Therapeutic Endoscopic Retrograde Cholangiopancreatography in Pediatric Patients with Acute Recurrent and Chronic Pancreatitis: Data from the INSPPIRE (INternational Study group of Pediatric Pancreatitis: In search for a cuRE) Study

Abstract

Objectives

To characterize utilization and benefit of therapeutic endoscopic retrograde cholangiopancreatography (ERCP) in children with acute recurrent pancreatitis (ARP) or chronic pancreatitis (CP).

Methods

From August 2012 to February 2015, 301 children with ARP or CP were enrolled in the INSPPIRE study. Physicians reported utilization and benefit of therapeutic ERCP at enrollment. Differences were analyzed using appropriate statistical methods.

Results

117 children (38.9%) underwent at least one therapeutic ERCP. The procedure was more commonly performed in children with CP compared to ARP (65.8% vs 13.5%, p<0.0001). Utility of therapeutic ERCP was reported to be similar between ARP and CP (53% vs 56%, p=0.81) and was found to be helpful for at least one indication in both groups (53 of 99 patients, 53.5%). Predictors for undergoing therapeutic ERCP were: presence of obstructive factors in ARP and CP, Hispanic ethnicity, or white race in CP.

Conclusions

Therapeutic ERCP is frequently utilized in children with ARP or CP and may offer benefit in selected cases, specifically if ductal obstruction is present. Longitudinal studies are needed to clarify the efficacy of therapeutic ERCP and to explore subgroups that might have increased benefit from such intervention.

INTRODUCTION

Acute pancreatitis (AP) in children was considered to be uncommon, but its incidence is now recognized to approach that reported in adults.^1–3 While many children recover without long-term effects, a subset of patients will have further episodes of pancreatitis and/or develop chronic pancreatitis (CP). With advanced non-invasive imaging modalities, endoscopic retrograde cholangiopancreatography (ERCP) is infrequently utilized solely as a diagnostic modality. However, ERCP is frequently applied as a therapeutic modality to manage patients with sequelae of acute recurrent pancreatitis (ARP) and CP. Endoscopic intervention such as therapeutic ERCP have traditionally been undertaken after failure of non-invasive maneuvers such as eliminating possible instigating environmental factors, or when medication trials are ineffective and before more invasive surgical interventions are pursued.

The literature regarding ARP and CP in the pediatric population is sparse, and even less data are published on appropriate interventions to ameliorate or prevent recurrences of pancreatitis or consequences of CP.^4–7 Given its favorable benefit/risk profile compared with surgical interventions, therapeutic ERCP is still recommended by most experts to be the first reasonable approach to treat select patients, for example with CP.⁸ Collaborative efforts among adult providers have led to significant advances in understanding etiologies and clinical outcomes from surgical and endoscopic interventions in adult patients with ARP and CP.^9,10

The INSPPIRE (INternational Study group of Pediatric Pancreatitis: In search for a cuRE) consortium was formed to address gaps in knowledge of pediatric ARP and CP. The present study analyzes baseline information entered on patients in this database to describe the utilization and efficacy of endoscopic intervention on pediatric patients with ARP or CP as reported by the providers responsible for patient enrollment. We identify predictors of therapeutic ERCP utilization and report its utility.

MATERIAL AND METHODS

Study Design and Participants

Demographic and clinical data were collected from 16 institutions using patient/parent and physician questionnaires for children who fulfilled the criteria for ARP or CP and were ≤19 years of age at the time of enrollment, as previously described.¹¹. Patient information was entered into the REDCap^™ (Research Electronic Data Capture¹², Vanderbilt University, Nashville, TN) database from September 2012 to July 2015, and represented baseline information of the INSPPIRE cohort. All centers obtained Institutional Review Board approval or the equivalent for their country. These 301 subjects (155 ARP and 146 CP) have been reported in different contexts in previous reports.^13,14

