High Incidence of PRSS1 and SPINK1 Mutations in Korean Children With Acute Recurrent and Chronic Pancreatitis

Abstract

Objectives: 

We evaluated the frequencies and clinical consequences of mutations in the genes encoding cationic trypsinogen, serine protease 1 (PRSS1), and serine protease inhibitor Kazal type 1 (SPINK1) in children with acute recurrent pancreatitis (ARP) and chronic pancreatitis (CP).

Patients and Methods: 

The study population consisted of 32 children with ARP or CP and 28 healthy controls. We analyzed clinical data and the sequences of the entire coding region and the intron-exon boundaries of the PRSS1 and SPINK1 genes from each patient.

Results: 

Fifteen (46.9%) of the 32 patients had at least 1 PRSS1 or SPINK1 mutation. Four (12.5%) of the 32 patients carried the p.N29I, p.R122H, or p.N29T mutation or a p.G208A variant of the PRSS1 gene in a heterozygote state. Eleven (34.4%) of the 32 patients carried either the IVS3 + 2T>C or p.N34S mutation of the SPINK1 gene. No PRSS1 or SPINK1 mutations were identified in the control group. In particular, mutations were identified in 4 of our patients who experienced pancreas divisum with CP, whereas the remaining 2 patients with pancreas divisum and ARP did not have mutation.

Conclusions: 

The frequencies of the PRSS1 and SPINK1 mutations are relatively high in Korean children with ARP or CP. Mutations in the PRSS1 and SPINK1 genes are highly associated with the development of childhood ARP or CP. Our findings suggest that patients with genetic mutations combined with pancreas divisum tend to develop CP early.

Biliary disease, systemic disease, and drug side effects are common causes of childhood pancreatitis ^(1,2). Many cases of pancreatitis resolve after conservative therapy and removal of the underlying cause, for example, surgical correction of choledochal cysts or drug withdrawal in cases of drug-induced pancreatitis. Hence, determination of the etiology of pancreatitis is particularly important for successful treatment. Despite the use of adequate biochemical and radiologic studies, combined with endoscopic retrograde cholangiopancreatographic imaging analyses, in some cases it is difficult to determine the cause of pancreatitis. In such patients, genetic testing may be helpful ^(3,4). Recent data have demonstrated a high prevalence of genetic mutations in pediatric patients with pancreatitis ^(5,6).

Four decades after Comfort and Steinberg ⁽⁷⁾ described the familial tendency of pancreatitis, the activation of trypsinogen to trypsin was shown to play a key role in the pathogenesis of pancreatitis ^(8–10). In addition, a mutation in the trypsinogen gene on chromosome 7q35 was implicated in pathogenesis ^(8–10) and was defined as an autosomal dominant trait with a penetrance of 80% ^(4,6). Various sites of cationic trypsinogen, serine protease 1 (PRSS1; OMIM 276000) mutations have since been reported. The serine protease inhibitor Kazal type 1 (SPINK1; OMIM 167790), which is encoded on chromosome 5q32 and is secreted from pancreatic acinar cells into the pancreatic juice, is a potent trypsin inhibitor that prevents the trypsin-catalyzed premature activation of zymogens in the pancreas and pancreatic duct. Mutations in the inhibitor gene are also thought to be associated with hereditary pancreatitis ⁽¹¹⁾.

Despite the potential importance of the topic, there have been few genetic studies on childhood pancreatitis. Therefore, we screened Korean children with acute recurrent pancreatitis (ARP) and chronic pancreatitis (CP) for mutations, or variants thereof, in the PRSS1 and SPINK1 genes.

PATIENTS AND METHODS

Selection of Patients and Controls

The institutional review board of the Asan Medical Center, Seoul, Korea, approved the present study. The study population consisted of 22 children with ARP and 19 with CP who were enrolled in our study between April 2006 and March 2009. The criterion for ARP included at least 2 documented episodes of acute pancreatitis in a patient without evidence of CP on imaging ⁽¹²⁾. CP was defined as persistent inflammatory disease of the pancreas, characterized by irreversible morphological changes and typically causing pain and/or permanent loss of function. Of 41 patients with ARP or CP, 32 patients consented to the genetic study and were enrolled. Of the 9 excluded patients, 5 had choledochal cysts and 3 had common bile duct stone, pancreas divisum, and autoimmune pancreatitis, respectively. We failed to determine etiology in 1 patient. Twenty-eight healthy children of similar age were selected as a control group.

Clinical Assessment

We retrospectively reviewed sex; onset age; accompanying pancreatic or biliary disorders; any history of current infection or trauma; and laboratory findings such as levels of amylase, lipase, aspartate aminotransferase, alanine aminotransferase, total bilirubin, alkaline phosphatase, and hemoglobin. Anatomic abnormalities were demonstrated by ultrasound, computed tomography, magnetic resonance cholangiopancreatography, or endoscopic retrograde cholangiopancreatography, and all of the patients underwent at least 1 of these studies. We also obtained the complete family history of pancreatitis in first-degree relatives of the study subjects.

