of the AVF.
Table 1 Clinical and biochemical parameters of the patients.
| Patient parameters | Fistula matured (Group A) | Fistula failed (Group B) | P value | All patients |
|---|---|---|---|---|
| Total patients (N) | 150 | 150 | 300 | |
| Mean age (Years±SD) | 45±13 | 43±15 | 0.847 | 44±12 |
| Gender male [n, (%)] | 118 (51.3) | 112 (48.7) | 0.792 | 230 (76.6) |
| Gender female [n, (%)] | 32 (45.7) | 38 (54.2) | 0.527 | 70 (23.3) |
| Ejection fraction (%±SD) | 59±5 | 58±6 | 0.691 | |
| NYHA class 1 (n) | 101 | 97 | 0.449 | 198 |
| NYHA class 2 (n) | 49 | 53 | 0.521 | 102 |
| Atherosclerotic risk factors: | ||||
| Smoker [n, (%)] | 84 (56) | 79 (52.6) | 0.679 | 163 (54.3) |
| Ischemic heart disease [n, (%)] | 24 (16) | 27 (18) | 0.592 | 51 (17.0) |
| Systolic BP (mmHg±SD) | 142±13 | 143±12 | 0.773 | |
| Diastolic BP (mmHg±SD) | 93±9 | 94±11 | 0.431 | |
| Diabetes mellitus [n, (%)] | 61 (40.6) | 52 (34.6) | 0.511 | 113 (37.6) |
| Biochemical parameters: | ||||
| HDL cholesterol (mg/dL±SD) | 49.4±11.9 | 50.2±12.2 | 0.629 | |
| LDL cholesterol (mg/dL±SD) | 101.4±32.8 | 100.7±37.2 | 0.482 | |
| Triglycerides (mg/dL±SD) | 192.2±83.4 | 195.5±91.1 | 0.745 | |
| HbA1c in diabetic patients (%±SD) | 7.9±0.8 | 8.1±0.9 | 0.091 | |
| White cell count (109/L±SD) | 7.4±2.1 | 8.1±1.9 | 0.024 | 7.9±1.8 |
| Neutrophils (109/L±SD) | 4.8±1.7 | 5.6±1.4 | 0.068 | 4.9±2.1 |
| Lymphocytes (109/L±SD) | 1.7±0.2 | 1.8±0.4 | 0.215 | 1.7 ±0.5 |
| Monocytes (109/L±SD) | 0.5±0.1 | 0.7±0.4 | 0.414 | 0.6±0.1 |
| NLR (×± SD) | 2.2±0.9 | 3.3±0.5 | 0.011 | 2.6±0.7 |
| Hematocrit (%±SD) | 29.4±3.4 | 33.2±2.9 | 0.082 | 31.2±1.9 |
| RBCs (×1012/L±SD) | 3.1±0.8 | 3.6±0.5 | 0.049 | 3.4±0.6 |
| Hemoglobin (g/dL±SD) | 9.2±1.1 | 10.3±1.5 | 0.075 | 9.8±1.1 |
| MCV (fL±SD) | 88.4±8.9 | 89.9±7.4 | 0.871 | 89.2±6.9 |
| MCH (pg±SD) | 30.1±2.9 | 30.6±3.1 | 0.894 | 30.4±2.6 |
| RDW (%±SD) | 13.6±1.1 | 15.9±2.9 | 0.020 | 14.4±1.9 |
| Platelets (×103/µL±SD) | 229±75 | 258±79 | 0.061 | 248±81 |
Patients were regularly followed up in the clinic at 2 weeks, 4 weeks, 2 months, and then every 3 months for a minimum of one year. A functionally mature AVF is defined by Kidney Disease Outcome Quality Initiative (KDOQI) guidelines as one that can be easily cannulated and has at least six successful consecutive dialysis sessions.11) The primary outcome of the study was failure of the fistula.
