Indian Pediatrics 2004; 41:704-711
Pancreatic Diseases in Children in a North Indian Referral Hospital
Srikant Das, N.K. Arora, D.K. Gupta*, A.K. Gupta#, Prashant Mathur and Anjali Ahuja
From the Departments of Pediatrics, *Pediatric Surgery and #Radiodiagnosis, All India Institute of Medical Sciences, New Delhi 110 029, India.
Correspondence to: Dr. N.K. Arora, professor, 3063, Department of Pediatrics, All India Institute of Medical Sciences, New Delhi 110 029, India. E-mail: email@example.com
Manuscript received: January 24, 2002, Initial
review completed: April 19, 2002;
There is a paucity of literature regarding the profile of pancreatic disorders in children(1-4). Buntain, et al.(1) after a review of pancreatitis in childhood over a 15 years period concluded that pancreatitis was not rare in childhood and had a significant over all mortality rate of 30% or greater as compared to 12% in adults(5). At least three other series between 1965 and 1992 reported a mortality rate ranging from 12 to 26% of pancreatitis in childhood(2-4). The divergence of opinion between authors regarding etiology, patho-physiology, definitions and classification of pancreatic disorders and non-uniformity of vocabulary makes it difficult to compare different studies(6-9).
Disease entities like tropical pancreatitis and malnutrition related diabetes mellitus are almost exclusively found in tropical countries, with probably the highest prevalence in India (6,7,10). From our country, only case reports of various pancreatic disorders in children are available(11-14). The present study was conducted to determine the profile of pancreatic disorders in children presenting to our center, between 1994 and 2001.
Subjects and Methods
This study included children who were diagnosed to have pancreatic disorders at the Division of Pediatric Gastroenterology, Hepatology and Nutrition of AIIMS during the years 1994-2001.
Pancreatic disorders were suspected if the child had abdominal pain associated with raised serum lipase or amylase or/and suggestive ultrasound (USG)/CT scan abdomen/ERCP (endoscopic retrograde choliangio pancreatography) showing altered echogenecity, change in size, shape, duct caliber or calcification of pancreas. These patients were worked up further for assigning etiologic label according to a pre-defined protocol and classified as having acute pancreatitis, chronic pancreatitis, hereditary pancreatitis and congenital anomalies of the pancreas. Medical / surgical conditions that were / could potentially be causing pancreatic pathology with temporal association were labeled as etiologic factors. Medical / Surgical condition(s) that were present in addition to etiologic factor(s) and/or those conditions which were unlikely to be cause of pancreatic pathology in the specific patient, were labelled as associated conditions.
Patient work up
Children with suspected pancreatic disorders were worked up according to a uniform protocol. The pre-tested form consisted of demographic features, and clinical (history and examination) details of patients. This included a two generation family history, history of trauma to abdomen, existence of other systemic disorders (like cystic fibrosis, celiac disease and others), exposure to drugs, toxins, recent infections (particularly clinical/laboratory evidence of salmonella typhi, mumps and hepatitis B) and metabolic disorders.
Following investigations were undertaken as and when required and feasible to establish presence of pancreatic disorder and its potential etiologies.
Blood: Serum amylase, C-reactive protein, lipid profile, serum calcium, renal and liver function tests, blood sugar and glucose tolerance test (GTT).
Radiologic: Plain X-ray abdomen, abdominal ultrasonography, contrasts enhanced CT-scan of abdomen and magnetic resonance cholangiopancreatography (MRCP).
Endoscopy: Upper gastrointestinal endosopy and ERCP.
Other investigations: Appropriate laboratory tests to confirm specific etiologic factors were done according to the clinical indications. Facilities for serum lipase estimation were available only during later part of the study and hence could be done in 15 patients (7 acute pancreatitis and 8 chronic pancreatitis). As per published evidences, we used a cut off value of upper limit of normal (>165 IU/L) to be suggestive of pancreatitis which is reported to have sensitivity and specificity in the range of 86.5% and 100%(15). Similarly, glucose tolerance test (12 chronic pancreatitis) and 24 hour fecal fat estimation (8 chronic pancreatitis) were systematically done during the last two years of the study only.
