Abstract
INTRODUCTION:
False-positive lipase elevation is a diagnostic challenge in acute pancreatitis (AP) but has not been systematically studied. We evaluated patients presenting with serum lipase elevation and correlated this with abdominal imaging findings, considered the gold standard for AP diagnosis.
METHODS:
Patients with abdominal pain, serum lipase ≥3-fold the upper limit of normal (ULN), and abdominal imaging within 48 hours of their lipase check were identified. An independent expert pancreas radiologist blinded to clinical data reviewed the images for features of AP. For patients without imaging changes of AP, repeat imaging obtained within 30 days was reviewed. Among patients with elevated lipase but no imaging changes of AP, other etiologies of lipase elevation were identified. Patients were stratified by the degree of lipase elevation, and imaging findings were compared between groups.
RESULTS:
Two hundred thirty-four patients had lipase ≥3-fold the ULN, abdominal pain, and underwent CT or MRI with intravenous contrast. 60.2% had AP findings on their initial imaging, and another 13% had AP or sequelae on imaging within 30 days. In patients without imaging findings of AP, the use of opioid analgesics was the most commonly suspected cause of lipase elevation (24.7%), followed by abdominal trauma/surgery (12.9%). Among lipase elevations, a cutoff of ≥6-fold the ULN had the highest accuracy (70.1%) for diagnosing AP.
DISCUSSION:
Lipase elevation without imaging findings of AP is found in nearly one-fourth of patients and may suggest non-AP causes of lipase elevation. This has clinical implications, but increasing lipase cutoff to ≥6-fold ULN has only a modest increase in diagnostic accuracy, suggesting the need for a better biomarker for AP diagnosis.
KEYWORDS: lipase, acute pancreatitis, diagnosis
INTRODUCTION
The diagnostic criteria for acute pancreatitis (AP) most widely used for clinical and research purposes were initially outlined in the Atlanta Classification of 1992 and remained unchanged in the revision in 2012. AP is diagnosed by the presence of 2 of the following 3: (i) abdominal pain consistent with AP, (ii) serum lipase or amylase levels ≥3-fold the upper limit of normal (ULN), and (iii) characteristic imaging findings of AP on contrast-enhanced CT (CE-CT) or MRI (1). However, it is increasingly being recognized that these diagnostic criteria have limitations. There is no “characteristic” abdominal pain for AP as it is a visceral disorder where pain can be diffuse, poorly localized, and often referred because of viscerosomatic sensory convergence at the spinal cord (2). The serum biomarkers for diagnosing AP, amylase, and lipase can be elevated because of a myriad of nonpancreatic causes (1). Although imaging is arguably the “gold standard” for diagnosing AP, it is not always obtained and can be normal if obtained shortly after the onset of abdominal pain. The impact of these limitations cannot be overstated, as they lead to unnecessary diagnostic studies and procedures (3).
In this study, we retrospectively evaluated the correlation of lipase elevation among patients with suspected AP and the radiographic findings. An independent radiologist assessed the most common imaging findings in patients with AP. Patients without AP on imaging were screened for other possible causes of serum lipase elevation.
METHODS
Between September 2019 and April 2020, patients with lipase greater than 3-fold the ULN were prospectively identified using a real-time notification system in the electronic medical system (Epic Systems). AP was diagnosed using cross-sectional imaging within 48 hours of the lipase check, which was considered the gold standard for AP diagnosis. CE-CT scan was the most frequently ordered imaging study, but if not obtained, MRI imaging with contrast was acceptable. All of the relevant imaging from the time of elevated lipase detection to 30 days afterward was reviewed by an independent expert pancreas radiologist who was blinded to the clinical and laboratory data. Patients were excluded from the study if the serum lipase was ordered for a reason other than abdominal pain (e.g., chemotherapy protocol, transplant workup, asymptomatic patients), if no imaging was performed, or if CT was performed without intravenous contrast. All images were reviewed on a diagnostic Picture Archiving and Communication System workstation (Centricity Picture Archiving and Communication System RA1000; GE Healthcare, Barrington, IL). The imaging features used in the diagnosis of AP included qualitative assessment of pancreatic gland edema or enlargement, presence of indistinct pancreatic margins and stranding, presence of a peripancreatic fluid collection, and pancreatic necrosis (4).