The data was analyzed using the Statistical Package for Social Sciences version 20.0. The numerical data, such as age was calculated as the mean and standard deviation. Categorical data, such as gender was recorded as frequency and percentage and compared between the groups using a Chi-Square test. Patients were divided into two groups. Group A was defined as patients with a mature fistula, and Group B was defined as a failed fistula. The comparison between the two groups was performed using a Student’s t test. The association of laboratory markers with the maturation (yes) or failure (no) of AVF was analyzed using univariate logistic regression. Stepwise reduction was performed using multivariate regression for those variables when the P value in univariate regression was significant. A receiver operating characteristic (ROC) curve analysis was done to establish the cut-off value for all of the statistically significant variables of multivariate regression. Odds Ratios (OR) and 95% Confidence Intervals (CI) were calculated. The results were considered statistically significant if the P value was found to be less than or equal to 0.05.
Results
A total of 365 consecutive patients fulfilling the inclusion criteria were included in the study. Of these, 65 were excluded (39 died and 26 lost to 1 year follow up). There were 150 patients with a failed fistula. Of the 150 patients with a failed AVF, 31 (20.7%) had early failure (defined as failure to mature within the first 8 weeks) and 119 (79.3%) had late failure. The average time to failure in this group was 5±4 months. Patients with a failed fistula were assigned to Group B (Case). Group A comprising of similar number of consecutive patients who had their fistula matured were selected for comparison (Control).
The minimum age of patients was 21, and the maximum age was 62 years with mean age of 44±12 years. Out of these 300 cases, 230 (76.6%) were males and 70 (23.3%) were females. The male to female ratio was 3.2 : 1. There was no difference in age and gender between the two groups. Similarly, major risk factors for atherosclerotic disease, such as diabetes, hypertension, and smoking were not statistically significant different between the two groups. Furthermore, there were no differences in cardiac status when compared in terms of ejection fraction (59±5 mature versus 58±6 failed group; P=0.691) and NYHA class (Table 1).
Patients with a failed fistula had a higher WBC count when compared to the patients with matured fistulae (8.1±1.9 versus 7.4±2.1×109/L; P=0.024). Similarly, they had higher hematocrit (33.2±2.9 versus 29.4±3.4%; P=0.082), red blood cells (RBCs) count (3.6±0.5 versus 3.1±0.8×1012/L; P=0.049), and hemoglobin (10.3±1.5 versus 9.2±1.1 g/dＬ; P=0.075). The mean NLR levels in patients with failed fistula group were higher in comparison to the matured fistula group (3.3±0.5 versus 2.2±0.9) with a P value of 0.011, which is statistically significant. Furthermore, when RDW was compared, it also showed a statistically significant difference between the two groups (15.9±2.9 in fistula failure group versus 13.6±1.1% in fistula mature group) with a P value of 0.02. No significant differences were found in the remainder of the biochemical parameters analyzed.
On further analysis of the failed group, the mean NLR was higher in patients who had early failure of fistula when compared to late failure group (3.5±0.5 versus 3.1±0.4; P=0.049). The difference was statistically significant. For RDW, no significant difference was found between the two failure groups (15.7±3.1 early versus 16.0±2.6 late failure; P=0.054). No significant difference was found between the failed sub-groups in the other biochemical parameters described in Table 1.
The results of univariate analysis show that an increase in WBC count (OR 1.16; 95%CI 1.02 to 1.45; P=0.044), NLR (OR 1.12; 95%CI 0.95 to 1.54; P=0.011), RBC count (OR 0.51; 95%CI 0.32 to 0.98; P=0.039) and RDW (OR 1.05; 95%CI 0.99 to 1.41; P=0.011) significantly increase the risk of fistula failure (Table 2). Since hemoglobin and hematocrit are strongly correlated with RBCs count (r2=0.81 and 0.85 respectively, both with a P value of <0.001), they were excluded from multivariate analysis. Neutrophils and WBCs count were significantly correlated with NLR (r2=0.82 and 0.83 respectively; P<0.001), so they were also excluded. Multivariate analysis revealed that NLR (OR 1.39; 95%CI 1.02 to 2.08; P<0.001) and RDW (OR 1.39; 95%CI 1.11 to 1.69; P<0.001) are strong independent predictors of AVF failure.
