Chronic pancreatitis (CP) is a rare clinical entity in children, with an unknown incidence and prevalence (1). Recurrent attacks of acute pancreatitis (RAP; defined as ≥2 episodes of pancreatitis) and CP are characterized by recurrent or ongoing inflammation of the pancreas leading to abdominal pain, maldigestion (exocrine tissue destruction), and diabetes (endocrine tissue destruction). Early changes of CP include irregularity of the smaller ducts and branches of the pancreas. With the progression of disease, the main pancreatic duct develops strictures, calcification, and dilatation. At a later stage, fibrosis and destruction of endocrine parenchyma lead to diabetes mellitus. Intractable pain is a significant problem in these patients and is often resistant to medical therapy.
Chronic alcohol use and cholelithiasis are the most common causes of RAP in adults. The most common causes of RAP and CP in children have been thought to be anatomic abnormalities, trauma, metabolic, and drug toxicity (2); however, in a large study of patients of all ages, no cause could be found in 51% of patients (3). Improved imaging techniques and genetic testing have allowed the identification of patients at high risk for developing recurrent pancreatitis and CP (4). Advances in the field of genetics led to the discovery that mutations in the cationic trypsinogen (PRSSI), serine protease inhibitor Kazal type 1 (SPINK1), cystic fibrosis transmembrane conductor regulator (CFTR), and chymotrypsin C (CTRC) genes can predispose patients to develop CP and RAP (5). There is growing evidence that a significant proportion of patients with “idiopathic” RAP and CP have mutations in one of the above-mentioned genes. The aim of our study was to estimate the prevalence and describe the clinical characteristics and outcome of RAP and CP in children with mutations in ≥1 of the PRSS1, CFTR, and SPINK1 genes.
We reviewed the charts of children ages 18 years or younger with RAP or CP diagnosed from 2000 to 2009 at the Children's Hospital of Wisconsin, Milwaukee. We analyzed the demographic data including sex, age of presentation, age of diagnosis, number of documented episodes of pancreatitis, presenting symptoms of first episode, family history, radiological and laboratory workup, and disease course. Because there are no accepted definitions of RAP and CP in children, RAP was defined as ≥2 episodes of acute pancreatitis presenting with abdominal pain and 3-fold increased serum amylase or lipase or imaging changes suggestive of acute pancreatitis. CP was defined as ≥2 episodes of pancreatitis associated with pancreatic duct abnormalities such as dilatation, strictures, intraductal plugs containing calculi, or pancreatic insufficiency (6). Children with pancreatitis secondary to gallstones, trauma, medication, infection, or metabolic disorders were excluded. Two children with pancreas divisum did not undergo genetic testing, although the role of pancreas divisum as a cause of RAP or CP remains controversial. The diagnosis of AP or CP associated with genetic mutation was made when the patient had a mutation in the PRSS1, SPINK1, or CFTR genes. Genetic testing was performed by a commercial laboratory (Ambry Genetics, Aliso Viejo, CA). We included all of the patients undergoing a complete pancreatitis panel using the Ambry test: pancreatitis (full gene analysis for CFTR, PRSS1, and SPINK1). Only one of our patients was tested for CTRC because testing was unavailable at the time that most of our patients were seen. Patients who were found to have mutations in any of the above-mentioned genes were included in the study. All of the information was entered into a Microsoft Excel spreadsheet for analysis. The present study was approved by the institutional review board of the Children's Hospital of Wisconsin.
The Fisher exact test was used to compare the prevalence of CP in boys and girls, and the proportion of CP in double versus single heterozygotes. The Mann-Whitney test was used to compare the age of onset of CP in boys versus girls.
Among 36 children with the diagnosis of RAP or CP, 7 patients were excluded because of gallstones (n = 3), pancreas divisum (n = 2), metabolic disorder (n = 1), or medication (n = 1) (Fig. 1). These were thought to be the cause of pancreatitis, and genetic testing was not performed. Of the 29 patients, 6 had negative genetic testing and 23 patients had a mutation in ≥1 of the above-mentioned gene mutations. Fourteen of 29 (48%) patients had CFTR mutations, 7 of 29 (29%) had PRSS1 mutations, and 8 of 29 (27%) had SPINK1 mutations.