Sweat chloride >60mEq/L on two occasions was considered as diagnostic of cystic fibrosis. Alpha-l antitrypsin (al AT) deficiency was diagnosed on the basis of absence of a-globulin band on simple gel electrophoresis and serum a1 AT levels below 60 mg/dL. Screening for a1AT deficiency was done among patients recruited after 1998, and phenotyping was not available.
Stata-6.0 software (STATA Corpn, Texas, USA) was used for statistical analysis. Descriptive statistics were computed for baseline demographic, clinical and laboratory features according to etiology. Frequency tables were generated to reflect the etiology of pancreatic disorders in childhood.
During the study period, 54 cases of pancreatic disorders presented to our center. Acute (28/54, 51.9%) and chronic (23/54, 42.5%) pancreatitis constituted most of the patients with pancreatic disorders. De novo presentation as pseudocyst was identified in a single case though pseudocyst was associated as a complicating feature in nearly 25% cases of chronic pancreatitis (6/23) and 41% (11/28) of acute pancreatitis. One girl had pancreato-blastoma while another boy presented with a pancreatic mass due to Nocardia infection.
Etiology of 28 children with acute pancreatitis is summarized in Table I.
TABLE I Etiological Factors of Acute Pancreatitis in Children (N = 28)
*Prednisolone, Carbamazepine, Phenytoin, L-Asparaginase # Acute HAV Infection
Ten children (35.7%) experienced more than one episode of acute pancreatitis over a period of time (0-18 months). Recurrent episodes of acute pancreatitis occurred up to 24 months of follow up in 10 patients. None of these children had manifest clinical or laboratory features of chronic pancreatitis. Three of these cases with recurrent acute pancreatitis were due to underlying alpha-l antitrypsin deficiency, hypertriglyceridemia, and hereditary pancreatitis.
Pulmonary tuberculosis(2), seizure disorder(2), acute leukemia(1), hypertension (1) and HBsAg carrier state without liver dysfunction(1) were the incidental conditions identified in 7 children (25%) that did not appear to have any direct contribution to the development of acute pancreatitis. The child with acute leukemia developed acute pancreatitis after treatment for one month with steroids and L-asparaginase. No complications were observed in 61 % (17/28) patients. Pseudocyst was the commonest complication (11/28; 39%) of acute pancreatitis. Splenic vein thrombosis was observed in a child who also had pseudocyst formation.
All but one patient required hospitalization. Thirteen children required more than one hospitalization because of acute pancreatitis and/or its complications. Laprotomy was required in 9/28 (32.1 %) patients and in seven of them pancreatic pathology was documented. One child (3.6%) with acute pancreatitis died during the acute phase of illness.
Demographic and clinical profile of these cases is depicted in Table II.
TABLE II Demographic and Clinical Profile of Acute Pancreatitis in Children.
*No. of patients in whom information was available.
The laboratory parameters for acute pancreatitis were fulfilled in most of the patients. Elevated serum amylase (>200 IU/L) and/or ultrasound evidence of pancreatitis was present in 21/22 (95%) cases. Abnormal CT scan and/or amylase >200 IU/L was documented in 19/21 (90%) patients.
Etiology of 23 children with chronic pancreatitis is shown in Table III. Both children with hereditary pancreatitis were born out of non-consanguineous marriages. Although no records were available, according to historical details the father and paternal uncle of the first child were also suffering from chronic pancreatitis and were on pancreatic supplements. The mother of second child had chronic pancreatitis and died due to illness. The child with a1-AT deficiency was product of a consanguineous marriage and did not have elevated hepatic transaminases. No liver biopsy was done in this child.
TABLE III Etiological Factors of Chronic Pancreatitis in Children (N = 23).
Comparison between the categories of Chronic Calcific and Chronic Non-calcific Pancreatitis: Fischer’s Exact Test = 0.134, * Valproate intake for more than 6 years.
In 4 out of 23 patients (17.4%), there were associated conditions viz., tuberculosis, seizure disorder and hepatitis B. However there was no evidence that these could be contributing to the development of chronic pancreatitis.
Pseudocysts were identified in six children (25%), four of whom were having chronic non-calcific pancreatitis. Vascular complica-tions viz., mesenteric vein thrombosis(1), splenic vein thrombosis, splenic artery aneurysm(2) and gastroduodenal artery aneurysm(1) were observed in four children with chronic pancreatitis.
