What Is Known
- The course of pediatric acute pancreatitis is poorly understood.
- Up to 20% of children with acute pancreatitis will develop local and systemic complications, and up to 15% will develop recurrent episodes of acute pancreatitis.
- Prognostication models in children perform poorly and lack consensus.
What Is New
- Age is the only parameter predicting acute pancreatitis severity in children.
- None of the parameters predicted future development of acute pancreatitis recurrence.
Acute pancreatitis (AP) is characterized by pancreatic acinar cells injury, spillage of proteases with resultant inflammation and further pancreatic damage (1). AP incidence increased over the last 2 decades reaching an estimated annual incidence of 13 per 100,000 children (2–4). AP in children usually follows a mild course and resolves with conservative management for complete recovery. Up to 20% of patients may, however, develop local or systemic complications with risk of systemic inflammatory response syndrome, organ failure, and rarely death; overall mortality was reported to range between none and 11% (5–9).
Early prediction of severity is important to ensure appropriate treatment for those patients at risk of complications. Several prognostication models were developed and validated, and were demonstrated to predict mortality in adults (10–13). Nevertheless, in children these scores performed poorly with low sensitivity and/or specificity (14–17). Multiple studies in children reported on the association of specific laboratory test results on admission and pediatric AP severity including white blood cell count (WBC), lactate dehydrogenase (LDH), calcium, albumin, blood urea nitrogen (BUN), and lipase; and while specific pediatric scores were formulated based on these values, with or without additional clinical data, performance results are conflicting and there is no consensus regarding their prediction capacity (15–25).
Furthermore, a significant portion of patients with a primary episode of AP will continue to have acute recurrent episodes of pancreatitis (ARP). In children, ARP was reported to develop in 5.7% to 15.3% of patients presenting with AP and was associated with pancreatic anatomic abnormalities, metabolic disorders, and genetic predisposition (26–29).
This study assessed the utility of commonly available demographic data, laboratory results upon admission, and clinical parameters as predictors of AP severity and the risk of ARP following a first episode of AP. We also specifically subanalyzed patients with a first episode of AP.
PATIENTS AND METHODS
We conducted a retrospective study of children with AP, hospitalized at the Dana-Dwek Children's Hospital, in The Tel-Aviv Medical Center, a tertiary pediatric medical center, between January 1995 and June 2016. Patients were identified through the local electronic health records database according to discharge codes: “pancreatitis,” “acute pancreatitis,” and “hyper-amylasemia.” We included all children (younger than 18 years) diagnosed with AP, with a minimal follow-up of 6 months. The clinical diagnosis of AP was validated, based on patient charts, according to the 2012 Revised Atlanta Criteria, a consensus reached by the Acute Pancreatitis Classification Working Group; namely, fulfilling 2 of the following 3 criteria: typical abdominal pain, elevated serum pancreatic enzyme (lipase or amylase) level >3× upper limits of normal (ULN), and characteristic findings on imaging studies (30). Exclusion criteria included patients with missing data and chronic pancreatitis, defined as abdominal pain consistent with pancreatic origin or evidence of exocrine or endocrine insufficiency and imaging findings suggestive of chronic pancreatic damage (31); special attention was given while reviewing charts of patients with recurrent episodes of pancreatitis for evidence of chronicity. The study was reviewed and approved by the local ethics committee.
A recurrent episode was defined by a ≥1-month pain-free interval or normalization of serum pancreatic enzyme levels with resolution of pain between episodes (31). Pancreatitis severity was classified, according to the 2012 Revised Atlanta Criteria, as severe—characterized by persistent (>48 hours) organ (cardiovascular/respiratory/renal) failure (defined by a score of ≥2 using the modified Marshall scoring system (32), or otherwise clinically diagnosed); moderate—transient organ failure and/or local or systemic complications; and mild—pancreatitis without complications or organ failure (30).
Collected data included patient demographics, medical history, clinical course, blood test results on admission, and imaging studies. Weight was recorded as percentiles for sex and age according to the CDC growth charts (https://www.cdc.gov/growthcharts/clinical_charts.htm). The hospital pain management policy follows the World Health Organization recommendations, namely use of nonopioids (paracetamol, ibuprofen, diclofenac, and dipyrone) for mild pain, mild opioid (tramadol) for moderate pain, and strong opioids (morphine, oxycodone, buprenorphine, and pethidine) for moderate to severe pain. Analgesics were recorded by class—nonopioids, mild, and strong opioids (http://apps.who.int/iris/bitstream/10665/44540/1/9789241548120_Guidelines.pdf).