Definitions

The definitions utilized for ARP and CP have previously been described.¹⁵ In summary, ARP was defined as at least 2 distinct episodes of AP with complete resolution of pain (>= 1-month pain-free interval between diagnoses of AP) or complete normalization of pain and serum pancreatic enzymes levels (amylase and lipase) before the subsequent episode of AP was diagnosed. CP was diagnosed when pancreatic imaging findings suggestive of CP were identified in the setting of either chronic pancreatic type pain, evidence of exocrine pancreatic insufficiency or endocrine pancreatic insufficiency. Suggestive imaging findings of CP included: ductal changes (irregular contour of the main pancreatic duct or its radicles; intraductal filling defects; calculi, stricture or dilation) or parenchymal changes (generalized or focal enlargement, irregular contour [accentuated lobular architecture], cavities, calcifications, heterogeneous echotexture). Imaging modalities may include computed tomography (CT), magnetic resonance imaging (MRI), magnetic resonance cholangiopancreagraphy (MRCP), ERCP; transabdominal ultrasound (US); or endoscopic ultrasound (EUS) (in which at least 5 EUS features [as defined by the Rosemont classification¹⁶] must be fulfilled).

Statistical Analysis

Differences were compared using Fisher’s exact test or chi square test for categorical variables depending on sample size and the student t-test for continuous variables. All differences were considered significant at a two-sided p<0.05. Data were analyzed utilizing SAS 9.4 (SAS Institute Inc., Cary, NC).

RESULTS

We analyzed the 301 patients entered into the INSPPIRE database as of July 2015; 155 (51.5%) were diagnosed with ARP and 146 (48.5%) with CP. Patient characteristics are presented in Table 1. Therapeutic ERCP was undertaken in 117 (38.9%) of patients.

TABLE 1.

Patient Characteristics

Total	N (%)
Age at enrollment, mean (SD), y	11.9 (4.6)
Age at first attack, mean (SD), y	8.7 (4.9)
Sex, female	172 (57)
Ethnicity, Hispanic	63/276 (23)
Race, n = 265
White	214 (81)
Multi-racial	20 (8)
Black	10 (4)
Asian	14 (5)
Other	7 (3)
ARP	155 (52)
CP	146 (48)
Exocrine insufficiency	46/259 (18)
Endocrine insufficiency	13/278 (5)
Risk factors
Toxic-metabolic	23/195 (12)
•Medications	7/287 (2)
•Ethanol exposure	4/287 (1)
•Active smoker	24/269 (9)
•Passive smoke exposure	30/220 (14)
Autoimmune	97/296 (33)
Obstructive factors
Genetic risk factors
•PRSS1	65/196 (33)
•SPINK1	35/177 (20)
•CFTR	54/193 (28)
•CTRC	9/121 (7)
Therapies Attempted	194/294 (66)
•Medical therapies	145/289 (50)
•Therapeutic ERCP utilization	117/301 (39)
•Surgery (non-TPIAT)	61/292 (21)
•TPIAT	29/291 (10)

SD: standard deviation, ARP: acute recurrent pancreatitis, CP: chronic pancreatitis, PRSS1: protease, serine 1, SPINK1: serine peptidase inhibitor, Kazal type 1, CFTR: cystic fibrosis transmembrane conductance regulator, CTRC: chymotrypsin C, ERCP: endoscopic retrograde cholangiopancreatography, TPIAT: total pancreatectomy and islet autotransplantation.

Of the 155 children with ARP, 21 (13.5%) underwent therapeutic ERCP. The results of a univariate analysis evaluating predictors for application of therapeutic ERCP in patients with ARP are included in Table 2. Children suspected to have obstructive factors contributing to their disease process were significantly more likely to undergo therapeutic ERCP.

TABLE 2.

