Three-hour post-ERCP amylase level: a useful indicator for early prediction of PEP

Background : To evaluate the value of the 3-hour post-ERCP serum amylase level for early prediction of post-ERCP pancreatitis (PEP). Method: A study of 206 patients performed ERCP at a single centre was collected from 2011 to 2016. The serum amylase or lipase level was measured at 3 h after ERCP. The patients with PEP were recorded. ROC curves were used to statistically analyze the data: The enrolled patients were divided into two groups according to gender, then be analyzed respectively. We comprehensively evaluated the predictive value of PEP by serum amylase level 3-hour post-ERCP based on the results above. Results : In the 206 patients, the 3-hour post-ERCP pancreatic amylase level was used as the test variable, and the PEP occurrence as the state variable to plot the ROC curve. The optimal cut-off value was 351U/L (sensitivity 76.19%, specificity 83.24%, positive likelihood ratio 4.55, negative likelihood ratio 0.29, Youden index 59.43%). There were 83 patients with both 3-hour post-ERCP amylase level and lipase level detected. The area under the ROC curve for the 3-hour post-ERCP serum amylase was 0.780, and the 95% confidence interval was 0.676-0.864. The optimal cut-off is 380U/L, and there was no statistically significant difference between the two for diagnostic accuracy. According to gender, there was no statistically significant difference in the diagnostic accuracy of the two groups. In the male group, 436 U/L serum amylase provided the greatest diagnostic accuracy with sensitivity(SE) of 70.5%, specificity(SP) of 89.2%, positive predictive value (PPV) 87.5%, and negative predictive value (NPV) 78.1%. Whereas, in the female group, 357U/L serum amylase provided the greatest diagnostic accuracy with sensitivity of 76.9%, specificity of 81.2%, positive predictive value of 80.4%, negative predictive value of 77.9%.

predicting post-ERCP pancreatitis. 2. There was no significant difference between serum amylase and lipase 3-hour post-ERCP for predicting PEP. 3. There was no statistically significant difference between male and female using the 3-hour post-ERCP serum amylase level to prediction PEP . For female, the optimal cut-off value was 357 U/L, whereas male 436U/L .

Background
Since first reported by McCunne in 1968(1),ERCP has been widely used in diagnosis and treatment of pancreatic and biliary diseases, especially in the treatment of common bile duct stones. Compared to traditional surgery, it is safer, more effective, with less damage and shorter hospitalization, benefits when repeated performance for stone extraction is required (2). However, its postoperative complications are sometimes difficult to avoid (3).
Post-ERCP pancreatitis (PEP) is one of the most common complications after ERCP procedure. It is reported that the incidence of PEP is 1.6-15.7% (4). Although 90% are mild or moderate (5), 1% may develop acute necrotizing pancreatitis (6), so early diagnosis and timely treatment is very important (7,8). So far, the determination of serum amylase is still the most commonly used indicator for the diagnosis of PEP, but there is still no clear standard for the early prediction of PEP at different time points and different levels of amylase (9,10). In this study, the ROC curve of the diagnostic test was used to evaluate the value of early prediction of PEP by 3-hour post-ERCP serum amylase level. The inclusion and exclusion criteria were as follows:

Methods
Inclusion criteria: (1) preoperative serum amylase and lipase levels were normal; (2) age ≥18 years; (3) serum amylase level was measured 3 hours after ERCP; (4) patients have not used trypsin inhibitor before the diagnosis of PEP.
Exclusion criteria: (1) preoperative diagnosis of acute and chronic pancreatitis (2) abnormal renal function (serum creatinine > 92umol / L) (3) pregnant women According to the inclusion and exclusion criteria above, a total of 206 patients were enrolled, including 92 males and 114 females, among them, 84 cases were simultaneously detected for serum lipase.
The diagnosis of pancreatitis after ERCP is based on the consensus reached by Cotton et al (11), and the international consensus on the classification of acute pancreatitis in Atlanta in 2012(12).