Statistical analysis was computed using SAS/STAT software V. 9.4 (SAS Institute Inc, Cary, NC). Univariate analysis was used to determine the relationships between pancreatitis severity and demographic and clinical factors. Continuous variables that followed a normal distribution were reported by mean and standard deviation and compared using 2-sample t test; continuous variables that did not follow a normal distribution were reported by median and interquartile range and compared using the 2-sample Wilcoxon test. Categorical variables were reported by their relative frequencies and compared by the Pearson χ2-test or Fisher exact test; all tests were done using 2-sided tails with a confidence interval (CI) of 95%. The association between continuous variables was evaluated by Pearson correlation coefficient. Kaplan-Meier method was used to compute product-limit estimates for event-free AP recurrences.
Seventy-four children hospitalized with an admission diagnosis of AP during the study period were identified and screened. Three patients were excluded for a final diagnosis of chronic pancreatitis and 3 for missing data. Thus, the study cohort included 68 patients accounting for 117 separate hospitalizations for AP (supplementary Fig. 1, Supplemental Digital Content 1, http://links.lww.com/MPG/B590).
Sixty (51.3%) episodes were diagnosed based on 2 and 57 (48.7%) on 3 of the Atlanta criteria. Nearly all patients were admitted to the pediatric or pediatric-surgical wards; 2 patients (2.9%) were admitted to the intensive care unit. While most episodes, 103 (88.0%), were treated conservatively with bowel rest, analgesics, and nutritional support, in 14 (12.0%) episodes patients underwent therapeutic interventions including surgery (1), computed tomography (CT)-guided (1) and endoscopic (1) fluid collection drainage, endoscopic retrograde cholangiopancreatography (ERCP) (4) and ERCP with stenting (3 following stone removal and 4 for stenotic bile and pancreatic ducts). A presumed etiologic diagnosis was established in 49 (41.9%) episodes, 68 episodes remained idiopathic (Table 1). Overall prognosis was excellent, and all patients recovered completely without sequela.
Mild Versus Moderate-severe Disease
In 15 (12.8%) episodes patients developed local complications including acute peripancreatic fluid collections (14, 12.0%), acute necrotic collections (2, 1.7%), and portal vein thrombosis (1, 0.9%). These complications were evident on 10 of 12 transabdominal ultrasound (US) examinations, 8 of 8 CT scans, and 2 of 2 of magnetic resonance imaging scans; 1 patient with pancreatic necrosis and 1 with peripancreatic fluid collection were diagnosed only by CT and magnetic resonance imaging, respectively, following unalarming US examinations. Two of these patients developed systemic inflammatory response and required oxygen supplementation and inotropic support. Subsequently, 2 patients (1.7%) were classified as severe pancreatitis, 13 (10.2%) as moderate and the remaining as mild pancreatitis. For characterization and analysis purposes, we combined the severe and moderate episodes for a total of 15 moderate-severe episodes.
Patients with moderate-severe AP compared to patients with mild AP were treated more often with antibiotics (47% vs 17%, P < 0.01) and total parenteral nutrition (40% vs 8%, P < 0.01), and trended for longer hospitalizations (median of 8 [IQR 4–17] vs 5 [IQR 4–8] days, P = 0.06). Surprisingly, no further clinical features were significantly different, including risk for invasive procedures, maximal analgesic treatment class, or final presumed etiology (Table 1).
Analysis of demographic data and initial laboratory test results on admission demonstrated a younger age at presentation for patients with a moderate-severe AP compared to mild AP: median (IQR) of 8.3 (4.0–14.4) versus 13.8 (8.1–16.0) years, P = 0.02 (Supplementary Table 1, Supplemental Digital Content 2, http://links.lww.com/MPG/B591). In a logistic regression univariate analysis, young age conferred a risk for a moderate-severe disease (odds ratio [OR] 0.9, CI 0.79–0.97); for practical implementation, OR were 3.8 (CI 1.2–12.1) for children younger than 12 years, and 5.8 (CI 1.6–21.4) for children younger than 6 years. A minor difference was also noted in the sodium levels on admission; however, both IQRs were well within normal range and did not suggest a biologically significant difference and thus further analysis was not performed. No additional differences were noted between the 2 groups, including weight percentiles, WBC, hemoglobin, amylase, and lipase levels (Table 1).