Table 2 Outcome of the univariate and multivariate analysis.
| Biochemical parameters | Univariate analysis ORs (95%CI) | P value | Multivariate analysis ORs (95%CI) | P value |
|---|---|---|---|---|
| HDL cholesterol (mg/dL±SD) | 0.98 (0.94–1.1) | 0.351 | ||
| LDL cholesterol (mg/dL±SD) | 0.99 (0.97–1.1) | 0.561 | ||
| Triglycerides (mg/dL±SD) | 1.00 (1.00–1.1) | 0.766 | ||
| White cell count (109/L±SD) | 1.16 (1.02–1.45) | 0.044 | ||
| Neutrophils (109/L±SD) | 1.22 (1.00–1.69) | 0.061 | ||
| Lymphocytes (109/L±SD) | 1.25 (0.97–1.85) | 0.598 | ||
| Monocytes (109/L±SD) | 0.65 (0.05–1.21) | 0.368 | ||
| NLR (×±SD) | 1.12 (0.95–1.54) | 0.011 | 1.39 (1.02–2.08) | <0.001 |
| Hematocrit (%±SD) | 0.88 (0.76–0.99) | 0.063 | ||
| RBCs (×1012/L±SD) | 0.51 (0.32–0.98) | 0.039 | 0.96 (0.20–3.69) | 0.893 |
| Hemoglobin (g/dL±SD) | 0.80 (0.65–1.02) | 0.075 | ||
| MCV (fL±SD) | 0.99 (0.89–1.05) | 0.726 | ||
| MCH (pg±SD) | 0.92 (0.77–0.99) | 0.594 | ||
| RDW (%±SD) | 1.05 (0.99–1.41) | 0.011 | 1.39 (1.11–1.69) | <0.001 |
| Platelets (×103/μL±SD) | 0.99 (0.91–1.03) | 0.742 |
ROC curve analysis was done to determine the cut-off values for NLR and RDW. The optimal cut-off value for NLR was 2.65. The area under curve (AUC) was 0.792 (0.711–0.843) with a specificity of 80% and sensitivity of 98% (Fig. 1). Figure 2 shows the cut-off value for RDW as 15.1, with a specificity of 79%, sensitivity of 98% and AUC as 0.821 (0.764–0.901).
Fig. 1 ROC curve analysis for NLR.
Fig. 2 ROC curve analysis for RDW.
Discussion
Studies suggest that inflammation causes neointimal hyperplasia, which results in thrombosis at the anastomotic site and is responsible for failure of the AVF.4,12,13) Neointimal hyperplasia is pathologically similar to atherosclerosis.14) Studies have clearly demonstrated that atherosclerosis is the main pathology in coronary occlusive disease,6) bare metal stent failure7) and autologous saphenous vein graft failure8); and NLR has been shown as a strong independent predictor in all of these scenarios. Considering atherosclerosis as the primary mechanism responsible for AVF stenosis and failure, and NLR being a strong predictor of atherosclerosis elsewhere; it can also be a positive predictor of AVF failure. In our study, we have revealed a clear association between inflammation and AVF failure using NLR. The results are consistent with other studies showing similar association between NLR and inflammation, leading to AVF failure.2–4)
In our study, we have clearly demonstrated using multivariate logistic regression that raised preoperative NLR is a strong independent predictor of AVF failure (OR 1.39; 95%CI 1.02–2.08; P<0.001). Similar results are advocated by Yilmaz et al. in his logistic regression model of 108 patients (OR 4.734; 95%CI 2.987–7.321; P=0.001).15) Yilmaz et al. concluded that a raised NLR is a strong predictor of AVF failure, and its correlation with failure is due to inflammation and atherosclerosis resulting in neointimal hyperplasia.15)
The mean normal value for NLR is variable depending upon age and ethnicity.16) Its reported values in the African population is 2.8, while it is much lower in the American white population (1.76).16,17) It was even lower in healthy Chinese population (1.5).18) Studies also suggest that the mean NLR is lower in a healthy age group of 18–50 years when compared with 51–80 years.16) Considering this variation, we performed this study determine the cut-off values of NLR in our South Asian population. Various studies have concluded that cut-off values for NLR could be useful in predicting the AVF failure. Çildağ et al. proposed a cut-off value of NLR as 3.18 for the determination of primary patency (sensitivity 81.4%, specificity 51.4%).19) Yilmaz et al. concluded a cut-off value of 2.70 for AVF stenosis (sensitivity 98.4%, specificity 75%).15) In our study, we also calculated the cut-off value using the ROC curve analysis. The value was 2.65 with a specificity of 80% and sensitivity of 98%.
