Successful treatment of pediatric IgG4 related systemic disease with mycophenolate mofetil: case report and a review of the pediatric autoimmune pancreatitis literature

Autoimmune pancreatitis is frequently associated with elevated serum and tissue IgG4 levels in the adult population, but there are few reports of pediatric autoimmune pancreatitis, and even fewer reports of IgG4 related systemic disease in a pediatric population. The standard of care treatment in adults is systemic corticosteroids with resolution of symptoms in most cases; however, multiple courses of corticosteroids are occasionally required and some patients require long term corticosteroids. In these instances, steroid sparing disease modify treatments are in demand. We describe a 13-year-old girl with IgG4 related systemic disease who presented with chronic recurrent autoimmune pancreatitis resulting in surgical intervention for obstructive hyperbilirubinemia and chronic corticosteroid treatment. In addition, she developed fibrosing medianstinitis as part of her IgG4 related systemic disease. She was eventually successfully treated with mycophenolate mofetil allowing for discontinuation of corticosteroids. This is the first reported use of mycophenolate mofetil for IgG4 related pancreatitis. Although autoimmune pancreatitis as part of IgG4 related systemic disease is rarely reported in pediatrics, autoimmune pancreatitis is also characterized as idiopathic fibrosing pancreatitis. All pediatric autoimmune pancreatitis cases reported in the world medical literature were identified via a PUBMED search and are reviewed herein. Twelve reports of pediatric autoimmune pancreatitis were identified, most of which were treated with corticosteroids or surgical approaches. Most case reports failed to report IgG4 levels, so it remains unclear how commonly IgG4 related autoimmune pancreatitis occurs during childhood. Increased evaluation of IgG4 levels in patients with autoimmune pancreatitis may shed further light on the association of IgG4 with pancreatitis and the underlying pathophysiology.


Background
The concept of immune mediated pancreatitis was first suggested by Sarles et al. in 1961 after observing massive lymphoplasmocytic infiltrates and fibrosis on pancreatic histology in patients with pancreatitis of unknown etiology [1]. Autoimmune pancreatitis (AIP) was further documented and characterized by Yoshida et al. in 1995 [2] with 11 common characteristics that have become the basis for diagnostic criteria of AIP [3][4][5] (Tables 1  and 2). These criteria are based on characteristic changes seen on imaging, pathology, serology, and disease modification (and usually resolution) with corticosteroid treatment [2][3][4][5]. When Hamano et al. identified an abnormal fraction of gamma globulins in patients with AIP in 2001, elevated serum IgG4 level became the hallmark lab value for AIP in adults [6]. Since then there have been many reports of adults with isolated AIP and multisystem sclerosis with elevated IgG4 levels [7]. However, there have been very few reports of pediatric patients with AIP [8][9][10][11][12][13], and all of those have had normal or not reported IgG4 levels. There is one report of a pediatric patient with elevated IgG4 levels; however, she did not have imaging changes or histology available to make a more definitive diagnosis of AIP [14]. Herein, an adolescent case of AIP [by Japanese Pancreas Society (JPS) 2006 criteria [4] (Table 1) and Mayo Clinic HISORt criteria [3,5] (Table 2)] with elevated IgG4 levels and evidence of systemic disease treated successfully with mycophenolate mofetil (MMF) is reported. A review of the literature on the topic of IgG4 related systemic disease and AIP in childhood is included.