1.
Acute pancreatic abdominal pain within 24 hours after ERCP; 2. Serum amylase more than 3 times the upper limit of normal within 72 hours after ERCP(normal value is 30-110U/L) or lipase greater than 3 times the upper limit of normal within 96h after ERCP(normal reference range 23-300U/L);

3.
Contrast-enhanced CT, MRI, abdominal ultrasound showing acute pancreatitis changes (pancreatic enlargement, exudation, necrosis and other AP signs); having two of the three criteria will lead to a diagnosis of PEP.

Equipment Description
Duodenoscope JF-260, Olympus Optical Corporation, Tokyo, Japan), triple lumen sphincterotome (Papillotome, ENDO-FLEX GmbH, Germany), guide wire (Hydra Jagwire  The non-parametric method was used in this paper, the serum amylase level 3-hour post-ERCP as the test variable, the PEP occurrence as the state variable and the ROC curve was drawn. The result is shown in Fig. 4 (the solid line is the drawn ROC curve, and the dotted line is 95% confidence interval), In tables 1, the area under the ROC curve was 0.816, the diagnostic value is good, the standard error was 0.0507, the 95% confidence interval was 0.756-0.866, and the significance test Z value was 6.235, P<0.0001. The area under the ROC curve was statistically significant, and the optimal cut-off value was 351 U/L. The sensitivity, specificity, positive and negative likelihood ratio, positive and negative predictive value, and Youden index of different cut-off values are shown in Table 2: As seen from the table 3, when 351U/L was taken as the optimal cut-off value, a better positive predictive value (81.90) could be obtained, the Youden index was the highest In order to investigate whether different genders groups need different diagnostic criteria, ROC curve was applied to evaluate this issue in our study. 206 patients were divided into two groups according to gender, and the ROC curves were plotted respectively ( Fig. 5, male; Fig. 6, female). The results are shown as follows: The ROC curves drawn for the two groups both had statistical significance (P = 0.001 for both groups), the area under the curve for the male group was 0.859, the standard error was 0.065, and the area under the curve for the female group was 0.794, and the standard error was 0.068 ( of 77.9%. The NPV in the male group was higher than that of the female group, which indicated that the proportion of PEP in the male group was also higher than that of the female group. This was consistent with the objective situation that women have a higher incidence of PEP than men.

Matched-pair analysis of the 3-hour post-ERCP serum amylase and lipase levels
The ROC curve matched-pair testing was used to compare the serum amylase and serum lipase 3 hours after ERCP. A total of 83 patients with both the serum amylase and lipase levels 3-hour post-ERCP were selected. Among them, 15 patients were diagnosed as PEP.
The ROC curves were drawn (Fig. 7) and the results were compared as follows: The ROC curves of the two groups were statistically significant (P < 0.001). The area under the ROC curve of lipase levels 3-hour post-ERCP was 0.778, with a 95% confidence interval of 0.673-0.862; The area under the ROC curve of amylase 3-hour post-ERCP was 0.780, and the 95% confidence interval was 0.676-0.864, indicating that the accuracy of predicting PEP was high (Table 7). Matched-pair testing was performed between the two groups, and the results showed that there was no statistically significant difference in the accuracy of predicting PEP (Z score 0.0598, P = 0.9523, P>0.05) ( Table 8), and the negative predictive value of lipase 3-hour post-ERCP was not high (0.688) ( Table 9).