Genetic Analysis

Genomic DNA was extracted from peripheral blood leukocytes using PUREGENE DNA isolation kits (Gentra, Minneapolis, MN) from 32 patients and 28 healthy children. The coding region and flanking intergenic sequences of the PRSS1 gene (NT_007914.14) and the SPINK1 gene (NT_029289.10) were sequenced using an ABI3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA).

Statistical Analysis

Patients were categorized into 3 groups based on the presence of a mutation or variant in either PRSS1 or SPINK1 genes. Mutation frequencies were determined. The frequencies of mutant or variant alleles and clinical and laboratory data were compared among the groups using the χ² test, Fisher exact test, and/or the Kruskal-Wallis test. All of the analyses were performed using SPSS version 12.0 for Windows (SPSS Inc, Chicago, IL). P < 0.05 was considered statistically significant.

RESULTS

Incidence of Mutations

Among 18 ARP and 14 CP patients, 15 (46.9%) carried a mutation in 1 of the 2 genes analyzed (PRSS1 and SPINK1). However, none of the patients had mutations in both genes, and none of the control patients had mutations in either gene (Table 1). The patients who had 1 mutation were significantly more likely than the controls to have ARP or CP (P < 0.0001, odds ratio 50.49).

TABLE 1: Frequencies of PRSS1 and SPINK1 mutated alleles in patients with ARP, patients with CP, and control individuals

PRSS1 Mutations

Four of the 32 patients (12.5%) with ARP or CP had missense mutations in the PRSS1 gene; however, none of the 28 control patients had a PRSS1 mutation. We identified heterozygous p.N29I c.89A>T (p.N29I), p.R122H c.365G>A (p.R122H), and c.86A>G (p.N29T) mutations, and ac.623G>C (p.G208A) variant (Table 1).

SPINK1 Mutations

The SPINK1 gene mutations were identified in 11 of the 32 patients (34.4%) with ARP or CP and none of the control patients. The splicing mutation, c.194 + 2T>C (IVS3 + 2T>), in the SPINK1 gene was observed in 10 patients with an allele frequency of 54.5% (ie, 12 of 22 alleles). Eight of the 10 patients with the above mutation were heterozygous and 2 were homozygous. The remaining patient had a heterozygous c.101A>G (p.N34S) missense mutation.

Clinical Aspects

Seventeen of the 32 patients presented with unknown etiologies and 15 patients presented with associated pancreaticobiliary disorders such as choledochal cysts (CDCs) (n = 4), biliary sludge/stones (n = 6), and pancreas divisum (n = 6). Among 14 patients with CP, 8 patients were idiopathic cases and 6 patients presented with associated pancreaticobiliary disorder. CP was more common in pancreas divisum (n = 4) than in CDC (n = 1) or biliary sludge/stone (n = 1). In idiopathic cases, the number of ARP and CP was equivalent. Genetic mutations were identified in 9 of the 17 patients with unknown disease etiology. None of the 4 patients with CDC and 2 of the 6 patients with biliary sludge or stones had mutations. Four of the 6 patients (66%) with pancreas divisum had genetic mutations, including a heterozygous p.N29I mutation in the PRSS1 gene (n = 1) and heterozygous (n = 1) and homozygous (n = 2) IVS3 + 2T>C mutation in the SPINK1 gene (Table 2). Each of the 4 patients with pancreas divisum who had CP with pancreatic duct stricture had a single mutation, whereas the remaining 2 patients with pancreas divisum with ARP did not have any mutation, although this difference was not significant (P = 0.067).

TABLE 2: Patients associated with pancreaticobiliary disorders according to genetic mutations

Among the 32 patients, the mean age of symptom onset was 8.2 ± 3.6 years (range 0.8–13.9 years). The mean age of symptom onset in the 17 patients without mutations, in the 4 patients with PRSS1 mutations, and in the 11 patients with SPINK1 mutations was 7.3 ± 3.4 years, 7.0 ± 4.3 years, and 9.9 ± 3.3 years, respectively. The patients with SPINK1 mutations showed an onset age similar to those in the other groups statistically (P = 0.128). Diagnosis, sex, and frequency of recurrence were similar in all 3 groups. Laboratory data were also similar among the 3 groups. Serum amylase, lipase, total bilirubin, and alkaline phosphatase levels were somewhat higher in patients without mutations compared with others, but this finding was not statistically significant.

Two patients with CP had a family history of pancreatitis. One patient with a heterozygous PRSS1 mutation (p.N29I) had a father and grandfather who had experienced CP. The other patient had a father, grandmother, sibling, and cousin who had experienced acute or CP, whereas genetic testing for PRSS1, SPINK1, and the full sequences of 27 exons and exon-intron boundaries of CFTR (cystic fibrosis transmembrane conductance regulator) gene, which is also known as another causative gene in CP ⁽¹³⁾, showed normal results.