Case Presentation
A 13-year-old Caucasian female with chronic abdominal pain initially developed fever, headache, and joint symptoms. Her symptoms persisted and progressed until she was hospitalized for shortness of breath several months later after developing what appeared to be a right lung scar and pericardial effusion by imaging. She was also diagnosed with pancreatitis and jaundice with a serum bilirubin level of 4.5 mg/dL. A computerized tomography (CT) scan revealed multiple pancreatic masses and biliary dilatation, lab work revealed an elevated IgG4, and she was transferred to our institution for further work up and management. Because of her persistent and recurrent pancreatitis symptoms, along with abnormal imaging she was diagnosed with AIP, meeting both JPS criteria (criterion 1 and criterion 2) and HISORt criteria (group B and after initiation of steroids, group C). Radiographic imaging [endoscopic ultrasound (EUS) and endoscopic retrograde cholangiopancreatography (ERCP)] and surgical placement of a pancreatic stent were performed. Prior to stent placement her serum lipase was elevated at 313 U/L (10-180). Other significant laboratory findings included the following results: amylase 39 U/L (39-100), AST 106 U/L (10-30), ALT 88 U/L (10-30), and alkaline phosphatase 417 U/L (93-386). EUS/ERCP revealed a stricture at the distal common bile duct, multiple focal hyperechoic areas in the pancreas with benign fine needle aspirate (FNA) consistent with chronic pancreatitis (specifically revealing a lymphocytic inflammatory infiltrate; no staining was performed on pancreatic aspiration specimens because there was insufficient sample), and a pancreatogram consistent with chronic pancreatitis (stricture of distal common bile duct and biliary dilatation). Following stent placement her lipase improved to 84 U/L. The stent was removed two months later as previously scheduled, and shortly after the procedure she developed worsening abdominal pain. A CT of the abdomen demonstrated dilatation of the intrahepatic and extrahepatic bile ducts without evidence of choledocholithiasis, mild dilation of the pancreatic duct and multiple foci of hypodensity within the kidneys. Her bilirubin level and liver function tests began to increase. One week after stent removal she was taken back for an ERCP and another stent was replaced. Her pain and jaundice improved, and the liver enzymes quickly returned to normal. When her stent was removed for the second time, multiple brushings and biopsies were taken. Pathology on duodenal biopsies revealed marked expansion of atypical lymphoid tissue, positive staining for CD20, BCL-6, BCL-2, CD10, and KI-67 and negative for CD5, CD23, Tdt, BCL-1, and CD30. cMYC translocation was negative and there was a light chain but not heavy chain rearrangement present. IgG4 staining was not done on this sample. This immunochemistry suggested a diffuse large B cell lymphoma; however, follow up on following FNA, biopsy and imaging did not demonstrate a monoclonal lymphoid population nor suggest malignancy. Autoantibody studies were all negative including: ANA, anti-dsDNA, anti-smooth muscle antibodies, anti-RNP, SSA/SSB autoantibodies, scleroderma-70 antibodies, and ANCA. In addition, C3, C4, and angiotensin converting enzyme levels were within normal limits for age. She returned again in 3 months for stent removal which she tolerated well. EUS revealed calcifications in the head of the pancreas (a finding that is unusual for Table 1 Clinical diagnostic criteria of Autoimmune Pancreatitis as proposed by the Japanese Pancreas Society,  revised 2006 (4) 1. Diffuse or segmental narrowing of the main pancreatic duct with irregular wall and diffuse or localized enlargement of the pancreas by imaging studies, such as abdominal ultrasonography, computed tomography, and magnetic resonance imaging.
2. High serum gamma-globulin, IgG, or IgG4, or the presence of autoantibodies, such as antinuclear antibodies and rheumatoid factor.
3. Marked interlobular fibrosis and prominent infiltration of lymphocytes and plasma cells in the periductal area, occasionally with lymphoid follicles in the pancreas. autoimmune pancreatitis) and a large (27 mm × 16 mm) area of stranding within the mediastinum, consistent with fibrosing mediastinitis. As treatment for her pancreatitis, corticosteroids had been prescribed previously resulting in improvement of pain but with significant weight gain, and her abdominal pain/symptoms would return shortly after tapering the corticosteroid dose. At that time the rheumatology service was consulted for the possibility of AIP.
To spare corticosteroids and to treat the IgG4-related pancreatitis and mediastinitis, she was started on MMF and tapered up to a maximum dose of 1,000 mg twice daily. She improved quickly and remarkably with resolution of abdominal pain and the ability to return to an enteral diet within days of starting MMF. The addition of MMF allowed for significant improvement of symptoms, eventually in the absence of corticosteroids, as well as normalization of laboratory and imaging findings consistent with pancreatitis. In terms of her laboratory findings, her lipase improved from 444 to 53 U/L (normal, 10-180) and amylase lowered from 63 to 41 U/L within 3 days of MMF. The initial IgG4 level was 226 mg/dL (normal range, 11-157) prior to MMF; after 15 months of initiating treatment with MMF, the IgG4 subclass level was 73.4 mg/dL. Inflammatory markers, ESR and CRP, were not measured until after treatment was started, at which point they were within normal limits. A repeat chest and abdominal CT one year after initiating treatment did not reveal any residual mediastinal or hilar lymphadenopathy. There was no evidence of pulmonary nodules or focal pancreatic lesions, and there was resolution of the hypodense renal lesions and fatty infiltration of the liver. She was ultimately tapered off of all immunosuppressive therapy within 1 year of treatment with MMF. The child currently remains in good health off of all immunosuppressive medicine after an additional period of over 7 months. She, thus, represents the first pediatric patient with AIP and elevated IgG4 levels to be successfully treated with MMF allowing for discontinuation of corticosteroids, and eventually all immunosuppressive therapy.

