New predictor of acute necrotizing pancreatitis: Red cell distribution width

Background. Acute pancreatitis (AP) is inflammation of the pancreas of various severity ranging from mild abdominal pain to mortality. AP may be classified as acute interstitial edematous pancreatitis (AEP) or acute necrotizing pancreatitis (ANP), according to the revised Atlanta criteria. Most of the patients with AP are AEP (75–85% of patients), while 15–25% of patients have ANP. The mortality rate is 3% in AEP and 15% in ANP. Thus, it is important to predict the severity of AP to decrease the morbidity and mortality. Objectives


Introduction
Acute pancreatitis (AP) is a disease with inflammation of the pancreas of various severity ranging from mild abdominal pain to mortality. 1 AP is characterized by 3 phases.In the 1 st phase, enzyme activation and cellular damage cause early symptoms.In the 2 nd phase, systemic inflammatory response and intrapancreatic inflammatory reaction occur by the release of proinflammatory and antiinflammatory mediators.In the 3 rd phase, complications of AP occur. 2 AP is classified as acute interstitial edematous pancreatitis (AEP) or acute necrotizing pancreatitis (ANP) according to the revised Atlanta classification. 3AEP constitutes 75-85% of patients with AP, while ANP consists of 15-25% of patients with AP.The mortality rate is 3% in AEP and 15% in ANP. 4 Since mortality is 5 times higher in ANP than in AEP, it is important to discriminate ANP from AEP to predict morbidity and mortality.Several scoring systems have been developed to detect the severity of AP.These scoring systems include the Ranson criteria, Acute Physiology and Chronic Health Evaluation II (APACHE II), systemic inflammatory response syndrome criteria, bedside index of severity in acute pancreatitis, harmless acute pancreatitis score, and Balthazar score. 5Recent studies have shown newer markers to detect AP severity.These include pancreatic protease activation peptides, interleukin 6 and interleukin 8, polymorphonuclear elastase, procalcitonin, mean platelet volume (MPV), and proteinuria. 6,7Red cell distribution width (RDW) is calculated by dividing the standard deviation of red blood cell volume by mean corpuscular volume (MCV) and multiplying it by 100 to express the results as percentages.RDW reflects the variability of the size of the circulating erythrocytes. 8Several studies have shown that RDW is significantly associated with inflammatory markers such as C-reactive protein and fibrinogen. 9,10In this study, we aimed to evaluate the RDW levels in patients with AEP and ANP.

Material and methods
The study included 180 patients who were admitted to the gastroenterology department of Adana Numune Research and Educational Hospital (Turkey) with a diagnosis of AP through the time frame of January 2014-May 2016.AP was diagnosed using the revised Atlanta classification. 3atients' age, sex, etiology of AP, and RDW were recorded.
Exclusion criteria comprised history of chronic pancreatitis or pancreas carcinoma, heart failure, hyperlipidemia, peripheral vascular disease, hematologic disorders, accompanying acute or chronic inflammatory diseases, any other accompanying carcinomas and chronic liver diseases.

Statistical analysis
Descriptive statistics were used to define the continuous variables and are expressed as mean ±standard deviation.Student's t-test was used for independent and normally distributed variables.Mann-Whitney U test was used for independent and not normally distributed variables.Receiver operating characteristic (ROC) curve analysis was performed for the variables found to be significant in univariate analysis to determine the cut-off point.Statistical significance level was determined as 0.05.The analysis was made using MedCalc Statistical Software v. 12.7.7 (MedCalc Software BVBA, Ostend, Belgium).

Results
A total of 180 patients were included in the study.Eighty patients (44%) were male and 100 patients (56%) were female.Mean age was 56.25 ±18.3 years (52.66 ±14.4 years for males; and 59.84 ±20.2 years for females).There was no statistically significant difference between the age of male and female patients.Patients were grouped into AEP and ANP.Patients' age was 53.75 ±13.3 years and 55.65 ±15.6 years in AEP and ANP groups, respectively.There was no statistically significant difference between patients' age.
Patient characteristics are shown in Table 1.The drug--induced AP among the patients was caused by azathioprine, furosemide, and thiazide type diuretics.Imaging studies (ultrasonography, computed tomography and magnetic resonance cholangiopancreatography) and laboratory values revealed no other etiologic factors in the idiopathic pancreatitis group.
One hundred forty-four patients (80%) were included in the AEP group, and 36 patients (20%) were included in the ANP group.There was a statistically significant difference in the RDW values between the groups (p = 0.011).There was no statistically significant difference between age, sex and etiologic factors in the patient groups.(Table 1).

