Background
Acute pancreatitis (AP) is an inflammatory process in which local pancreatic injury leads to systemic inflammation through activation of cytokine cascades [
1]. The clinical extent of AP varies widely from no symptoms to systemic inflammatory response syndrome (SIRS), persistent organ failure (POF), and death [
2]. The clinical presentation of AP is both unreliable and nonspecific and exhibits a sensitivity less than 40% for the prediction of adverse outcomes [
3]. Also, the underlying pathophysiology behind the progression of local pancreatic injury to SIRS is not fully understood [
4]. Due to the diverse presentations of AP and its unknown pathophysiology, multiple severity scoring systems have been designed to help clinicians in triaging AP patients and predicting their prognosis. The Ranson score, the Acute Physiologic Assessment and Chronic Health Evaluation II (APACHE II) score, the Bedside Index for Severity in Acute Pancreatitis (BISAP) score, and the Glasgow-Imrie criteria are currently in wide use. However, these systems are time-consuming and difficult to apply to patients outside of intensive care settings because they use many variables [
5]. Also, they are unsuitable for the evaluation of patients at the time of admission or shortly thereafter. Simplified serum markers such as C-reactive protein (CRP), procalcitonin, interleukin-6, and interleukin-8 have been applied to predict the prognosis or severity of AP, but they are expensive, not readily available, and cannot adequately predict the prognosis or severity of AP [
6].
Recently, many research groups have studied the value of hematological components, such as the neutrophil to lymphocyte ratio (NLR) and the platelet to lymphocyte ratio (PLR), in predicting disease severity and outcomes across a variety of diseases, including inflammation, cardiovascular disease, and neoplastic states [
2]. The superiority of NLR over total white blood cell (WBC) count, which is used in the Ranson, APACHE-II, and Glasgow-Imrie scoring systems, has been demonstrated in a variety of medical conditions [
7]. Furthermore, a few studies have shown that PLR is superior to NLR as a prognostic factor in certain disease conditions [
8‐
10]. Increased NLR and PLR ratios have been associated with inflammatory conditions, and poor outcomes in severe AP are explained by uncontrolled SIRS and its progression to multi-organ dysfunction syndrome [
6].
Although a few studies have considered NLR and its prognostic value in AP, no studies have yet examined the prognostic value of PLR in AP. In the present study, we evaluated NLR and PLR values as independent prognostic factors for adverse outcomes in AP and sought to improve previous scoring systems by incorporating NLR and PLR.
Discussion
In the present study, we investigated the value of NLR and PLR as predictive markers of AP severity. We found that NLR and PLR were well correlated with other scoring systems in patients with gallstone AP. In addition, NLR and PLR showed significant predictive ability for POF in patients with gallstone AP. However, in patients with alcoholic AP, NLR and PLR were not correlated with other scoring systems.
NLR was first introduced as an easily measurable parameter assessing systemic inflammation and stress in critically ill patients [
11]. Later, PLR was also found to be an inflammatory marker, and the role of platelets as a critical link between inflammation and microvascular dysfunction has since been investigated [
12‐
14]. The prognostic value of these two parameters has been confirmed in a variety of clinical conditions, and PLR was shown to be superior to NLR in certain cancers [
8‐
10]. AP is an inflammatory condition characterized by activation of both innate and adaptive immune responses. Activation and modulation of neutrophils and platelets play a core role in establishing host defenses in settings of systemic inflammation; however, excessive inflammatory response causes massive cell transmigration to the pancreas and subsequent release of aggressive defense molecules, resulting in destruction of the pancreas and organ failure [
15‐
18].
A few studies have investigated the relationship between NLR and outcome in patients with AP. Azab et al. [
7] first applied the concept of NLR to patients with AP in 2011. They found that NLR was a better predictor of ICU admission or prolonged hospitalization in AP than was total WBC count and suggested a cutoff value of < 4.7 as a predictor of a poor outcome [
7]. However, in-hospital mortality was extremely low in that study, and the investigators failed to assess records of organ failure. Suppiah et al. [
6] revealed an association between NLR measured in the first 48 h and the risk of AP developing into a more severe form. However, that study was limited by a small sample size (
n = 146), and the AP cases included were mostly mild, with no local/systemic complications or organ failure. Recently, Gulen et al. [
19] investigated the association between NLR and early mortality and argued that NLR is not a significant independent prognostic factor. Zhang et al. [
2] demonstrated that elevated NLR is associated with POF, ICU stay longer than > 7 days, and increased in-hospital mortality in a Chinese population.
However, despite the demonstrated superiority of PLR over NLR in predicting the outcome of inflammation in several clinical conditions, no previous study has investigated the predictive value of PLR at the time of admission on outcomes in patients with AP. Furthermore, although they are grouped together as pancreatitis, gallstone AP and alcoholic AP each has a distinct pathophysiology, and no study has compared NLR or PLR between them. Therefore, we investigated the value of PLR in predicting AP outcomes and compared differences between NLR and PLR patterns in two distinct forms of AP. We excluded pancreatitis caused by factors other than gallstone or alcohol. When all AP cases were combined, NLR and PLR were not significant independent predictive factors of POF. However, after subgrouping AP by etiology, both NLR and PLR were independent predictive factors of POF in gallstone AP. In alcoholic AP, NLR was a significant predictor, but PLR was not. This can be explained by the different mechanism of alcohol AP. Alcoholic AP is usually associated with chronic liver disease. In our results, the number of liver cirrhosis patients was larger and the platelet count was lower in alcoholic AP compared to gallstone AP. Thrombocytopenia is related to chronic liver disease due to impaired platelet production and decreased hepatic synthesis of thrombopoietin [
20]. Therefore, PLR can vary according to liver function as well as systemic inflammation. Interestingly, CRP, a marker traditionally used to assess the severity of inflammation [
21‐
23], failed to predict POF in gallstone AP and alcoholic AP. This suggests the superiority of NLR and PLR to CRP in predicting the course of gallstone AP.
Although the exact mechanism of alcoholic AP has not been elucidated, our findings imply a fundamental difference in pathophysiology between the two subgroups. Although elevated NLR and PLR can be used as predictive biomarkers in AP, interpretation should follow confirmation of the etiology. Furthermore, our findings challenge the rationale of applying a uniform prognostic scoring system to all AP. Replacing WBC count with NLR or PLR in traditional prognostic scoring systems could improve their performance [
10].
Also, pancreatic cancer can induce impairment of the patient’s immune system through systemic inflammation [
24]. In this aspect, NLR and PLR can reflect the status of the immune system in patients with pancreatic cancer. Several studies have reported that NLR and PLR were correlated with poor overall survival in patients with pancreatic cancer [
25,
26]. A recent study demonstrated that NLR was a predictive marker for the presence of invasive carcinoma in patients with intraductal papillary mucinous neoplasm [
27]. Therefore, NLR and PLR have a role as biomarkers in pancreatic malignancy.
This study has several limitations. First, the number of patients enrolled in this study was small, and this study was performed in a tertiary care center, which could have resulted in disproportional inclusion of patients with severe disease status and tendency to progress to POF. Such selection bias might have overestimated the predictive value of elevated NLR or PLR. Second, we did not compare NLR or PLR with other biochemical markers, such as procalcitonin and IL-6. Third, we did not describe changes in NLR or PLR during treatment, which could estimate the prognosis of AP. Despite these limitations, this study also has strengths. This is the first prospective study investigating the predictive value of PLR in AP and the difference between NLR and PLR in two subgroups of AP. Also, all laboratory values were obtained within 1 h of initial presentation, minimizing changes in WBC and platelet counts caused by hydration and medication.