Review of the AIP literature and focus on pediatric cases
Pancreatitis due to an inflammatory, autoimmune process was first suggested by Sarles et al. in 1961 with the description of 10 adult cases of chronic pancreatits, all with massive cellular infiltrates and increased inflammation [1]. AIP was later named and described by Yoshida et al. [2] in 1995 when a patient without evidence of Sjogren syndrome developed pancreatitis and responded to corticosteroid therapy. The patient had evidence of autoimmune abnormality with a positive ANA titer, anti-thyroglobulin antibody, and anti-microsomal antibodies; all other autoantibodies tested were negative [2].
The authors described 11 characteristics of AIP that have been used as the basis for the development of recent diagnostic criteria including; the JPS diagnostic criteria of AIP [4] and the Mayo Clinic HISORt criteria [3,5] (Table 1).

Diagnostic criteria for AIP
Diagnostic criteria for AIP were first developed by the JPS in 2002 and were revised in 2006 after reevaluating additional patients with AIP and with likely AIP. The foundation of the diagnosis rests on meeting characteristic findings on imaging, along with supportive serology and/or histology ( Table 1). The JPS criteria emphasize the importance of excluding malignancy during the work up and the responsiveness to steroids, although they did not include it as a diagnostic criterion [4]. The Mayo Clinic, Chari et al. [3,5], used a cohort of patients with AIP diagnosed by histology to develop criteria in five categories that make 3 diagnostic groups ( Table 2). Those categories and criteria are: Histology, [at least one of the following: periductal lymphoplasmacytic infiltrate with obliterative phlebitis and storiform fibrosis (LPSP) or lymphoplasmacytic infiltrate with storiform fibrosis showing abundant (≥10 cells/hpf) IgG4-positive cells]; Imaging, typically diffusely enlarged gland with delayed (rim) enhancement, diffusely irregular, attenuated main pancreatic duct or alternatively, focal pancreatic mass/enlargement, focal pancreatic duct stricture, pancreatic atrophy, pancreatic calcification, or pan creatitis; Serology, elevated serum IgG4 level (>140 mg/dL); Other organ involvement, hilar/intrahepatic biliary strictures, persistent distal biliary stricture, parotid/lacrimal gland involvement, mediastinal lymphadenopathy, retroperitoneal fibrosis; and response to steroid therapy, resolution/marked improvement of pancreatic/extrapancreatic manifestations with steroid therapy [3,5]. Of the previously diagnosed idiopathic fibrosing pancreatitis (IFP) pediatric patients included in this review, by imaging and histology, they are more consistent with AIP, although immunoglobulin levels were not measured and steroids were not given (surgically treated) in any of the cases [15][16][17][18] (Table 3). Thus, for many pediatric patients reported in the literature to have IFP, it is difficult to establish whether they meet AIP criteria.