RDW is another simple blood parameter, which is routinely reported in automated laboratory blood count equipment. RDW is calculated as standard deviation of red blood cell volume divided by the mean corpuscular volume (MCV). RDW has been reported as a strong independent predictor of mortality in patients with coronary occlusive disease. The exact mechanism is unknown; however, it is suggested that high RDW is an indicator of poor red cell production, malnutrition, and inflammation.9) One can argue that a raised RDW can be due to anemia reflected by low hematocrit, hemoglobin, and RBCs levels, suggesting that confounding factors in the results. Univariate analysis of the two variables (hemoglobin and hematocrit) did not show an association between them and the raised RDW, since the P values were not significant in either case. RBCs levels were subjected to multivariate analysis, and the P value failed to show any statistically significant association between RBCs and RDW. We concluded that the increased RDW is an independent predictor for AVF failure. Similar results were proposed by Bojakowski et al. in 2012, when they concluded that raised RDW is a better predictor of AVF failure than other inflammatory markers such as raised WBC count and CRP, suggesting RDW as an independent predictor of AVF failure.10)
The strength of this study is its prospective design, non-purposive consecutive sampling, standardized preoperative biochemical blood parameter measurements, univariate and multivariate analysis of parameters, and proposed cut-off values of statistically significant parameters for a South Asian population. However, this is a single center study with no randomization of the study groups. Other factors, which may affect maturation of fistula, such as small vein caliber, obesity, low flow rate, anatomical location of access site in terms of depth and tortuosity, non-standardization of the fistula access protocols in terms of experience of dialysis staff, and patient education could be potential confounding factors affecting the final results. We propose larger multicenter randomized trials with stricter preoperative vascular assessment and standardized protocols for dialysis access to keenly assess the association between these simple markers and the fistula failure rate.
Conclusion
Raised preoperative NLR and RDW were found to be associated with a high rate of AVF failure in a South Asian population.
Funding
None.
Disclosure Statement
All authors have no conflict of interest.
Author Contributions
Study conception: RU
Data collection: RU, MN
Analysis: RU, MJ
Investigation: RU, MN
Writing: RU, MJ
Critical review and revision: RU, MN
Final approval of the article: all authors
Accountability for all aspects of the work: all authors
References
- 1).Stolic R. Most important chronic complications of arteriovenous fistulas for hemodialysis. Med Princ Pract 2013; 22: 220-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2).Juncos JP, Grande JP, Kang L, et al. MCP-1 contributes to arteriovenous fistula failure. J Am Soc Nephrol 2011; 22: 43-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3).Marrone D, Pertosa G, Simone S, et al. Local activation of interleukin 6 signaling is associated with arteriovenous fistula stenosis in hemodialysis patients. Am J Kidney Dis 2007; 49: 664-73. [DOI] [PubMed] [Google Scholar]
- 4).Roy-Chaudhury P, Khan R, Campos B, et al. Pathogenetic role for early focal macrophage infiltration in a pig model of arteriovenous fistula (AVF) stenosis. J Vasc Access 2014; 15: 25-8. [DOI] [PubMed] [Google Scholar]