All images were evaluated by 2 independent radiologists who were aware of the study design and objectives but were blinded to all other clinic information. For patients without AP on initial imaging, records were reviewed for repeat imaging within 30 days to assess for imaging findings of AP or its complications. In patients with elevated lipase and without findings of AP on imaging, the medical record was reviewed for the most likely alternative etiology of lipase elevation (1). Patients with elevated serum lipase were stratified into groups based on a serum lipase cutoff value. The frequency of AP imaging features was compared between the groups using the χ2 test and Fisher exact test. Cohen kappa test was used to evaluate the agreement between the judgment of the 2 radiologists. A P value < of 0.05 was regarded as statistically significant. All statistical analyses were performed using IBM SPSS 26.0.
RESULTS
In total, 5,767 notifications of lipase greater than 3-fold the ULN were received. Five hundred fifty-seven unique patients were found to have lipase elevation. Among these, 395 patients had abdominal pain, and 295 patients underwent abdominal imaging (CE-CT/MRI [n = 234], noncontrast CT [n = 61]). Patients with noncontrast CT were excluded, and 234 patients were included in the study population. The demographics of these patients are described in Table 1. Of these patients, 141 (60.2%) were found to have imaging features of AP. Repeat scans were available for 23 of the 93 patients without AP findings on initial imaging. Among these patients, 3 (13%) developed AP imaging findings within 30 days of presentation.
Table 1.
Clinical characteristics and imaging studies of patients with lipase elevation ≥3-fold the upper limit of normal and abdominal pain (n = 395)
| Clinical characteristics and comorbidities | |
| Age (mean ± SD) | 50.7 ± 16.6 |
| Female sex (n, %) | 186 (47) |
| Diabetes (n, %) | 104 (26.3) |
| CKD of any stage (n, %) | 14 (2) |
| Chronic opioid use (n, %) | 37 (9.4) |
| Systemic inflammatory immune response (n, %) | 103 (26) |
| ICU admission (n, %) | 47 (12) |
| Abdominal imaging | |
| CE-CT or MRI of the abdomen (n, %) | 234 (75) |
| Noncontrast CT of the abdomen (n, %) | 61 (15.4) |
CKD, chronic kidney disease; CE-CT, contrast-enhanced computed tomography; ICU, intensive care unit.
Pancreatic edema/enlargement was seen in 109 (77.3%), indistinct pancreas margin/fat stranding was seen in 140 (99.2%), pancreatic/peripancreatic fluid collection was seen in 101 (71.6%), and pancreatic/peripancreatic necrosis was seen in 25 patients (17.7%, Table 2). There was discordance in 45 results (19.2%) among the 2 radiologists. Cohen kappa showed moderate agreement between the radiologist's judgments, κ = 0.609 (95% confidence interval [CI], 0.605–0.613), P < 0.001.
Table 2.
Imaging characteristics of patients with lipase elevation more than 3-fold the upper limit of normal with abdominal pain who underwent cross-sectional abdominal imaging (Contrast-enhanced CT/MRI) (n = 234)
| Imaging characteristics | |
| AP imaging features on radiologist review (n, %) | 141 (60.2) |
| Specific imaging features in patients with AP (n, %) | |
| Pancreatic edema/enlargement | 109 (77.3) |
| Indistinct pancreas margin/Stranding | 140 (99.2) |
| Pancreatic/peripancreatic fluid collection | 101 (71.6) |
| Pancreatic/peripancreatic necrosis | 25 (17.7) |
AP, acute pancreatitis; CE-CT, contrast-enhanced computed tomography.
On chart review of the patients with elevated lipase and without evidence of AP, the use of opioid analgesics (24.7%) was the most common possible cause, followed by a history of recent abdominal surgery or trauma (Table 3). The accuracy of lipase as a diagnostic test was measured at different cutoff values, considering the presence of AP imaging findings as the gold standard (Table 4, Figure 1). The greatest sensitivity was found at a lipase cutoff value ≥ 4-fold the ULN (90.8% [53.7%–95%]), and the greatest specificity was found at a lipase cutoff value ≥10-fold the ULN (77.4% [67.6%–85.4%]). A cutoff value of lipase ≥6-fold the ULN was determined to have the highest accuracy for diagnosing AP (70.1% [63.7%–75.8%]). Among patients with lipase elevation between 3-fold and 6-fold the ULN, 35.4% had AP findings on imaging, compared with 72.9% of patients with lipase ≥6-fold ULN.