DISCUSSION

Mutations in the PRSS1 gene are thought to play an important role in hereditary pancreatitis via modulation of trypsin activity ^(8,9). We have identified identical heterogeneous p.R122H, p.N29I, and p.N29T mutations in the PRSS1 genes of children with ARP and CP. In the present study, we identified p.G208A variation of the PRSS1 gene in a child with CP. We recently examined another 7-year-old patient with p.G208A variant in the PRSS1 gene, who suffered from severe necrotizing pancreatitis with no explainable cause (unpublished data). Several reports have suggested that this variant plays a minor role in the development of ARP or CP when mutations in the CFTR gene are present ^(14,15). However, our 2 children with this variation did not possess the cystic fibrosis phenotype and they experienced CP and severe acute necrotizing pancreatitis without any other explainable cause.

In the present study, SPINK1 mutations were more common than PRSS1 mutations in patients with ARP or CP. The most common SPINK1 mutation identified in our study was IVS3 + 2T>C (allele frequency 54.5%), consistent with a recent study of Korean adults. The prevalence of the SPINK1 IVS3 + 2T>C mutation is 26.8% in adult patients with idiopathic and familial pancreatitis ⁽¹⁴⁾. The p.N34S and IVS3 + 2T>C mutations in the SPINK1 gene are most frequently observed in Japan ⁽¹⁶⁾. The prevalence of the SPINK1 IVS3 + 2T>C splicing mutation is presumably more common in Asia than in Western countries ^(5,15,16), and the prevalence of the p.N34S mutation was lower in our population than in a Japanese population ⁽¹⁶⁾.

Pancreatitis in Korean children differed in a way from that in Western populations. Although cystic fibrosis is a known cause of pancreatitis in Western countries, cystic fibrosis itself is rare in Asian people, and our 32 patients did not have cystic fibrosis phenotypes. In addition, most children with mutations in the PRSS1 or SPINK1 gene did not have a family history of pancreatitis.

Interestingly, 4 of our 6 patients with pancreatic divisum had at least 1 mutation, whereas mutations were not identified in patients with choledochal cysts (Table 2). The 4 patients with both pancreas divisum and a genetic mutation also had CP accompanying pancreatic duct stricture rather than ARP. Only a few studies have investigated the association between pancreatic divisum and genetic mutations. Garg et al ⁽¹⁷⁾ recently reported that the patients with pancreas divisum presenting with idiopathic pancreatitis had a higher frequency of SPINK1 mutation compared with healthy controls, and the mean age of the patients with pancreas divisum and SPINK1 mutation was relatively less than that of patients without mutation. In another study, analysis of CFTR function in patients with pancreas divisum and ARP showed a significant difference in CFTR function between patients and healthy control individuals ⁽¹⁸⁾. Dray et al ⁽¹⁹⁾ also reported 2 patients with ARP with pancreatic divisum possessing the IVS8-5T-12TG allele of the CFTR gene. One Polish ⁽⁵⁾ study described 4 patients with pancreatic divisum, each of which had a PRSS1, CFTR, or SPINK1 mutation. These previous reports and our current findings suggest that the mutations may contribute to the early development of CP in patients with pancreas divisum. The prevalence of pancreatic divisum among normal individuals is thought to range from 5% to 10%, based on autopsy results ⁽²⁰⁾. Pancreas divisum is asymptomatic in most cases, but pancreatitis sometimes develops. Therefore, genetic study is warranted for patients with ARP or CP and pancreas divisum. Despite the absence of statistical significance due to our small study population, in our study the patients with pancreas divisum and a genetic mutation were more likely to suffer from CP than from ARP; specifically, these mutations in patients with pancreas divisum may lead to early and irreversible morphological pancreatic changes. In general, as seen in our research, pancreas divisum can result in both ARP and CP. Therefore, this finding may support the concept of “progression”; in other words, ARP and CP are not different diseases but rather different stages of the same disease, although ARP is generally thought of as actually CP ab initio ^(12,21).

The limitation of our study was the small number of patients enrolled in each group. It is also not clear whether these genes, especially the SPINK1 gene, are causative factors of pancreatitis or merely bystanders in the progression of pancreatitis. In addition, we did not find any statistically significant clinical characteristics caused by the existence of mutations in these genes. Furthermore, we did not investigate other possible candidate genes of CP or ARP, such as CFTR, chymotrypsinogen C (CTRC), α1-antitrypsin, anionic trypsinogen gene, mesotrypsinogen gene, and calcium sensing receptor (CASR) genes in our study. Because pancreatitis may result from an imbalance between proteases and their inhibitors, genetic study of these genes may still be necessary. Nevertheless, our study indicates several important findings: First, the PRSS1 and SPINK1 mutations are common among Korean children with ARP and CP; second, we noted a relatively high incidence of mutations in patients with pancreas divisum; and finally, mutations in pancreas divisum patients may lead to chronic changes at an early age.