Predictors of Therapeutic Endoscopy Utilization in Pediatric ARP

Variables	Underwent therapeutic-ERCP, n=21 n (%)	No therapeutic-ERCP, n=134 n (%)	P
Sex, female	14 (67)	74 (55)	0.325
Ethnicity, Hispanic	(n=20) 11 (55)	(n=125) 93 (74)	0.074
Race	(n=17)	(n=121)
White	14 (82)	98 (81)	0.414
Multi-racial	3 (18)	9 (7)	–
Black	0 (0)	5 (4)	–
Asian	0 (0)	7 (6)	–
Other	0 (0)	2 (2)	–
Age at first acute attack, mean (SD), y	(n=18) 9.1 (5.1)	(n=122) 9.0 (5.0)	0.967
Exocrine insufficiency	0/16 (0)	2/111 (2)	1.0
Endocrine insufficiency	0/19 (0)	6/122 (5)	1.0
Alcoholic	0/19 (0)	2/131 (2)	1.0
Active smoker	0/20 (0)	1/130 (1)	1.0
Passive smoking exposure	1/19 (5)	11/121 (9)	1.0
Autoimmune pancreatitis	3/13 (23)	13/99 (13)	0.394
Other autoimmune diseases	3/19 (16)	11/124 (9)	0.400
Genetic Risk Factors:
PRSS1	0/11 (0)	15/77 (19)	0.199
SPINK1	3/10 (30)	7/68 (10)	0.113
CFTR	3/10 (30)	27/79 (34)	1.0
CTRC	0/6 (0)	5/42 (12)	1.0
Obstructive factors	14/21 (67)	36/131 (27)	0.0004*
Pancreas divisum	4/20 (20)	9/126 (7)	0.081
Sphincter of Oddi disorders	3/20 (15)	2/124 (2)	0.019*
Gallstones	3/19 (16)	6/128 (5)	0.093
Pancreatic duct mal-union	2/20 (10)	5/126 (4)	0.245
Duct obstruction	0/20 (0)	1/128 (1)	1.0
Annular pancreas	0/20 (0)	3/128 (2)	1.0
Choledochal cyst	3/20 (15)	2/128 (2)	0.018*
Medications
Pain medications	7/13 (54)	34/102 (33)	0.217
Medical therapies	8/19 (42)	36/130 (28)	0.198
Pancreatic enzymes	7/19 (37)	25/130 (19)	0.130
Pancreatic enzymes (not for PI)	7/19 (37)	24/130 (18)	0.076
Vitamins/anti-oxidants	1/18 (6)	11/127 (9)	1.0
Steroids	0/18 (0)	0/125 (0)	–
Octreotide	0/19 (0)	2/128 (2)	1.0

^*Statistically significant.

ARP: acute recurrent pancreatitis, SD: standard deviation, PRSS1: protease, serine 1, SPINK1: serine peptidase inhibitor, Kazal type 1, CFTR: cystic fibrosis transmembrane conductance regulator, CTRC: chymotrypsin C, PI: pancreatic insufficiency, ERCP: endoscopic retrograde cholangiopancreatography.

Of the 146 patients with CP, 96 (65.8%) underwent therapeutic ERCP. Patients with CP were more likely to undergo therapeutic ERCP compared with ARP (65.8% vs 13.5%, p<0.0001). The results of a univariate analysis evaluating predictors for utilization of therapeutic ERCP in patients with CP were included in Table 3. Children who were of Hispanic ethnicity, white race, and those who were suspected to have obstructive factors contributing to their disease process were significantly more likely to have therapeutic ERCP. In contrast, patients with alcohol exposure or autoimmune pancreatitis were significantly less likely to have therapeutic ERCP. Table 4 shows benefit of various endoscopic interventions reported by the physician questionnaires. While therapeutic ERCP, particularly pancreatic interventions such as sphincterotomy, stone removal and stenting, was utilized more commonly in the CP group, its efficacy was not significantly different comparing patients with ARP and those with CP. Biliary stenting was infrequently performed in both cohorts.

TABLE 3.