Table 3.
Etiologies of lipase elevation in patients without AP findings on imaging (n = 93)
| Cause of lipase elevation | Frequency (%) |
| Opioid analgesics for abdominal pain | 23 (24.7) |
| Abdominal surgery or trauma | 12 (12.9) |
| Renal impairment or nephrolithiasis | 11 (11.8) |
| Choledocholithiasis or biliary obstruction | 11 (11.8) |
| Pancreatic or ampullary malignancy | 7 (7.5) |
| Gastritis or enteritis | 6 (6.5) |
| Post-ERCP abdominal pain (without pancreatitis) | 6 (6.5) |
| Cholangitis or cholecystitis | 4 (4.3) |
| Extrapancreatic malignancy | 4 (4.3) |
| Hepatitis or cirrhosis | 3 (3.2) |
| No etiology identified | 2 (2.1) |
| Medications | 2 (2.1) |
| Uncontrolled diabetes or ketoacidosis | 1 (1) |
| Ascites | 1 (1) |
The one most likely etiology is reported per patient.
AP, acute pancreatitis; ERCP, endoscopic retrograde cholangiopancreatography.
Table 4.
Sensitivity, specificity, and accuracy of lipase as a diagnostic test for AP at different cutoff levels
| Lipase cutoff value | Sensitivity | Specificity | Accuracy |
| Lipase ≥ 4-fold ULN | 90.8% (53.7%–95%) | 28% (19.1%–38.2%) | 65.6% (59.3%–71.9%) |
| Lipase ≥ 5-fold ULN | 84.4% (77.3%–90%) | 47.3% (36.9%–057.9%) | 69.6% (63.3%–75.5%) |
| Lipase ≥ 6-fold ULN | 80.1% (72.6%–86.4%) | 54.8% (44.2%–65.2%) | 70.1% (63.7%–75.8%) |
| Lipase ≥ 7-fold ULN | 75.2% (67.2%–82.1%) | 60.2% (49.5%–70.2%) | 69.2% (62.9%–75.1%) |
| Lipase ≥ 8-fold ULN | 67.4% (59%–75%) | 71% (60.6%–79.9%) | 68.8% (62.4%–74.7%) |
| Lipase ≥ 9-fold ULN | 62.4% (53.9%–70.4%) | 75.3% (65.2%–83.6%) | 67.5% (61.1%–73.4%) |
| Lipase ≥ 10-fold ULN | 61.7% (53.1%–69.8%) | 77.4% (67.6%–85.4%) | 67.9% (61.5%–73.9%) |
Bold indicates best performing values.
AP, acute pancreatitis; ULN, upper limit of normal.
Figure 1.
Sensitivity, specificity, and accuracy of lipase as a diagnostic test of AP at different cutoff values. AP, acute pancreatitis; ULN, upper limit of normal.
DISCUSSION
In this retrospective cohort study of 234 patients who presented with abdominal pain and serum lipase ≥ 3-fold ULN who underwent CE-CT or MRI, we found that approximately 60.2% of patients had AP findings on initial imaging, with an additional 13% of patients having AP findings on imaging within 30 days of presentation. Approximately one-third of patients had abdominal pain and elevated serum lipase likely due to a cause other than AP.
