Neutrophil‐to‐lymphocyte ratio can specifically predict the severity of hypertriglyceridemia‐induced acute pancreatitis compared with white blood cell

Objectives We aimed to evaluate the values of neutrophil‐to‐lymphocyte ratio (NLR) and white blood cell (WBC) in predicting severity of acute pancreatitis (AP) with different etiologies. Methods We compared NLR and WBC levels in patients with different etiologies and AP severity. The optimal cutoff value for them to predict severe acute pancreatitis (SAP) was determined by receiver operating characteristic (ROC) curve analysis. Results Both NLR and WBC were elevated in patients with SAP. After subgrouping AP by etiology, NLR was predictive of SAP only in hypertriglyceridemia‐induced AP (HTG‐AP), while WBC could effectively predict severity in both gallstone and HTG‐AP. The best cutoff value of WBC for predicting SAP in gallstone AP patients was 12.81 × 109/L, with sensitivity and specificity of 78.9% and 70.2%. The best cutoff value for NLR and WBC to differentiate HTG‐SAP was more than 5.88 and 15.89 × 109/L, respectively, with sensitivity and specificity of 87% and 50% for NLR and 56.5% and 75.76% for WBC. Conclusions Our study firstly demonstrated that NLR selectively played a role in HTG‐AP, while WBC could predict the severity of both gallstone and HTG‐AP. Furthermore, we firstly elucidated that NLR was more sensitive and accurate in judging the severity of HTG‐AP compared with WBC.

can be easily obtained at low cost through an automatic hematology analyzer, many research groups have further confirmed the value of NLR in predicting disease severity and worse clinical outcomes in a variety of diseases, including cerebrovascular diseases, 6-8 cardiovascular diseases, 9 and neoplasm. 10 Besides, cumulative evidence has suggested that NLR was associated with AP and better than other serum markers in predicting severity and prognosis of AP. However, there was a lack of data regarding their utilities and comparison in the development of AP with different etiologies. The aim of this study was to explore the value of NLR in determining the severity of AP with different causes compared with white blood cell (WBC).

| Patients
We recruited 268 AP patients admitted to digestive department of our hospital from January 2012 to January 2017. AP was diagnosed based on patients' clinical symptoms, radiographic examination, and laboratory results. Disease severity was assessed using Atlanta typing 2012. 11 The most three common causes of AP are gallstone, hypertriglyceridemia, and alcohol. 1,12 There were 223 cases with mild acute pancreatitis (MAP) and 45 cases with moderately severe acute pancreatitis (MSAP) or SAP, of which 123 were gallstone AP, 56 were alcoholic AP, and 89 were hypertriglyceridemia-induced AP (HTG-AP). Because the sample size was not big enough, MSAP was also classified as SAP in our study. Exclusion criteria are as follows: (a) other causes induced AP; (b) acute exacerbation of chronic pancreatitis; (c) tumor induced AP; (d) AP onset was more than 24 hours when at hospital; and (e) missing data were available.

| Data collection
Peripheral blood samples were obtained within 24 hours after AP onset and detected by a full-automatic hemolytic analyzer. WBC, neutrophil, and lymphocyte counts were obtained, and NLR was calculated.
Demographic variables (including age and gender) were also included.

| Statistical analysis
Continuous variables were presented as mean ± standard deviation.
Frequency and percentage (%) were used to describe categorical data. The analysis of variance and t test were applied in comparing the intergroup difference of measurement data. The optimal cutoff values of NLR and WBC were determined by ROC curve analysis.
Test which has the larger AUC has the better diagnostic value. A Pvalue <0.05 was considered statistically significant. All analyses were performed using SPSS 17.0 Software (SPSS, Inc., Chicago, IL, USA).

| Clinical characteristics
A total of 268 patients with AP were enrolled, consisted of 140 males and 128 females with an average age of 55 years old (range: 28-75).
We next classified patients according to the etiology, of which gallstone accounted for 45.90% (n = 123), alcohol 20.90% (n = 56), and hyperlipidemia 33.21% (n = 89). According to the Atlanta typing 2012, SAP had a proportion of 16.8% (n = 45). There were no statistical differences in age and sex between SAP group and MAP group. Thus, we excluded age and sex as latent impurity interferon of SAP. Patients with SAP mainly came from HTG-AP and gallstone AP groups (Table 1). In our study, SAP accounted for 15.45%, 5.36%, and 25.84% in gallstone AP, alcoholic AP, and HTG-AP, respectively (Table 2).
The mean levels of NLR and WBC in patients with SAP were significantly higher than those in the MAP group (12.24 ± 8.37 vs 8.32 ± 7.24, P = 0.001; 16.17 ± 5.00 vs 12.14 ± 4.35 × 10 9 /L, P = 0.000, respectively; 13.85 ± 4.11 × 10 9 /L, P = 0.020, respectively) than in MAP. In particular, the difference in NLR was more obvious than in WBC. In the alcoholic AP group, our study showed neither NLR nor WBC could be regarded as indicators of the disease severity (Table 5).

| Values of WBC and NLR in predicting SAP by ROC curve
We calculated the AUCs of NLR and WBC for predicting SAP by ROC curve in gallstone group and HTG group. For the gallstone AP, AUC value of NLR and WBC was 0.657 and 0.786, respectively.
The best WBC cutoff value on predicting SAP in gallstone AP patients was 12.81 × 10 9 /L, with sensitivity and specificity of 78.9% and 70.2% (Figure 1). For the HTG-AP group, AUC value of NLR and WBC was 0.706 and 0.653, respectively. The differences were both statistically significant. It seemed that NLR was better than WBC in predicting the severity of HTG-AP. According to the ROC curve, the best cutoff value for NLR and WBC to differentiate HTG-SAP with HTG-MAP was more than 5.88 and 15.89 × 10 9 /L, respectively, with sensitivity and specificity of 87% and 50% for NLR and 56.5% and 75.76% for WBC ( Figure 2). Thus, we came to conclusion that NLR was more sensitive and accurate in judging the severity of HTG-AP, while WBC was more suitable for gallstone AP.

| D ISCUSS I ON
It was reported that there was a decrease in lymphocyte numbers and an increase in lymphocyte apoptosis, associated with lymphocyte dysfunction. 13 NLR integrates two opposing and complementary components of the immune pathway and represents the balance between inflammatory activating factor neutrophils and inflammatory regulatory factor lymphocytes. Besides, the higher NLR value represents a more imbalanced inflammatory state. 14 It was first reported as a parameter assessing systemic inflammation which could be easily measured by Zahorec. 15 The course of SAP is complicated, and the prognosis is poor. Imbalance of immune response is one of the most important causes leading to severe pancreatitis. 16 Therefore, we assumed that NLR should play an important role in predicting the severity of AP. and reactive oxygen species. 24 It was reported that higher NLR value represented a more imbalanced inflammatory state. 14 Therefore, our results which showed NLR could selectively predict the severity of HTG-AP and was more sensitive and accurate than WBC in judging the severity can be explained as NLR could be regarded as a representative of imbalanced inflammatory cascade reaction. Because numbers of patients with different causes of AP were small, and data were from single center, the study on the role of NLR and WBC in the process of the severity of AP needs a larger sample and a more reasonable research plan.

ACK N OWLED G M ENTS
This work was supported in part by Natural Science Foundation