We further subanalyzed 59 admissions of patients with first-time episodes of AP, including 8 (13.6%) with moderate-severe and 51 (86.4%) with mild pancreatitis. Clinically, patients with moderate-severe disease were treated more often with total parenteral nutrition (50.0% vs 7.8%, P < 0.01) and antibiotics (75.0% vs 25.5%, P = 0.01), underwent more invasive interventions (25.0% vs 2.0%, P = 0.05), and had longer hospitalizations: median (IQR) of 9 (5–24) versus 5 (4–8) days, P < 0.01, (Table 2).
Analyzing demographics and laboratory results on admission, moderate-severe AP patients were younger with a median (IQR) age of 5.3 (2.9–10.4) versus 12.0 (6.3–15.8) years, P = 0.03 (Supplementary Table 1, Supplemental Digital Content 2, http://links.lww.com/MPG/B591); and had higher levels of WBC counts with a median (IQR) of 22.8 (11.8–31.3) versus 11.0 (8.1–14.6) 109/L, P < 0.01. In a logistic regression univariate analysis, young age conferred a risk for a moderate-severe disease (OR 0.85, CI 0.72–0.99); for practical implementation, OR was 7.5 (CI 1.5–38.2) for children younger than 6 years. Elevated WBC conferred a risk for severe disease (OR 1.2, CI 1.1–1.3), with OR of 5.3 (CI 1.1–25.4) for WBC counts higher than 15 × 109/L and OR of 24 (CI 3.3–173.8) for counts above 20 × 109/L. Additional differences were noted in creatinine and BUN levels; however, both medians and IQRs fell within normal limits and were not biologically different and thus further analysis was not performed (Table 2).
Acute Pancreatitis Versus Acute Recurrent Episodes of Pancreatitis: Clinical Course and Characteristics
We identified 59 first-time AP episodes and 58 AP recurrences. None of the patients with recurrent episodes had features suggestive of chronic pancreatitis, that is, patients experienced prolonged periods of good health in the interims and lacked chronic pancreatitis-compatible findings on sequential imaging studies.
Moderate-severe disease was equivalent in both groups. ARP patients, however, presented earlier (median [IQR] of 1 [1–2] vs 2 [1–5] days, P < 0.01) for medical evaluation, were more often girls (62% vs 42%, P = 0.03), had higher weight percentiles for age and sex (median [IQR] of 71 [54–79] vs 54 [17–74], P = 0.03), and on admission, their lipase levels were higher (median [IQR] of 840 [488–1992] vs 485 [191–1129] U/L, P = 0.02). Additional minute differences were noted in initial chloride, alanine transaminase and gamma-glutamyltransferase levels without biological significance (Supplementary Table 2, Supplemental Digital Content 3, http://links.lww.com/MPG/B592). Furthermore, ARP patients underwent more invasive procedures—mostly ERCP (19.0% vs 5.1%, P = 0.02) and were more likely to receive opioids for analgesia (33.3% vs 12.2%, P = 0.02); but were treated less often with antibiotics (8.6% vs 32.2%, P < 0.01). Lastly, a presumed diagnosis of anatomic anomaly was directly associated with ARP patients (27.6% vs 5.1%, P < 0.01), whereas the diagnosis of medication-induced AP was inversely related (3.4% vs 20.3%, P < 0.01) with them.
Prediction of Acute Recurrent Episodes of Pancreatitis
Fourteen (23.7%) of the 59 patients with first-time episodes, continued to have recurrent episodes of AP during the study period. Six (10.2%) had 1 additional admission for ARP and 8 (13.6%) had 2 or more. The estimated one and 2-year risks for recurrence of AP were 17% and 26%, respectively; actually, only 1 patient had recurrence over 2 years after his first AP admission (Supplementary Fig. 2, Supplemental Digital Content 1, http://links.lww.com/MPG/B590).
Interestingly, analysis of demographics, laboratory test results upon admission, clinical course including severity of the first episode, and presumed etiology failed to demonstrate significant risk factors as predictors for future ARP. Essentially, patients who developed ARP following the initial episode did not differ from those who did not (Supplementary Table 3, Supplemental Digital Content 4, http://links.lww.com/MPG/B593).
In this retrospective study, we describe a cohort of 68 patients and 117 separate episodes of AP in a tertiary medical center over the course of 20 years. We further subanalyzed a selective and more homogenous cohort of 59 patients with a first-time episode of AP. Despite 15 (13%) moderate or severe episodes in the general cohort, and 8 (14%) in the subanalysis, prognosis was excellent. Young age was associated with a more severe course of disease (along with elevated WBC on admission in the subanalysis). Overall, patients with ARP episodes were more often girls, had higher weight percentiles for age, and had higher levels of admission lipase. Although nearly quarter (24%) of the patients with a first episode of AP, however, had recurrence, we failed to identify predictors of future ARP.