Diagnosis of AP using the Revised Atlanta Classification is challenging as abdominal pain is a nonspecific symptom, lipase quantification assays suffer from low specificity, and elevated lipase may not reflect pancreatic pathology. A chief complaint of abdominal pain has a vast differential diagnosis. Recent surveys show that up to 21.8% of patients reported abdominal pain in the past month and that 40.3% of patients suffer from a functional gastrointestinal disorder (5,6). The most common methods of lipase quantification are titrimetric and colorimetric assays, which suffer from low accuracy (7,8). Serum lipase, the most commonly used biomarker to diagnose AP, is elevated in many nonpancreatic conditions that cause abdominal pain and thus does not always correlate with pancreatic pathology on imaging (1). Decompensated cirrhosis, renal failure, opioid analgesic use, acute illness, biliary obstruction, and recent abdominal trauma are the common causes noted in previous literature (1,9). In our study, opioid analgesic use was the most common cause, followed by abdominal trauma or surgery. Although the mechanism for opioid-induced lipase elevation is not completely understood, previous studies have suggested that it may result from decreased gastrointestinal motility and increased smooth muscle tone, leading to pancreatic duct sphincter spasm (10–12). In the Nardi test for sphincter of Oddi dysfunction, Lobo et al (12) demonstrated that administration of 10 mg of morphine in healthy volunteers led up to a 4-fold increase in serum lipase levels, even in the absence of reproducible abdominal pain. Therefore, opioid use can confound the interpretation of pancreatic enzyme levels, and in patients with chronic abdominal pain, lipase elevation may reflect the opioid effect rather than underlying organic pathology (3). An incorrect diagnosis of AP can have significant clinical and research implications. From a direct patient care perspective, incorrect diagnosis of AP may result in unnecessary discontinuation of medications such as checkpoint inhibitors in patients with cancer and immunosuppressives in those with inflammatory bowel disease. On the other hand, it can also lead to unnecessary dietary restriction and the initiation and continuation of drugs such as opioid analgesics for pain and pancrelipase enzyme supplements. Furthermore, this may result in unwarranted procedures such as endoscopic retrograde cholangiopancreatography, cholecystectomy, and more invasive pancreatic drainage and resection operations. From a research standpoint, misclassification inflates the reported incidence of AP and can lead to the inclusion of patients without true pancreatic inflammation in clinical trials. This not only dilutes the efficacy signals of therapeutic interventions but also risks falsely attributing adverse drug reactions to AP based on nonspecific enzyme elevations (13). The net result is increased medical expenditure, use of medical resources, morbidity and mortality, and patients forgoing essential medical therapies (14). Adopting a higher diagnostic threshold for AP—ensuring that diagnostic criteria such as imaging confirmation or characteristic clinical presentation are met in addition to enzyme elevation—could substantially reduce these risks.
We also showed that using a serum lipase cutoff ≥6-fold the ULN for AP diagnosis has a better correlation with radiographic findings of AP compared with using serum lipase levels ≥4-fold the ULN. Specifically, using a cutoff of ≥6-fold the ULN results in 79% sensitivity, 52% specificity, and 67.4% accuracy for an AP diagnosis compared with using a cutoff of ≥4-fold the ULN, which has 91% sensitivity, 26.6% specificity, and 64% accuracy for an AP diagnosis (P value < 0.00001). Interestingly, another study found that 16% of patients with lipase <3-fold the ULN had imaging features of AP compared with 34% with lipase ≥10-fold the ULN (P = 0.0042) (15). Both studies suggest that a higher serum lipase cutoff value may be helpful to improve the accuracy in the diagnosis of AP but is still limited.
Future research should quantitatively explore the impact of AP misdiagnosis, which may lead to increased research and motivation to improve the diagnostic criteria. Further studies should explore the role of imaging modalities in AP diagnosis, characterizing the most specific imaging findings reflecting true AP and assessing interobserver variability among radiologists. Numerous studies have examined the advantages and disadvantages of ultrasound, CT, MRI, and magnetic resonance cholangiopancreatograhy. Notably, MRI is advantageous in identifying morphology and signal changes in most patients with pancreatic edema compared with CT and is also superior for describing the extent of edema and inflammatory changes in AP (16). The sensitivity of imaging modalities for diagnosing AP varies. According to a systematic review and meta-analysis, CT has a sensitivity of approximately 73% and MRI has a sensitivity of 92% (17,18). Ultrasound is less sensitive than CT and MRI, with one study showing a sensitivity of 52% for diagnosing AP (19). Biomarkers such as amylase have also been extensively studied, and current evidence suggests that while amylase can be useful, it is not a preferred biomarker. Serum amylase may remain within the normal range in up to 20% of patients with AP, particularly in patients with alcohol-induced pancreatitis. Similar to lipase, elevated amylase can be seen in several nonpancreatic conditions, reducing its specificity (3,20). There have been attempts to develop additional specific biomarkers in the diagnostic evaluation of AP, such as pancreatic elastase and serum trypsin (21). While some of these biomarkers show promise, the newer biomarkers need to be evaluated and validated in large sample studies.