Predictors of Therapeutic Endoscopy Utilization in Pediatric CP

Variables	Underwent therapeutic ERCP, n=96 (%)	No therapeutic ERCP, n=50 (%)	P
Sex, female	60 (62)	24 (50)	0.093
Ethnicity, Hispanic	(n=85) 19 (22)	(n=46) 3 (7)	0.021*
Race	(n=80)	(n=47)	0.010*
White	68 (85)	34 (72)	–
Multi-racial	4 (5)	4 (8)	–
Black	6 (6)	0 (0)	–
Asian	1 (1)	6 (13)	–
Other	2 (3)	3 (7)	–
Age at first acute attack, mean (SD), y	(n=81) 8.3 ± 4.6	(n=38) 8.1 ± 4.9	0.827
Age at first diagnosis CP, mean (SD), y	(n=72) 9.7 ±4.4	(n=32) 11.3 ± 4.6	0.106
Exocrine insufficiency	29/89 (33)	15/43 (35)	0.793
Endocrine insufficiency	6/93 (6)	1/44 (2)	0.429
Alcoholic	1/90 (1)	4/47 (9)	0.028*
Active smoker	1/90 (1)	2/47 (4)	0.271
Passive smoking exposure	9/84 (11)	3/45 (7)	0.540
Autoimmune pancreatitis	2/68 (3)	12/40 (30)	<0.0001*
Other autoimmune diseases	5/92 (5)	3/46 (7)	1.0
Genetic Risk Factors:
PRSS1	34/75 (45)	16/33 (48)	0.762
SPINK1	16/69 (23)	9/30 (30)	0.473
CFTR	14/71 (20)	10/33 (30)	0.233
CTRC	4/53 (8)	0/20 (0)	0.570
Obstructive factors	38/95 (38)	9/49 (18)	0.009*
Pancreas divisum	16/91 (18)	6/49 (12)	0.408
Sphincter of Oddi disorders	2/91 (2)	0/48 (0)	0.545
Gallstones	5/92 (5)	1/47 (2)	0.664
Pancreatic duct mal-union	5/91 (5)	1/48 (2)	0.664
Duct obstruction	4/93 (4)	0/48 (0)	0.299
Annular pancreas	1/93 (1)	0/48 (0)	1.0
Choledochal cyst	2/93 (2)	0/48 (0)	0.548
Medications
Pain medications	42/73 (58)	20/36 (56)	0.844
Medical therapies	72/94 (77)	29/46 (63)	0.093
Pancreatic enzymes	55/93 (59)	22/46 (48)	0.060
Vitamins/anti-oxidants	14/89 (16)	6/45 (13)	0.713
Steroids	3/90 (3)	5/45 (11)	0.116
Octreotide	4/89 (4)	1/46 (2)	0.661

^*Statistically significant.

CP: chronic pancreatitis, SD: standard deviation, PRSS1: protease, serine 1, SPINK1: serine peptidase inhibitor, Kazal type 1, CFTR: cystic fibrosis transmembrane conductance regulator, CTRC: chymotrypsin C, ERCP: endoscopic retrograde cholangiopancreatography.

TABLE 4.

Benefit of Endoscopic Interventions in Pediatric ARP and CP

Intervention	ARP, n = 155 n (%)	CP, n = 146 n (%)	P
Therapeutic ERCP performed	21 (14)	96 (66)	<0.0001*
Helpful for at least one indication	9/16 (56)	44/83 (53)	0.812
Biliary sphincterotomy performed	11/151 (7)	36/136 (26)	<0.0001*
Helpful	6/10 (60)	12/30 (40)	0.300
Biliary stenting performed	2/151 (1)	11/138 (8)	0.006*
Helpful	2/2 (100)	3/10 (30)	–
Pancreatic sphincterotomy performed	4/151 (3)	68/138 (49)	<0.0001*
Helpful	2/4 (50)	33/66 (50)	–
Pancreatic duct stenting performed	6/151 (4)	60/137 (44)	<0.0001*
Helpful	2/5 (40)	28/57 (49)	–
Pancreatic stone removal performed	1/151 (1)	30/137 (22)	<0.0001*
Helpful	1/1 (100)	18/18 (100)	–

^*Statistically significant.

ARP: acute recurrent pancreatitis, CP: chronic pancreatitis, ERCP: endoscopic retrograde cholangiopancreatography.