For the majority of pediatric patients, AP is a mild disease, amenable to supportive measures with a favorable prognosis. Multiple studies, however, showed that up to 36% of patients will develop local or systemic complications, indicating a moderate or severe disease, potentially warranting advanced imaging, invasive therapeutic interventions, and a worse prognosis (15–17,24). Identification of patients at risk for complications is highly important and will enable a prompt reply.
Several prognostication models were demonstrated to predict mortality in adults with AP including the Ranson, Modified Glasgow, Acute Physiology and Chronic Health Evaluation II, and the Bedside Index for Severity in Acute Pancreatitis scores (10–13). In children, these models perform poorly, questionably as a result of different etiologies and associated comorbidites (14–17). DeBanto et al (15) were the first to formulate a specialized pediatric score—the Pediatric Acute Pancreatitis Severity (PAPS) score, largely based on the Ranson and modified Galsgow scores. The PAPS score, combining age, weight, WBC, LDH on admission, and trough levels of calcium, albumin, the amount of fluid sequestration, and rise in BUN during the first 48 hours, was shown to perform better than the prototypes at predicting severity with 70% sensitivity and 79% specificity (15,22). In 2 similarly structured studies, Lautz et al (16), and Suzuki et al (21), however, found the PAPS score to have comparable sensitivity (50%–60%), specificity (86%–99%), and positive (24%–71%) and negative predictive values (80%–84%/96%–98%) as the Ranson and modified Glasgow scores, and that all 3 performed poorly (16,21). Suzuki et al (21) suggested a different score based on age/weight, BUN, LDH, platelet count, calcium, C-reactive protein, and clinical parameters, that performed better (80%, 96%, 62%, and 98%, respectively). Several other studies demonstrated independent associations of single parameters (WBC, trough calcium levels, BUN, lipase, and D-dimer) with AP severity (16,17,19,20). Likewise, Coffey et al (18,23) demonstrated that high levels of lipase ≥7× ULN predicted severity with 85% sensitivity and 56% specificity, and Bierma et al (23) suggested adding lipase drop on day 2 or calcium for enhanced specificity. Szabo et al (24), introduced a score based on the combination of lipase, albumin, and WBC on admission and demonstrated 68% sensitivity and 71% specificity. Finally, Izquierdo et al (25) reported recently 82% sensitivity and 64% specificity using a severity score based on hemoglobin and BUN. Moreover, a CT-based severity index (the Balthazar score) was demonstrated to be superior over clinical scoring systems in an earlier comparative study, and more recently was shown to predict need for admission to a pediatric intensive care unit; however, the utility of this score is limited by the decreasing use of CT in children (17,33,34). In summary, despite extensive research, the derived scores performances are limited, and exhibit either high sensitivity or specificity but not both. Adding to this, we found that young age and elevated WBC counts on admission were associated with a moderate-severe disease course in patients with a first-time AP. Analysis of the entire cohort of AP patients yielded even poorer results as young age alone was associated with moderate-severe disease. Our limited findings are in line with the past reports, indicating that AP severity in children is a poorly predicted disease.
A significant proportion of patients with a first episode of AP will continue to have ARP. In adults, ARP was described in up to 29% of first-time AP and was associated with alcohol and tobacco abuse, and biliary AP (35–39). These risk factors are, however, rare in children and thus the adult literature is of limited value. In children, ARP was described in up to 15% of patients presenting with AP (26,27,29). Recently, following an extensive literature review, a consensus statement by the International Study group of Pediatric Pancreatitis: In search for a cure group of experts reported several factors associated with ARP including anatomical pancreatic anomalies; metabolic disorders including hypertriglyceridemia, hypercalcemia, organic acidosis syndromes, and maple syrup urine disease; pancreatitis-predisposing genetic mutations including PRSS1, SPINK1, CFTR, and CTRC; and potentially celiac (28).