Histopathology of AIP
Similar to the patient reported herein, needle biopsy of the pancreas in the case report by Yoshida et al. [2] revealed severe fibrotic change with lymphocytic infiltration. LPSP was the descriptive name of AIP prior to the development of diagnostic criteria, a histologic pattern distinct from ordinary chronic pancreatitis, alcoholic pancreatitis, and obstructive pancreatitis secondary to cancer. The histologic findings of AIP used to include or exclude a diagnosis of AIP in case series and cohort studies are a periductal lymphoplasmacytic infiltrate with obliterative phlebitis and swirling fibrosis centered around ducts and veins [3][4][5][19][20][21][22][23][24][25]. The majority of lymphocytes are T cells, and eosinophils and neutrophils can be found. In one study, all 3 patients with ulcerative colitis were found to have neutrophilic infiltration of pancreatic ducts [21]. Two subtypes of histologic inflammatory pattern were described, ductocentric and lobulocentric [21]. A similar dichotomy was described by another group, interlobular lymphoplasmacytic inflammation and idiopathic duct-centric chronic pancreatitis (intralobular) [26]. Thus, the histopathologic findings in AIP are rather varied.

Imaging findings in AIP
Both the JPS [4] and the Mayo HISORt [3,5] criteria have specific imaging descriptions for AIP. A diffusely enlarged pancreas on CT with diffuse irregular narrowing of the pancreatic duct (both of which resolved following therapy) were included in the Yoshida et al. original description of AIP [2]. In fact, diffuse irregular narrowing of the pancreatic duct was part of the subjective differentiation between sclerosing pancreatitis with elevated IgG4 levels and ordinary chronic pancreatitis (irregular dilatation of the pancreatic duct or diffuse calcification) for Hamano et al. [6]. Typically, AIP will have imaging that reveals a "sausage-like" pancreas with delayed enhancement on CT scan, rim enhancement on T2 weighted MRI, or hypoechoic pancreas with echogenic spots on ultraound. AIP rarely causes calcification or pancreatic pseudocyst, but can cause localized enlargement which can contribute to the difficulty differentiating it from malignancy [7]. The JPS criteria for diagnosis of AIP require that pancreatic imaging is consistent with AIP [4]. The Mayo Clinic HISORt criteria, however, have 3 distinct diagnostic groups, only one of which includes imaging criteria [3,5]. Both criteria, though, have the same imaging requirement, an enlarged pancreas with diffuse irregular narrowing of the pancreatic duct [3][4][5].

Elevated IgG4 levels in AIP
In addition to the imaging and histologic characteristics of AIP, a sensitive and specific serologic association with elevated IgG4 serum levels was first described by Hamano et al. [6]. A serum IgG4 level of at least 135 mg/dL resulted in 95% sensitivity and 97% specificity in differentiating between AIP (referred to as sclerosing pancreatitis in Hamano et al., and diagnosed by imaging, histology, and response to steroids) and pancreatic cancer. They did not find elevated serum IgG4 levels in patients with chronic pancreatitis, primary biliary cirrhosis, primary sclerosing cholangitis, nor Sjogren syndrome [6]. These findings were supported by investigators at the Mayo Clinic; using a serum IgG4 level cutoff of 140 mg/dL, the sensitivity for AIP was 76% and specificity 93%. Moreover, the negative predictive value of serum IgG4 <140 mg/dL was 99%, but the positive predictive value was only 36% [20]. In both the Mayo Clinic cohort [20] and the Hamano et al. cohort [6], 3-10% of patients without AIP had elevations in serum IgG4, and in the Mayo Clinic cohort 11/45 (24%) patients with AIP had normal serum IgG4 levels. All patients diagnosed with AIP regardless of serum IgG4 levels had similar clinical responses to corticosteroids. The patients most likely to have normal IgG4 levels with a histologic diagnosis of AIP were more likely to be female, younger, and less likely to have other organ involvement. In addition, all patients in the Mayo Clinic cohort that had normal serum IgG4 levels and available histology had positive IgG4 immunostaining [20]. IgG4 immunostaining has become a method to both confirm pancreatic histology consistent with AIP and evaluate other manifestations of a similar disease process. In the study by Ghazale et al. [20], 19/20 patients who had IgG4 immunostaining of pancreatic tissue showed >10 IgG4 positive cells/hpf. Five patients with elevated serum IgG4 levels that did not have AIP by histology, had IgG4 immunostaining performed that was negative in four patients and showed nonspecific focal clusters of IgG4 positive cells in one patient [20]. In another cohort, all 21 AIP patients with pancreatic tissue IgG4 immunostaining had at least >5 IgG4 positive cells/hpf, and extrapancreatic specimens with lymphoplasmacytic infiltrate (3 biliary, 2 pulmonary, 3 salivary glands, 4 renal) when tested also showed significant IgG4 positive cells by immunostaining (3 biliary, 1 lung, 1 salivary gland, 4 renal) [21]. Since the pathology of the extrapancreatic lesions is the same as the pancreatic pathology, the concept of IgG4 related systemic disease was developed. Whether or not distinct categories of AIP based on IgG4 levels will be established in pediatric cases is yet to be determined.