DISCUSSION

In our cohort, therapeutic ERCP was relatively infrequently performed in our patients with ARP, especially compared with patients with CP. When utilized in patients with ARP, therapeutic ERCP was done more frequently in patients with obstructive factors or pancreatic anatomical anomalies. Otherwise, therapeutic ERCP was not utilized as frequently in patients with other contributing variables causing pancreatitis. Over half of the patients with ARP who had a therapeutic ERCP performed were reported to have improvement of symptoms. Our study volume was similar to one of the largest pediatric ERCP studies for pediatric pancreatitis by Agarwal et al, who described 221 ERCPs in 172 children, of whom 19 had ARP.⁶ Of these 19 children, 63% were reported to have improvement of symptoms after therapeutic ERCP, similar to our observed rate of improvement of 56%. Otto et al also evaluated a large series of pediatric ERCP for pancreatitis and found a similar rate of children with anatomical abnormalities (13/106) contributing to recurrent pancreatitis, but their study did not evaluate the effectiveness of ERCP guided endotherapy for relief of symptoms in children with ARP or CP.¹⁷ Cote et al demonstrated that biliary sphincterotomy alone may be as effective as dual biliary and pancreatic sphincterotomy to prevent AP episodes in adult patients with idiopathic ARP.¹⁸ Unfortunately, the small number of sphincterotomies performed in our ARP cohort did not allow us to perform an analogous analysis, but the reported utility of biliary and pancreatic sphincterotomies in our cohort were similar (60% for biliary vs 50% for pancreatic). From our data, although therapeutic ERCP was used much less frequently in children with ARP compared with CP, its benefit in children with ARP was similar to those with CP. Whether therapeutic ERCP should be utilized at a greater frequency in managing children with ARP remains to be determined, but evidence should best come from longitudinal prospective studies evaluating the outcomes of such intervention.

Similar to the pediatric INSPPIRE consortium, several adult centers in a collaborative effort formed the North American Pancreatitis Study 2 (NAPS2) and first published on their collective data on ARP and CP in adult patients in 2008.⁹ This was soon followed by a focused analysis of the NAPS2 data assessing the pattern of use of therapeutic ERCP and surgical therapies and their perceived effectiveness for only CP as based on completed physician questionnaires.¹⁰ In the present study from the pediatric-based INSPPIRE consortium, a similar analysis was performed yielding specific data for CP, but also included ARP (Tables 2 thru 4), but limited to only therapeutic ERCP. Several key comparisons between the current INSPPIRE CP data and that from the NAPS2 cohort are notably important to highlight. Despite notable differences in the mean ages of the patients (12 yrs in INSPPIRE vs 49 yrs in NAPS2) and primary identified risk factors (genetics/obstructive factors in INSPPIRE vs alcohol in NAPS2) the utilization of therapeutic ERCP (66% for INSPPIRE vs 61% for NAPS2) and the perceived utility of this therapy (53% for INSPPIRE vs 43% for NAPS2) were remarkably similar for patients with CP in both cohorts.¹⁰ Interestingly, biliary interventions were performed more often in the adult NAPS2 cohort compared with the current pediatric data (biliary sphincterotomy: 42% vs 26 %, biliary stenting: 14% vs 8%) possibly reflective of the less frequent occurrence of CP-related biliary strictures encountered in pediatric patients. In contrast, pancreatic interventions such as pancreatic stenting and stone removal were performed more often in the INSPPIRE cohort (44% and 22%, respectively) compared with the NAPS2 cohort (36% and 11%, respectively) possibly reflecting the fact that obstructive pathology within the pancreatic duct were more frequently identified in pediatric cohort as well. These differences in therapeutic ERCP utilization between adult and pediatric cohorts deserve further study, as they may provide clues into differences in underlying mechanisms of disease, disease progression, and which subgroups of patients may be more amenable to specific endoscopic interventions. These differences are also important to consider as endoscopists performing therapeutic ERCP in children attempt to apply the relatively abundant adult literature in this area to pediatric patients in which a paucity of literature currently exists for the purpose of guiding management decisions.