Discussion
There has been research to evaluate the relationship between RDW and mortality in patients with AP before.In our study, we found a relationship between the disease severity and the RDW values in patients with AP.Earlier studies included a small number of patients, while our study had the largest patient group.
RDW is calculated as part of complete blood count.It is used to distinguish the etiology of anemia and it shows heterogeneity.Normal values for RDW are 11.5-14.5%.There is no case of RDW levels below normal limits.Thus, RDW is expressed as normal or high.Increased RDW shows a higher variation of erythrocyte size than normal.RDW is increased in hematologic and inflammatory diseases.The RDW value shows no sensitivity and specificity for a specific disease, and also does not exclude a diagnosis.RDW values are increased in hemolytic anemia, in the case of erythrocyte transfusion, and deficiency of vitamin B 12 , folic acid and iron. 9,11,124][15][16][17] Kurt et al. showed a correlation of nonvalvular atrial fibrillation and higher RDW values in 320 patients. 13emir et al. conducted a study of 37 patients with cerebral sinus thrombosis and 101 patients with primary headache with a total of 138 patients.In that study, increased RDW values were correlated with the existence of vertebral sinus thrombosis.They showed that increased RDW values might be a predictor of cerebral sinus thrombosis in patients with primary headache. 14In a study of 367 patients with grade I-V chronic renal failure, Solak et al. reported a relationship between increased RDW values and flow mediated dilatation, which shows endothelial dysfunction independent of anemia, diabetes mellitus and inflammation. 15Kim et al. showed the correlation between increased RDW values and advanced fibrosis in 24,547 patients with nonalcoholic liver disease. 16he inflammation status of the disease may change the RDW values.The association of RDW with mortality may be explained by inflammation.Inflammation promotes the death of RBCs or inhibits the maturation of RBCs, thus decreasing RBC lifespan.RDW values may reflect the inflammation status of acute pancreatitis and may be used to predict severe acute pancreatitis.Bone marrow function and iron metabolism may be influenced by inflammation, and inflammatory mediators suppress erythrocyte maturation and cause larger and younger reticulocytes to enter the circulation, thus increasing RDW.Inflammation also increases oxidative stress leading to elevated RDW by reducing RBC survival and increasing the release of large, premature RBCs into the circulation.Inflammation itself also alters RBC membrane structure, contributing to changes in RBC morphology. 17cute pancreatitis may cause local inflammation and also systemic effects.Mortality is higher in ANP than in AEP, so early diagnosis and treatment of ANP may improve the prognosis. 17At present, there is no single marker showing ANP. Biochemical tests, imaging studies and scoring systems are used in all clinics to detect the severity of AP, but 20-30% of severe AP may not be diagnosed with these tests. 18An ideal marker -objective, simple, cheap, reproducible, sensitive and specific to AP, should be available in all clinics. 174][15][16][17] Wang et al. showed higher mortality rates with RDW values > 13.4%. 17In our study, we found sensitivity of 88.2% and specificity of 91.8%, with a cut-off value of 14.35% in ROC analysis.In a study including 103 patients, Şenol et al. observed that high RDW values on admission were associated with mortality. 19They found a cut-off value of 14.8% and predicted mortality in 77% of cases.They reported a sensitivity of 47.6% and a specificity of 96.3%. 19Yao and Lv also reported a cut-off value of 14.2%, and found higher RDW values in terminal patients with a sensitivity of 75% and specificity of 89.8% in 106 patients. 20Our study included 180 patients with AP. 144 patients had AEP, and 36 patients had ANP.We found that high RDW values on admission to the hospital were associated with severe disease.We found 16.4 as the cut-off value of RDW.Sensitivity was 29.2% and specificity was 89.83% for this value.
In conclusion, RDW can effectively distinguish ANP from AEP, and be used to evaluate the prognosis of AP.It is simple, cheap and reproducible.

Fig. 1 .
Fig. 1.The receiver operating characteristic (ROC) curve of red cell distribution width (RDW) values for predicting acute necrotizing pancreatitis in patients with acute pancreatitis