Although our finding of 24% ARP prevalence is higher than previously reported, it may still underestimate the true prevalence given the retrospective nature of the study. Although we found that patients with ARP episodes were more often girls, had higher weight percentiles for age, and had higher levels of admission lipase, analysis of patients with a first-time AP episode did not identify factors associated with future ARP (Supplementary Table 3, Supplemental Digital Content 4, http://links.lww.com/MPG/B593). Occasionally, patients with anatomical anomalies were not identified at first presentation but only on subsequent admissions. Furthermore, genetic testing for at-risk genes was not routinely done; nevertheless, none of the tested patients was found positive. Finally, lack of statistical power may have contributed for the negative results, as certain trends (ie, anatomic abnormalities and weight percentiles) reached significance when analyzed for all patients with ARP compared to first-time AP (Supplementary Table 2, Supplemental Digital Content 3, http://links.lww.com/MPG/B592).
Interestingly, analgesic treatments were comparable between mild and moderate-severe patients, despite the fact that moderate-severe patients were hospitalized longer and underwent invasive procedures more often. This suggests that subjective perceptive measures such as pain correlate poorly with disease severity. Moreover, ARP patients presented earlier to medical attention, and were treated more vigorously with analgesics compared to first-time AP patients, suggestive of a relative hypersensitivity in ARP patients. Chronic pancreatitis and ARP share common risk factors and occasionally ARP evolves to chronic pancreatitis. Subsequently, it was postulated that the 2 entities are not different diseases but rather at ends of a single spectrum (28,29,40). So far, central hyperalgesia was described in chronic pancreatitis but not ARP (41,42). Our finding of increased use of opioids in ARP patients may reflect peripheral (organ) or central (central nervous system) hypersensitivity. Further research is required to re-evaluate this clinical finding, and specifically to address whether ARP also causes central hyperalgesia as chronic pancreatitis.
The main limitation of our study is its retrospective nature. Relying on available historic data found in patients’ charts may lead to bias through different admission and discharge criteria, timing of obtaining the different studies, availability of laboratory testing, imaging studies interpretation variability, and nonprotocoled approach for conservative and invasive therapeutic measures. Specifically, lack of routine testing for pancreatic sufficiency and advanced imaging for patients with recurrent pancreatitis may have contributed for underdiagnoses of chronic pancreatitis. Finally, the fact that our hospital is a tertiary center with pediatric ERCP capacity may contribute to a selection bias. Indeed, our comparable findings of moderate-severe disease frequency and ARP incidence with those reported elsewhere support the reliability of our findings.
In conclusion, AP carries a favorable prognosis in children even in the face of moderate or severe presentation. Our main findings support past observations that demographic data or surrogate laboratory markers upon admission poorly predict AP severity in children, highlighting the complex nature of AP. Likewise we did not identify demographic, laboratory, or clinical predictors for future development of ARP following a first episode of AP. Finally, analgesic treatment requirements were independent of disease severity, but increased in ARP episodes over first-time AP episodes.
1. Bhatia M, Wong FL, Cao Y, et al. Pathophysiology of acute pancreatitis. Pancreatology
2. Lopez MJ. The changing incidence of acute pancreatitis in children: a single-institution perspective. J Pediatr
3. Nydegger A, Heine RG, Ranuh R, et al. Changing incidence of acute pancreatitis: 10-year experience at the Royal Children's Hospital, Melbourne. J Gastroenterol Hepatol
4. Morinville VD, Barmada MM, Lowe ME. Increasing incidence of acute pancreatitis at an American pediatric tertiary care center: is greater awareness among physicians responsible? Pancreas
5. Benifla M, Weizman Z. Acute pancreatitis in childhood: analysis of literature data. J Clin Gastroenterol
6. Bai HX, Lowe ME, Husain SZ. What have we learned about acute pancreatitis in children? J Pediatr Gastroenterol Nutr
7. Guo Q, Li A, Xia Q, et al. The role of organ failure and infection in necrotizing pancreatitis: a prospective study. Ann Surg
8. Vasilescu A, Cuffari C, Santo Domingo L, et al. Predictors of severity in childhood pancreatitis: correlation with nutritional status and racial demographics. Pancreas