- 5).Lippi G, Targher G, Montagnana M, et al. Relation between red blood cell distribution width and inflammatory biomarkers in a large cohort of unselected outpatients. Arch Pathol Lab Med 2009; 133: 628-32. [DOI] [PubMed] [Google Scholar]
- 6).Arbel Y, Finkelstein A, Halkin A, et al. Neutrophil/lymphocyte ratio is related to the severity of coronary artery disease and clinical outcome in patients undergoing angiography. Atherosclerosis 2012; 225: 456-60. [DOI] [PubMed] [Google Scholar]
- 7).Turak O, Ozcan F, Isleyen A, et al. Usefulness of the neutrophil-to-lymphocyte ratio to predict bare-metal stent restenosis. Am J Cardiol 2012; 110: 1405-10. [DOI] [PubMed] [Google Scholar]
- 8).Taşoğlu I, Turak O, Nazli Y, et al. Preoperative neutrophil-lymphocyte ratio and saphenous vein graft patency after coronary artery bypass grafting. Clin Appl Thromb Hemost 2014; 20: 819-24. [Epub ahead of print]. DOI: 10.1177/1076029613484086 [DOI] [PubMed] [Google Scholar]
- 9).Förhécz Z, Gombos T, Borgulya G, et al. Red cell distribution width in heart failure: prediction of clinical events and relationship with markers of ineffective erythropoiesis, inflammation, renal function, and nutritional state. Am Heart J 2009; 158: 659-66. [DOI] [PubMed] [Google Scholar]
- 10).Bojakowski K, Dzabic M, Kurzejamska E, et al. A high red blood cell distribution width predicts failure of arteriovenous fistula. PLoS One 2012; 7: e36482. DOI: 10.1371/journal.pone.0036482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11).Rayner HC, Besarab A, Brown WW, et al. Vascular access results from the Dialysis Outcomes and Practice Patterns Study (DOPPS): performance against Kidney Disease Outcomes Quality Initiative (K/DOQI) Clinical Practice Guidelines. Am J Kidney Dis 2004; 44 Suppl 2: 22-6. [DOI] [PubMed] [Google Scholar]
- 12).Kaygin MA, Halici U, Aydin A, et al. The relationship between arteriovenous fistula success and inflammation. Ren Fail 2013; 35: 1085-8. [DOI] [PubMed] [Google Scholar]
- 13).Morton SK, Rodríguez AJ, Morris DR, et al. A systematic review and meta-analysis of circulating biomarkers associated with failure of arteriovenous fistulae for Haemodialysis. PLoS One 2016; 11: e0159963. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14).Gagliardi GM, Rossi S, Condino F, et al. Malnutrition, infection and arteriovenous fistula failure: is there a link? J Vasc Access 2011; 12: 57-62. [DOI] [PubMed] [Google Scholar]
- 15).Yilmaz H, Bozkurt A, Cakmak M, et al. Relationship between late arteriovenous fistula (AVF) stenosis and neutrophil-lymphocyte ratio (NLR) in chronic hemodialysis patients. Ren Fail 2014; 36: 1390-4. [DOI] [PubMed] [Google Scholar]
- 16).Alexander NI. Reference values of neutrophil-lymphocyte ratio, platelet-lymphocyte ratio and mean platelet volume in healthy adults in north central Nigeria. J Blood Lymph 2016; 6: 143-6. [Google Scholar]
- 17).Azab B, Camacho-Rivera M, Taioli E. Average values and radical differences of neutrophil lymphocyte ratio among a nationally representative sample of United States subjects. PLoS One 2014; 9: e112361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18).An X, Mao P, Wei X, et al. Elevated neutrophil to lymphocyte ratio predicts overall and cardiovascular mortality in maintenance peritoneal hemodialysis patients. Int Urol Nephrol 2012; 44: 1521-8. [DOI] [PubMed] [Google Scholar]
- 19).Çildağ MB, Çildağ S, Köseoğlu ÖF. The Relationship between neutrophil-lymphocyte ratio and primary patency of percutaneous transluminal angioplasty in hemodialysis arteriovenous fistula stenosis when using conventional and drug-eluting balloons. Cardiovasc Intervent Radiol 2016; 39: 1702-7. [DOI] [PubMed] [Google Scholar]