Our study adds to the literature because it has a rigorous follow-up with every abdominal imaging study reviewed by an expert pancreas radiologist over a 30-day period. Although it is possible that imaging studies may not capture AP features if performed too early, there was at least a 24-hour interval between the onset of symptoms and the imaging among our patient cohort. Another limitation is that only around 25% of patients without findings of AP on imaging underwent another contrast-enhanced abdominal imaging within 30 days. However, delayed findings of AP were identified in only 13% of the patients and would not significantly affect the study results. Another limitation is that CE-CT, while considered the gold standard for identifying AP features, may be limited to capturing subtle changes in AP compared with MRI. In addition, the retrospective data collection and limited duration of the study inherently limit its generalizability. Finally, patients with lipase elevations <3-fold the ULN were excluded, precluding analysis of a subgroup that may present solely with abdominal pain and radiographic evidence of AP. Overall, this study shows that among patients with elevated lipase and abdominal pain who underwent cross-sectional imaging, there is only modestly improved accuracy in diagnosing AP on imaging using a higher serum lipase. The current definition of AP may lead to an increased incidence of false-positive diagnoses and affect patient care. Future diagnostic criteria for AP may use a higher serum lipase cutoff, identify better biomarkers, and place a higher weight on imaging findings in the diagnosis, especially among patients without relevant risk factors and clinical history.
CONFLICTS OF INTEREST
Guarantor of the article: Venkata S. Akshintala, MD.
Specific author contributions: V.S.A.: project conceptualization, review, co-authored. R.S.: data collection, analysis, review, co-authored; A.S.: data collection, analysis, review, co-authored; A.Z.: radiographic analysis, review, co-authored; F.B.: review, co-authored; A.K.: review, co-authored; P.J.L.: review, co-authored; M.A.A.H.: review, co-authored; M.F.: review, co-authored; E.A.: review, co-authored; V.K.S.: project conceptualization, review, co-authored.
Financial support: None to report.
Potential competing interests: V.A. is co-founder of Origin Endoscopy Inc., Solv Endotherapy Inc and consultant to Olympus Endoscopy, Amgen. Vikesh Singh is a consultant to, Amgen, Horizon Therapeutics, and Panafina. He is on the advisory board and an equity holder in Origin Endoscopy Inc. and Kyttaro. All other authors declare no competing interests.
Study Highlights.
WHAT IS KNOWN
✓ Acute pancreatitis is challenging to diagnose.
✓ There are nonpancreatic causes of lipase elevation.
WHAT IS NEW HERE
✓ We correlate imaging findings with degree of lipase elevation.
✓ We describe nonpancreatic causes of lipase elevation.
✓ We calculate the accuracy of different lipase cutoffs in diagnosing acute pancreatitis.
ABBREVIATIONS:
- AP
acute pancreatitis
- CE-CT
contrast-enhanced computed tomography
- ULN
upper limit of normal
Contributor Information
Rishi Subrahmanyan, Email: rsubrah1@jh.edu.
Anmol Singh, Email: drsinghanmol@gmail.com.
Atif Zaheer, Email: azaheer1@jhmi.edu.
Furqan Bhullar, Email: fbhulia1@jhmi.edu.
Ayesha Kamal, Email: akamal3@jh.edu.
Peter J. Lee, Email: peter.lee@osumc.edu.
Maisam A. Abu El Haija, Email: abuelhmm@ucmail.uc.edu.
Mahya Faghih, Email: mfaghih2@jh.edu.
Elham Afghani, Email: eafghan1@jhmi.edu.
Vikesh K. Singh, Email: vsingh1@jhmi.edu.