The median age of onset of the first episode of acute pancreatitis in gene-positive patients was 5 years (range 9 months–15 years) and the median age of diagnosis was 6.5 years (range 1–16 years). There was no significant difference (P = 0.41) in the age of onset between boys (median 5 years) and girls (median 8 years). Twenty-one were white and the rest were Hispanic, African American, and Arab. Fourteen patients were girls. The most common presenting symptoms were abdominal pain (100%), vomiting (74%), nausea (40%), and diarrhea (13%). Jaundice and fever were seen in 8% of patients. Family history was present in 5 of 29 gene-tested patients (17%): 3 with CFTR and 2 with PRSS1 mutations. None of our patients were below the 2 standard deviations line for weight. During acute crisis, all of the patients showed biochemical signs of pancreatitis with increased serum amylase or lipase >3 times normal. The patients with RAP had 2 to 8 episodes within 3.6 years on average (range 1–10 years) of follow-up. Eleven of 29 (38%) patients met the criteria for CP as documented by radiological studies (Table 1). The prevalence of CP was significantly higher in boys (77.8%) than in girls (28.6%; P < 0.036) with no known explanation. Six patients with CP had a combination of CFTR and SPINK1 or PRSS1 mutations. The median age of onset of the 6 double heterozygous patients with CP was 8.75 years (range 0.75–13.5 years), and the time to develop CP was 4 to 8 years. All of the patients who were double heterozygous for both CFTR and SPINK1 mutations had CP. The prevalence of CP was significantly higher (P < 0.021) in the double heterozygotes (100% in CFTR and SPINK and CFTR and PRSS1) than in the single heterozygotes (29.4%). Eight patients developed a chronic pain syndrome and 2 developed exocrine pancreatic insufficiency during the follow-up period.
CFTR mutations were identified in 48% (14/29) of patients as shown in Table 1. One patient was homozygous for p.L997F, and this mutation has been reported to be associated with pancreatitis (7). This patient presented at age 14 years with RAP, normal growth, pulmonary function, and equivocal sweat tests (50 mEq/L). Another patient was double heterozygous (R533X/A349 V). This patient presented with RAP with the first episode at 1 year of age, but maintained normal growth and pulmonary function. Six patients were heterozygous for the CFTR mutation (F508del, R297W, D1152H, R297Q, and I148T). Two other patients had double heterozygous mutations in F508 del/2789 + 5G > A and F508 del/5T variant. Four patients carried 5 alleles of 5T variants; 3 of them had combined 5T and 7T variants: (TG)11-5T/(TG) 11-7T, (TG)11-5T/(TG)11-5T, (TG)11-5T/(TG)12-7T, (TG)11-5T/(TG) 11-7T (8). Only 8 patients had a sweat chloride test done. One patient with F508del/2789 + 5G > A had an abnormal sweat test, another patient with p.L997/p. L997 had an equivocal sweat test (50 mEq/L), and the rest were normal (<40 mEq/L). All of the Wisconsin-born patients passed the newborn screening for CF.
A PRSS1 mutation was present in 7 patients (24%), 3 with RAP and 4 with CP (Table 1). The median age of presentation was 5.5 years (range 3–11 years). Two patients each had R122H and N29I mutations. Patients with N29I presented at age 2 years. Mutation p.C139F was present in 1 patient who presented at age 8 years with recurrent abdominal pain and later developed CP and a chronic pain syndrome. This mutation has been found only in a few patients, but clear segregation with disease and/or the biochemical and cell-biological properties of the mutant proteins are consistent with disease association (9). Another patient presented at age 9 months with vomiting and irritability. This patient was diagnosed as having RAP and was found to have a combined mutation in PRSS1 (p.R116C) and CFTR genes (I148T variant on exon 4). Hereditary pancreatitis has been described in carriers of the p.R116C mutation, and I148T variant was originally reported to be a disease-causing mutation (10). E79K mutation was found in another patient who also carried 2 copies of the 5T variants in the CFTR gene.