The clinical and demographic profile of calcific and non-calcific chronic pancreatitis was similar (Table IV). Weight for age z-scores of £ 2 were also comparable in the two groups. Steatorrhea (24 hour fecal fat >6g) was documented in one out of eight consecutive patients in whom the test was done. Similarly, abnormal GTT was present in 33.3% (4/12) patients.
TABLE IV Demographic and Baseline Features of Children with Chronic Pancreatitis
† No. of patients in whom information available; *Fisher’s Exact Test and Mann WhitneyU Rank Sum Test used for comparison.
The profile of pancreatic disorders as reported in this study was similar to that reported by Vane, et al.(16) and from Boston Children’s Hospital (17). In both these studies, trauma, congenital malformations and pseudo-cyst were listed separate from pancreatitis as etiologic categories. We however included these as factors contributing to development of pancreatitis. In both these case series(16,17) tumors contributed to 10% of all pancreatic disorders. This was higher than that reported in the present study (3.8%). At our institution, there is a separate center for cancers and hence pancreatic tumors were likely to be under represented. Our findings of acute pancreatitis resulting from diverse etiology are similar to that reported earlier(2-4, 18-20). Two of our patients had generalized tuberculosis with lymph nodes around pancreatic head. One of them also had ascites which showed evidence of tuberculous antigen on PCR. During follow up, these patients recovered completely on anti-tuberculous therapy. Mortality due to acute pancreatitis in the present study was 3.6%-as compared to 0-26% reported earlier (1-4,18,19). Increased prevalence of pseudo-cyst in our series (41%) as compared to 16-22% reported by others (18-20) may be explained by delay in seeking care and or referral to a tertiary center of patients with severe disease or on development of complications.
Recurrent episodes of acute pancreatitis occurred in ten patients (35.7%). It is possible that on follow up some of these patients might develop overt features of chronic pancreatitis. Three of these cases were due to underlying alpha-l antitrypsin deficiency, hypertrigly-ceridemia and hereditary pancreatitis, all of which have been reported to cause repeated episodes of acute pancreatitis(20). In remaining seven patients, no etiology could be determined.
In the present study, 65% (15/23) of chronic pancreatitis patients had pancreatic calcification and remaining 35% without any calcification in pancreas. Etiology in our series was comparable to that reported by Prasad, et al.(21). Two of our patients developed chronic pancreatitis due to prolonged pressure on pancreatic ducts/pancreas, caused by tuberculous lymph nodes and cholelithiasis in one case each.
In one of the early studies on chronic pancreatitis from India, Mehta, et al.(22) reported six cases of cystic fibrosis. All these patients had gastrointestinal symptoms. In the present study there was a single child with non-calcific chronic pancreatitis and high sweat chloride on two occasions (68 mEq/L and 75 mEq/L). This child had predominantly gastrointestinal symptoms and no respiratory involvement. Our study was undertaken in the gastroenterology division hence it was possible that cystic fibrosis patients with predominant pulmonary symptoms and pancreatic involvement but without gastro-intestinal manifestations were missed in the current study.
Both acute and chronic pancreatitis appeared to be diseases of older children. In previous studies also most of the children were in the age ranges of 3-16 years (18-20) with mean age varying between 7.4 years to 10.2 years. The proportion of children below 5 years was small in most previous studies.
It was possible that 13 (56.5%) idiopathic chronic pancreatitis patients including 9 with calcification might have included children with tropical pancreatitis. The etiology of tropical pancreatitis remains unknown(23) but in India and other geographic regions where the condition has been described it could be a distinct possibility if other known metabolic as well as acquired conditions causing chronic pancreatitis have been ruled out. Our youngest patient was an 11-month-old girl who did not have any identifiable congenital pancreato-biliary or metabolic abnormalities and was labeled as idiopathic chronic pancreatitis. More extensive work up for known infectious diseases and metabolic disorders is necessary to reduce the proportion of patients with indeterminate etiology.
Contributors: NKA designed and coordinated the study. NKA, SD, PM, AA and DKG were responsible for data collection. AKG provided the radiological support. Data were analyzed by NKA and SD. NKA, SD and PM were involved in the manuscript preparation and editing. NKA will be the guarantor.
Competing interests: None stated.