In uncomplicated CP, several adult studies have shown that therapeutic ERCP can be helpful in improving the chronic pain associated with this disease.^19–21 The technical goals of therapeutic ERCP in CP typically include alleviating pancreatic ductal obstructions due to stones or strictures, stenting of the main pancreatic duct and treatment of regional complications such as peri-pancreatic fluid collections and strictures of the intra-pancreatic portion of the bile duct. Clinical guidelines by the European Society of Gastrointestinal Endoscopy suggest that therapeutic ERCP should be considered as a first-line treatment for patients with uncomplicated CP.⁸ However, others advocate for reserving this intervention for those with imaging-proven main pancreatic duct obstruction.²² In our study cohort, a majority of children with CP who undergo therapeutic ERCP are suspected to have pancreaticobiliary obstruction. Adult studies have also shown that benign biliary strictures, such as those seen in CP, are very responsive to endoscopic stenting, with response rates between 80–90%.^23–25 As mentioned previously, this intervention was rarely undertaken in our pediatric cohort and curiously only 3 of the 10 children with CP who had biliary stents placed were reported to benefit from this intervention. We are unable to account for this discrepancy; however, it is possible that the diminutively sized biliary tree of the pediatric-aged patients may preclude some of the effective stenting therapies developed in adults (i.e. placement of multiple plastic or fully covered metal biliary stents).

As previously described by the INSPPIRE consortium, genetic mutations in pancreatitis-related genes are frequently identified in pediatric ARP and CP.¹³ Limited retrospective studies have suggested that therapeutic ERCP may be effective at decreasing rates of recurrent pancreatitis and help manage chronic pain for these patients.^5,7 Although therapeutic ERCP offers an effective alternative to treat the disease sequelae in this subgroup, it does not address the underlying genetic defect. Therefore, care should be taken in equating technical procedural success with improvements in pain and quality of life. The relatively invasive nature of therapeutic ERCP and the potential for adverse events invokes the need for ongoing studies to examine the long term effects of these procedures in children with ARP and CP. Unfortunately, the retrospective nature of our data collection regarding pain did not allow us to comment on patient or family perceptions as to whether or not therapeutic ERCP led to any tangible benefits in this regard. However, the majority of physician respondents believed that therapeutic ERCP was helpful for their patients although this result may be skewed due to bias on behalf of the provider. Clearly, prospective studies are still warranted to determine the effects of therapeutic ERCP on pain and quality of life in children with ARP and CP. Of particular interest is elucidating where in the treatment paradigm therapeutic ERCP should be considered for various subgroups of patients in relation to other potential treatment modalities such as medications, extracorporal shockwave lithotripsy, surgical interventions and total pancreatectomy with auto islet cell transplantation.

This study has several significant limitations that need to be taken into consideration. Subjects were recruited from tertiary care centers which may introduce referral bias towards children with more severe and disabling forms of ARP and CP. Another limitation is the subjective nature of procedural effectiveness that was utilized. In addition, the timing of the intervention in relation to the assessment of procedural effectiveness was not standardized as all patients were analyzed at enrollment. It is hoped that by incorporating annual follow-up questionnaires as well as more objective measures of effectiveness such as rates of recurrent pancreatitis, pain medication utilization, school days missed among others into the longitudinal INSPPIRE database will help overcome these limitations as the consortium continues its study of pediatric ARP and CP in a systematic manner. Nevertheless, the large number of well-phenotyped pediatric patients with ARP and CP allowed us to assess therapeutic ERCP utilization in these patients and gain insight into its perceived effectiveness according to providers.

In summary, this study is unique in that it is the first multicenter assessment of therapeutic ERCP utilization and perceived benefit in pediatric patients with ARP and CP. We document that therapeutic ERCP is more commonly undertaken in patients with CP compared with ARP, and in those identified to have obstructive factors contributing to their disease burden. Further systematic investigation is needed in this area, including prospective comparative studies evaluating therapeutic interventions in ARP and CP incorporating assessment of outcomes including qualify of life after these interventions.