A SPINK1 mutation was found in 27% (8/29) of patients; all of them were heterozygous (Table 1), 2 with RAP and 6 with CP. The most frequent mutation seen in 4 patients was N34S, and the median age of presentation of this subgroup was 10.7 years. One patient was heterozygous for a known disease-modifying mutation (P55S) (3). Two patients were heterozygous for a previously undescribed variant p.N64D (also known as c.190A > G) in exon 3. This variant results from A to G substitution at nucleotide position 190, which resulted in changes of amino acid from asparagine to aspartate. Another previously unreported variant, p.L9R (also known as c.26T > G), was found in exon 1 of the SPINK1 gene that resulted in change of amino acid from leucine to arginine. Both of the above patients carried 1 copy of the 5T variants in the CFTR gene.
CFTR Mutation Associated With SPINK1 or PRSS1 Mutations
Four patients were double heterozygotes for the CFTR and the SPINK1 gene mutations. Two other patients were double heterozygote for CFTR and PRSS1 mutation. All of the 6 patients developed radiological signs of CP.
CTRC (Chymotrypsin) Mutation
CTRC testing conducted for 1 patient was negative because the test was unavailable at the time that most of our patients were seen.
The present study shows that pediatric RAP or CP is often associated with SPINK1, PRSS1, and/or CFTR mutations. Twenty-three of 29 patients (79%) with RAP and/or CP were found to have a mutation in ≥1 of these genes. The median age of onset of clinical symptoms was 5 years and of diagnosis was 6.5 years. Our patients had a similar age of onset, but diagnosis of genetically inherited or familial pancreatitis occurred at a younger age. Previous reports describe a 9-year gap between the onset of clinical symptoms and diagnosis (11). In our study, the median age of the onset of clinical symptoms and diagnosis was 1.5 years, possibly reflecting an increasing awareness of the disease and the commercialization of genetic testing, making it more readily available.
CF is the most common inherited disease of the exocrine pancreas among whites. The association of CFTR mutations with idiopathic CP was initially described in 1998 (12,13). Symptomatic pancreatitis occurs in 22% of patients with cystic fibrosis and pancreatic sufficiency (14,15) and rarely occurs in patients with pancreatic insufficiency (16). A spectrum of pancreatic abnormalities occurs in patients with CF, especially in those with atypical CF mutations. The mild-variable and borderline mutations, when combined with a severe CFTR mutation, are associated with atypical CF (17,18). The mild-variable mutation accounts for the residual pancreatic function that protects these patients from CF lung disease and may be required for pancreatitis (19). Pancreatitis in these patients is thought to result from the viscid secretions producing pancreatic ductular obstruction, ineffective clearing of secretions from the pancreatic duct, and autodigestion of the pancreas by activated proteolytic enzymes. In a recent study by Ooi et al (20), patients with pancreatic-sufficient CF (277 with pancreatitis and 215 without pancreatitis) were analyzed. Patients with pancreatitis were more likely to have genotypes associated with mild (70%) than moderate to severe (30%) disease based on the pancreatic insufficiency prevalence score (P = 0.004).
In our series, at least 1 CFTR mutation was found in 48% (14/29) of patients. This is higher than that previously reported by Keiles et al (3). Of 381 patients with a primary diagnosis of chronic or recurrent pancreatitis, Keiles et al identified CFTR mutant alleles in 32% of 381 patients; however, our findings are consistent with the observation of Bishop et al (21), who found that at least 1 CFTR mutation or variant was carried in 18 of 40 patients (45%) with idiopathic CP and in 6 of 16 patients (38%) with idiopathic RAP. In addition, Cohn et al (12) studied 27 patients referred for an evaluation of idiopathic pancreatitis. DNA was tested for 17 CFTR mutations and for the 5T variants; 10 patients (37%) had at least 1 abnormal CFTR allele. In the Cohn et al study, the association between CFTR mutations and pancreatitis may have been underestimated because DNA was tested for only 17 CFTR mutations. The above 3 studies are from the adult population and hence head-to-head comparisons between our pediatric population and the adult study cannot be made. This is relevant because it would be expected that a pediatric population would have a higher chance of having genetic factors involved in pancreatitis manifesting in childhood, versus a group of mostly adults who are undergoing investigations for, most likely, adult-onset pancreatitis.