REFERENCES
- 1.Hameed AM, Lam VW, Pleass HC. Significant elevations of serum lipase not caused by pancreatitis: A systematic review. HPB (Oxford) 2015;17(2):99–112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cervero F, Laird JM. Visceral pain. Lancet 1999;353(9170):2145–8. [DOI] [PubMed] [Google Scholar]
- 3.Bush N, Akshintala VS. Interpretation of serum pancreatic enzymes in pancreatic and nonpancreatic conditions. Curr Opin Gastroenterol 2023;39(5):403–10. [DOI] [PubMed] [Google Scholar]
- 4.Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis—2012: Revision of the Atlanta classification and definitions by international consensus. Gut 2013;62(1):102–11. [DOI] [PubMed] [Google Scholar]
- 5.Sandler RS, Stewart WF, Liberman JN, et al. Abdominal pain, bloating, and diarrhea in the United States: Prevalence and impact. Dig Dis Sci 2000;45(6):1166–71. [DOI] [PubMed] [Google Scholar]
- 6.Sperber AD, Bangdiwala SI, Drossman DA, et al. Worldwide prevalence and burden of functional gastrointestinal disorders, results of Rome Foundation Global study. Gastroenterology 2021;160(1):99–114.e3. [DOI] [PubMed] [Google Scholar]
- 7.Guo J, Chen CP, Wang SG, et al. A convenient test for lipase activity in aqueous-based solutions. Enzyme Microb Technol 2015;71:8–12. [DOI] [PubMed] [Google Scholar]
- 8.Hasan F, Shah AA, Hameed A. Methods for detection and characterization of lipases: A comprehensive review. Biotechnol Adv 2009;27(6):782–98. [DOI] [PubMed] [Google Scholar]
- 9.Da BL, Shulman IA, Joy Lane C, et al. Origin, presentation, and clinical course of nonpancreatic hyperlipasemia. Pancreas 2016;45(6):846–9. [DOI] [PubMed] [Google Scholar]
- 10.Myhre J, Nesbitt S, Hurly JT. Response of serum amylase and lipase to pancreatic stimulation as a test of pancreatic function; the mecholyl-secretin and the morphine-secretin tests. Gastroenterology 1949;13(2):127–34. [PubMed] [Google Scholar]
- 11.Burke JO, Plummer K, Bradford S. Serum amylase response to morphine, mecholyl and secretin as a test of pancreatic function. Gastroenterology 1950;15(4):699–707. [PubMed] [Google Scholar]
- 12.Lobo DN, Takhar AS, Thaper A, et al. The morphine-prostigmine provocation (Nardi) test for sphincter of Oddi dysfunction: Results in healthy volunteers and in patients before and after transduodenal sphincteroplasty and transampullary septectomy. Gut 2007;56(10):1472–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lando HM, Alattar M, Dua AP. Elevated amylase and lipase levels in patients using glucagonlike peptide-1 receptor agonists or dipeptidyl-peptidase-4 inhibitors in the outpatient setting. Endocr Pract. 2012;18(4):472–7. [DOI] [PubMed] [Google Scholar]
- 14.El Halabi M, Bou Daher H, Rustom LBO, et al. Clinical utility and economic burden of routine serum lipase determination in the Emergency Department. Int J Clin Pract 2019;73(12):e13409. [DOI] [PubMed] [Google Scholar]
- 15.Glazer DI, Cochon LR, Raja AS, et al. Prevalence of imaging findings of acute pancreatitis in emergency department patients with elevated serum lipase. Am J Emerg Med 2021;50:10–3. [DOI] [PubMed] [Google Scholar]
- 16.Xiao B, Xu HB, Jiang ZQ, et al. Current concepts for the diagnosis of acute pancreatitis by multiparametric magnetic resonance imaging. Quant Imaging Med Surg 2019;9(12):1973–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sun H, Jian S, Peng B, et al. Comparison of magnetic resonance imaging and computed tomography in the diagnosis of acute pancreatitis: A systematic review and meta-analysis of diagnostic test accuracy studies. Ann Transl Med 2022;10(7):410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Arvanitakis M, Delhaye M, De Maertelaere V, et al. Computed tomography and magnetic resonance imaging in the assessment of acute pancreatitis. Gastroenterology 2004;126(3):715–23. [DOI] [PubMed] [Google Scholar]
- 19.Lawson TL. Sensitivity of pancreatic ultrasonography in the detection of pancreatic disease. Radiology 1978;128(3):733–6. [DOI] [PubMed] [Google Scholar]
- 20.Tenner S, Baillie J, DeWitt J, et al. , American College of Gastroenterology. American College of Gastroenterology guideline: Management of acute pancreatitis. Am J Gastroenterol. 2013;108(9):1400–16; 1416. [DOI] [PubMed] [Google Scholar]
- 21.Meher S, Mishra TS, Sasmal PK, et al. Role of biomarkers in diagnosis and prognostic evaluation of acute pancreatitis. J Biomark 2015;2015:519534. [DOI] [PMC free article] [PubMed] [Google Scholar]