9. Restrepo R, Hagerott HE, Kulkarni S, et al. Acute pancreatitis in pediatric patients: demographics, etiology, and diagnostic imaging. Am J of Roentgenol
10. Ranson JH. Etiological and prognostic factors in human acute pancreatitis: a review. Am J Gastroenterol
11. Blamey SL, Imrie CW, O’Neill J, et al. Prognostic factors in acute pancreatitis. Gut
12. Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med
13. Wu BU, Johannes RS, Sun X, et al. The early prediction of mortality in acute pancreatitis: a large population-based study. Gut
14. Meyer A, Coffey MJ, Oliver MR, et al. Contrasts and comparisons between childhood and adult onset acute pancreatitis. Pancreatology
15. DeBanto JR, Goday PS, Pedroso MR, et al. Acute pancreatitis in children. Am J Gastroenterol
16. Lautz TB, Chin AC, Radhakrishnan J. Acute pancreatitis in children: spectrum of disease and predictors of severity. J Pediatr Surg
17. Fabre A, Petit P, Gaudart J, et al. Severity scores in children with acute pancreatitis. J Pediatr Gastroenterol Nutr
18. Coffey MJ, Nightingale S, Ooi CY. Serum lipase as an early predictor of severity in pediatric acute pancreatitis. J Pediatr Gastroenterol Nutr
19. Antunes H, Nascimento J, Mesquita A, et al. Acute pancreatitis in children: a tertiary hospital report. Scand J Gastroenterol
20. Boskovic A, Pasic S, Soldatovic I, et al. The role of D-dimer in prediction of the course and outcome in pediatric acute pancreatitis. Pancreatology
21. Suzuki M, Saito N, Naritaka N, et al. Scoring system for the prediction of severe acute pancreatitis in children. Pediatr Int
22. Hashimoto N, Yotani N, Michihata N, et al. Efficacy of pediatric acute pancreatitis scores at a Japanese tertiary center. Pediatr Int
23. Bierma MJ, Coffey MJ, Nightingale S, et al. Predicting severe acute pancreatitis in children based on serum lipase and calcium: a multicentre retrospective cohort study. Pancreatology
24. Szabo FK, Hornung L, Oparaji JA, et al. A prognostic tool to predict severe acute pancreatitis in pediatrics. Pancreatology
25. Izquierdo YE, Fonseca EV, Moreno LÁ, et al. Multivariate model for the prediction of severity of acute pancreatitis in children. J Pediatr Gastroenterol Nutr
26. Werlin SL, Kugathasan S, Frqutschy BC. Pancreatitis in children. J Pediatr Gastroenterol Nutr
27. Park A, Latif SU, Shah AU, et al. Changing referral trends of acute pancreatitis in children: a 12-year single-center analysis. J Pediatr Gastroenterol Nutr
28. Gariepy CE, Heyman MB, Lowe ME, et al. The causal evaluation of acute recurrent and chronic pancreatitis in children: consensus from the INSPPIRE group. J Pediatr Gastroenterol Nutr
29. Poddar J, Yachha SK, Borkar V, et al. A report of 320 cases of childhood pancreatitis—increasing incidence, etiologic categorization, dynamics, severity assessment, and outcome. Pancreas
30. Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis—2012: revision of the Atlanta classification and definitions by international consensus. Gut
31. Morinville VD, Husain SZ, Bai H, et al. Definitions of pediatric pancreatitis and survey of current clinical practices. J Pediatr Gastroenterol Nutr
32. Marshall JC, Cook DJ, Christou NV, et al. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med
33. Lautz TB, Turkel G, Radhakrishnan J, et al. Utility of the computed tomography severity index (Balthazar score) in children with acute pancreatitis. J Pediatr Surg
34. Izquierdo YE, Fonseca EV, Moreno LA, et al. Utility of CT classifications to predict unfavorable outcomes in children with acute pancreatitis. Pediatr Radiol
35. Lankisch PG, Breuer N, Bruns A, et al. Natural history of acute pancreatitis: a long-term population-based study. Am J Gastroenterol
36. Takeyama Y. Long-term prognosis of acute pancreatitis in Japan. Clin Gastroenterol Hepatol
37. Cote GA, Yadav D, Slivka A, et al. Alcohol and smoking as risk factors in an epidemiology study of patients with chronic pancreatitis. Clin Gastroenterol Hepatol
38. Nojgaard C, Becker U, Matzen P, et al. Progression from acute to chronic pancreatitis: prognostic factors, mortality, and natural course. Pancreas
39. Yadav D, O’Connell M, Papachristou GI. Natural history following the first attack of acute pancreatitis. Am J Gastroenterol
40. Poddar U, Yachha SK, Borkar V, et al. Is acute recurrent pancreatitis in children a precursor of chronic pancreatitis? A long-term follow-up study of 93 cases. Dig Liver Dis
41. Buscher HC, Wilder-Smith OH, van Goor H. Chronic pancreatitis patients show hyperalgesia of central origin: a pilot study. Eur J Pain
42. Drewes AM, Krarup AL, Detlefsen S, et al. Pain in chronic pancreatitis: the role of neuropathic pain mechanisms. Gut