The 1 patient homozygous for p.L997F had an equivocal sweat chloride level and this mutation has been reported to be associated with pancreatitis (7). Our observation suggests that the p.L997F mutation can cause pancreatitis, although a recent review reported the mutation to be of unknown consequence (22). Another patient was double heterozygous for R533X/A349V, and both of these mutations are reported in patients with CF. Patients homozygous for R533X usually have severe pancreatic insufficiency (23,24). Other mutations (R297Q, D1152H, R297Q, and I148T) may be present in CF and CFTR-related disorders such as pancreatitis and obstructive azoospermia (25). The significance of different CFTR mutations observed in our study is summarized in Table 2(7,21,22,26,27).
The importance of a single CFTR mutation in the pathogenesis of RAP or CP remains controversial (28). Because the estimated carrier frequency of CFTR mutation is 1:32 in Wisconsin, we would expect that at most 1 or 2 of our patients would carry this mutation by chance. The presence of this mutation in 14 of 29 of our patients strongly suggests that a single mutant CFTR allele is one of the factors that may contribute to the pathogenesis of RAP and CP.
Hereditary pancreatitis, an autosomal dominant disease with incomplete penetrance (4), is associated with single-point mutations in the third exon of the cationic trypsinogen gene on chromosome 7. More than 20 mutations have been reported, R112H with N29I being the most commonly identified mutations. In our series, 24% (7/29) of our patients had a PRSS1 mutation, and all of these mutations are known to be disease causing except the p.C139F mutation, which is of uncertain significance (9). R112H and N29I were the most common mutations, being found in 4 patients, and this is consistent with previous reports. The frequency of PRSS1 mutation is higher than what was reported previously (9.4%) by Keiles and Kammesheidt (3). The Keiles and Kammesheidt population was predominantly adults, which may explain the difference in the frequency of PRSS1 mutations, and this is a limitation of our study.
It is interesting to note that the 2 patients in our series with N29I mutations presented at the early age of 2 years. This is in contrast to the Howes et al (29) report that described a later onset of symptoms in patients with an N29I mutation (14 years old) versus those with an R122H mutation (10 years old). A larger patient series is needed to determine whether specific mutations present at various ages.
Serine protease inhibitor, Kazal type 1 (SPINK1), is a potent protease inhibitor that is thought to be an inactivation factor of intrapancreatic trypsin activity. SPINK1 mutations are found in 1% to 3% of the population, and because of its low penetrance and association with CP from other causes, SPINK1 mutations are thought to be more disease predisposing or modifying rather than disease causing. SPINK1 mutations have been shown to increase the risk of RAP/CP in patients who already have had an episode for acute pancreatitis compared with patients with acute pancreatitis and no SPINK1 mutation (26); however, a study showed that 34% of patients with sporadic pancreatitis had SPINK1 variants with a median age of onset of pancreatitis at 13 years (27).