Discussion
ERCP has become an important method in clinical diagnosis and treatment for biliary and pancreatic diseases. However, its complications should not be ignored, especially Post-ERCP pancreatitis (PEP). As the most common post-ERCP complication, PEP has restricted the development of ERCP. The average incidence of PEP is reported to be 3.5% in the literatures, but in high-risk groups (such as previous PEP, difficulty in cannulation of the bile duct, etc.), the incidence can be as high as 30% (5). Although most cases of PEP were mild(5), they recovered after conservative treatment in less than 72 h, in some severe cases, the length of hospital stay can be extended, the costs can be increased, multiple organ failure may occur, even death (6). Since it is difficult to avoid PEP even if the procedure is performed by an endoscopic expert, early detection and timely treatment for the PEP are very important. Serum amylase measurement technology is simple, easy to obtain for low price with high sensitivity and early appearance [15]. Although the level of amylase cannot estimate the severity of pancreatitis, it is still one of the most commonly used indicators for diagnosis of PEP currently. However, there is still no clear criteria for various cut-off serum amylase levels at different time points in prediction and diagnosis of PEP (7,9,(13)(14)(15)(16).
The role of serum amylase levels in predicting the occurrence of PEP after ERCP has been recognized (17). Takayoshi Nishino (14)  Our study showed that the serum amylase level 3-hour after ERCP has a good accuracy in the prediction of PEP. The cut-off at 351U/L provides the optimal diagnostic accuracy, when setting the cut-off at 105U / L (close to the standard upper limit of normal), the negative likelihood ratio can reach 0.21, indicating that this is a valuable predictive value for eliminating PEP.
Studies have confirmed that female gender is one of the important risk factors for pancreatitis after ERCP (20), so whether different genders need to use different predictive diagnostic criteria is one of the purposes of this study. By using the ROC curve matchedpair testing, the results showed that the two groups have the best diagnostic accuracy at different cut-off values. For female, the optimal cut-off value was 357U/L, whereas male 436U/L. This conclusion may provide help for clinicians. In addition, nearly 25% of patients with type 2 diabetes have elevated levels of lipase and/or amylase without the symptoms of acute pancreatitis (21). Whether we need to establish different cut-off values for the diagnosis of PEP to obtain a better predictive effect for those patients, is still to be validated by more clinical studies.
Serum lipase is a specific diagnostic indicator of pancreatitis. Its activity is maintained longer than amylase, and there is no other source in the blood (22), so it is considered to be more sensitive than serum amylase (95%;79%). The British guidelines for the diagnosis and treatment of acute pancreatitis in 2005 indicated that the pancreas is the only source of lipase and is superior to amylase in terms of sensitivity, specificity and accuracy (23).
And the American guidelines for the diagnosis and treatment of acute pancreatitis in 2006 also had similar views (24). The European Guidelines for Digestive Endoscopy in 2014 clearly indicated that serum amylase <1.5*ULN or serum lipase<4*ULN monitored 2-4 hours after ERCP provide a high NPV for PEP, and the patients can be discharged without considering the risk of PEP (recommended level B) (17), which suggests that serum lipase has a high negative predictive value for PEP. Gumaste et al. (25) pointed out based on an analysis of 170 patients with abdominal pain, the negative predictive value for acute pancreatitis was 98%. However, we emphasized the early (3-hour) lipase and amylase levels after ERCP in this study. And the results showed that in terms of comparison the predictive value for PEP between serum lipase and amylase 3-hour post-ERCP, lipase was not better than amylase. It is well known that serum lipase starts to rise at 24-48h and reaches a peak at 72-96 hours and last for 7-10d. This seems to explain why the early lipase monitoring value is not significantly better than that of amylase. Based on the actual situations in China, the predictive value of the joint monitoring needs to be further verified, and the economic benefits of the two tests need further cost-benefit analysis.

Conclusions
In summary, we have come to the following conclusions: The 3-hour post-ERCP serum amylase level is a useful measurement for early predicting post-ERCP pancreatitis. With the 3-hour post-ERCP serum amylase level 1-1.5 times upper limit of normal, the negative predictive value is high.
There was no significant difference between serum amylase and lipase 3-hour post-ERCP for predicting PEP.
There was no statistical difference between male and female using the 3-hour post-ERCP serum amylase level for prediction PEP. For female, the optimal cut-off value was 357U/L, whereas male 436U/L.          Figure 1 Descriptive statistics Figure 2 The positively skew distribution of the serum amylase levels 3-hour after ERCP in the non-PEP group The positively skew distribution of the serum amylase levels 3-hour after ERCP in the PEP group