Eight of 29 (27%) of our patients had a SPINK1 mutation compared with previous reports of 23% and 8.9% (3). The N34S mutation has been reported to be the most frequent SPINK1 mutation, being found in 75% of patients with CP and SPINK1 mutations (26). This mutation was identified in 4 of our patients. Three patients had novel variants (p.N64D, p.L9R) that have not been described. The significance of these novel mutations is not certain because we lack an adequate control group. Two of the above patients carried 1 copy of the 5T variants in the CFTR gene and the combination of these 5T variants in both patients can contribute to their pancreatitis. Few studies reported a strong association between SPINK1 and CFTR mutations in CP (25). Schneider et al (27) showed that CFTR variant p.R75Q with a SPINK1 mutation increases the risk of CP with an odds ratio of 25.1. Four of our patients with CP were double heterozygous for a CFTR mutation and a SPINK1 mutation. This suggests that this association is also important in pediatric patients. SPINK1 homozygous mutations have been implicated in RAP and CP in the Indian subcontinent, and thus, in select populations, may be disease causing, although the majority of our study population is white (30). We could not conduct any analysis on the CTRC (chymotrypsin) test because it was given to only 1 patient, and this was reported to be negative.
The discovery of mutations in the above-mentioned genes known to be associated with CP helped answer a variety of questions posed by families. All of the families were offered and accepted genetic counseling. This avoids the additional, fruitless search for the cause of the pancreatitis in these patients and allowed at the parents’ request to screen siblings in selected cases.
In our small retrospective series, 7 patients were excluded because of known causes and did not undergo genetic testing by the attending physician. Two of these patients had pancreas divisum, which may not have been the underlying cause of their pancreatitis. Pancreatitis in patients with pancreas divisum has been associated with mutations in both SPINK1 and CFTR genes (31). Another limitation is that the genetic testing is done by a single commercial laboratory (Ambry), and our results are based on the genetic testing done only in this laboratory.
We found a high prevalence of genetic mutations or variants in patients without anatomic or metabolic abnormalities known to be associated with pancreatitis. This highlights the importance of genetic testing in patients with RAP and CP. We also found that the prevalence of CP was significantly higher in the double heterozygous patients (CFTR and SPINK and CFTR and PRSS1) than in the single heterozygous patients. Future studies are needed to ascertain the genetic causes of RAP and CP and to examine the relation between single CFTR mutation and mutation in SPINK1 genes.
The authors thank Dr Raymond G. Hoffmann and Mahua Dasgupta for analysis of the statistics.
1. Rosendahl J, Bodeker H, Mossner J, et al. Hereditary chronic pancreatitis. Orphanet J Rare Dis 2007; 2:1.
2. Werlin SL, Kugathasan S, Frautschy BC. Pancreatitis in children. J Pediatr Gastroenterol Nutr 2003; 37:591–595.
3. Keiles S, Kammesheidt A. Identification of CFTR, PRSS1, and SPINK1 mutations in 381 patients with pancreatitis. Pancreas 2006; 33:221–227.
4. Whitcomb DC, Gorry MC, Preston RA, et al. Hereditary pancreatitis is caused by a mutation in the cationic trypsinogen gene. Nat Genet 1996; 14:141–145.
5. Whitcomb DC. Genetic aspects of pancreatitis. Annu Rev Med 2010; 61:413–424.
6. Witt H, Becker M. Genetics of chronic pancreatitis. J Pediatr Gastroenterol Nutr 2002; 34:125–136.
7. Fanen P, Ghanem N, Vidaud M, et al. Molecular characterization of cystic fibrosis: 16 novel mutations identified by analysis of the whole cystic fibrosis conductance transmembrane regulator (CFTR) coding regions and splice site junctions. Genomics 1992; 13:770–776.
8. Groman JD, Hefferon TW, Casals T, et al. Variation in a repeat sequence determines whether a common variant of the cystic fibrosis transmembrane conductance regulator gene is pathogenic or benign. Am J Hum Genet 2004; 74:176–179.
9. Szmola R, Sahin-Toth M. Uncertainties in the classification of human cationic trypsinogen (PRSS1) variants as hereditary pancreatitis-associated mutations. J Med Genet 2010; 47:348–350.
10. Bozon D, Zielenski J, Rininsland F, et al. Identification of four new mutations in the cystic fibrosis transmembrane conductance regulator gene: I148T, L1077P, Y1092X, 2183AA-->G. Hum Mutat 1994; 3:330–332.
11. Rebours V, Boutron-Ruault MC, Schnee M, et al. The natural history of hereditary pancreatitis: a national series. Gut 2009; 58:97–103.
12. Cohn JA, Friedman KJ, Noone PG, et al. Relation between mutations of the cystic fibrosis gene and idiopathic pancreatitis. N Engl J Med 1998; 339:653–658.
13. Sharer N, Schwarz M, Malone G, et al. Mutations of the cystic fibrosis gene in patients with chronic pancreatitis. N Engl J Med 1998; 339:645–652.
14. Shwachman H, Lebenthal E, Khaw KT. Recurrent acute pancreatitis in patients with cystic fibrosis with normal pancreatic enzymes. Pediatrics 1975; 55:86–95.
15. Atlas AB, Orenstein SR, Orenstein DM. Pancreatitis in young children with cystic fibrosis. J Pediatr 1992; 120:756–759.
16. del Rosario JF, Putnam PE, Orenstein DM. Chronic pancreatitis in a patient with cystic fibrosis and clinical pancreatic insufficiency. J Pediatr 1995; 126:951–952.
17. Durno C, Corey M, Zielenski J, et al. Genotype and phenotype correlations in patients with cystic fibrosis and pancreatitis. Gastroenterology 2002; 123:1857–1864.
18. Zielenski J, Tsui LC. Cystic fibrosis: genotypic and phenotypic variations. Annu Rev Genet 1995; 29:777–807.
19. Durie PR. Pancreatic aspects of cystic fibrosis and other inherited causes of pancreatic dysfunction. Med Clin North Am 2000; 84:609–620.
20. Ooi CY, Dorfman R, Cipolli M, et al. Type of CFTR mutation determines risk of pancreatitis in patients with cystic fibrosis. Gastroenterology 2010;140:153–61.
21. Bishop MD, Freedman SD, Zielenski J, et al. The cystic fibrosis transmembrane conductance regulator gene and ion channel function in patients with idiopathic pancreatitis. Hum Genet 2005; 118:372–381.
22. Ooi CY, Gonska T, Durie PR, et al. Genetic testing in pancreatitis. Gastroenterology 2010; 138:2202–2206.
23. Walkowiak J, Herzig KH, Witt M, et al. Analysis of exocrine pancreatic function in cystic fibrosis: one mild CFTR mutation does not exclude pancreatic insufficiency. Eur J Clin Invest 2001; 31:796–801.
25. Rohlfs EM, Zhou Z, Sugarman EA, et al. The I148T CFTR allele occurs on multiple haplotypes: a complex allele is associated with cystic fibrosis. Genet Med 2002; 4:319–323.
26. Witt H, Luck W, Hennies HC, et al. Mutations in the gene encoding the serine protease inhibitor, Kazal type 1 are associated with chronic pancreatitis. Nat Genet 2000; 25:213–216.
27. Schneider A, Larusch J, Sun X, et al. Combined bicarbonate conductance-impairing variants in CFTR and SPINK1 variants are associated with chronic pancreatitis in patients without cystic fibrosis. Gastroenterology 2011; 140:162–171.
28. Scotet V, De Braekeleer M, Audrézet MP, et al. Prevalence of CFTR mutations in hypertrypsinaemia detected through neonatal screening for cystic fibrosis. Clin Genet 2001; 59:42–47.
29. Howes N, Greenhalf W, Rutherford S, et al. A new polymorphism for the RI22H mutation in hereditary pancreatitis. Gut 2001; 48:247–250.
30. Bhatia E, Choudhuri G, Sikora SS, et al. Tropical calcific pancreatitis: strong association with SPINK1 trypsin inhibitor mutations. Gastroenterology 2002; 123:1020–1025.
31. Garg PK, Khajuria R, Kabra M, et al. Association of SPINK1 gene mutation and CFTR gene polymorphisms in patients with pancreas divisum presenting with idiopathic pancreatitis. J Clin Gastroenterol 2009; 43:848–852.