Extrapancreatic manifestations of IgG4 related disease
Elevated serum IgG4 levels and positive IgG4 immunostaining have been associated not only with AIP, but also with retroperitoneal fibrosis, sialoadenitis, mediastinal adenopathy (as in the patient reported herein), cholangitis, and tubulointerstitial nephritis [3,21,22,27] and is one criterion for diagnosis in the HISORt criteria [3,5]. In 12 patients clinically diagnosed with idiopathic retroperitoneal fibrosis, 16 biopsies (retroperitoneal tissue, liver, kidney, colon, omentum) showed fibrosis with a lymphoplasmacytic infiltrate, and 14/16 biopsies had IgG4 positive plasma cells (not seen in one liver and one colon biopsy). Comparative biopsies from inflammatory and fibrotic diseases in similar organs failed to show any IgG4 positive cells [27]. Ghazale et al. [22] reported 4 patients with IgG4 positive immunostaining on bile duct resection with histology comparative to AIP, namely, lymphoplasmacytic infiltrate with storiform fibrosis and obliterative phlebitis around the bile duct, who had no evidence of pancreatic disease or involvement. Similarly, these patients responded to corticosteroids with improvement in all and resolution of strictures in most. Some patients, however, had relapse with corticosteroid taper, but instead of continuing low dose therapy they were started on other immunomodulatory medications, azathioprine, MMF, or cyclophosphamide with successful maintenance of treatment [22]. The similar pathology, immunostaining, and response to steroids have led to the proposal of an IgG4-related sclerosing disease, of which AIP is the pancreatic manifestation of a systemic process [7,21,23,27].

Treatment of AIP
Prior to the concept of AIP, patients with painless swelling of the pancreas were treated with surgery to rule out a malignant process. The histologic specimens collected following resection, where malignancy was ruled out and an inflammatory infiltrate was noted instead, have been used to create the Mayo Clinic HISORt criteria [3]. Yoshida et al. [2] was the first to report the successful use of corticosteroids in a patient with AIP, resulting in resolution of symptoms, imaging abnormalities, and laboratory abnormalities. A resolution or at least effective response to corticosteroids is part of the Mayo Clinic HISORt diagnostic criteria and is noted as a characteristic of AIP in the JPS criteria [3][4][5].
As the concept of AIP as part of a systemic IgG4 related disease developed, so did the need for more aggressive treatment. In most cases of AIP, patients respond fully to a single course of corticosteroids [2,3,25]; however, some patients with AIP, and especially those with extrapancreatic manifestations (including our patient), are unable to fully wean off of corticosteroids [22,25,27]. Azathioprine and low dose prednisone have been used to control symptoms and limit disease course with success [27]; in addition, MMF and cyclophosphamide have also been used in cases of IgG4 associated cholangitis (IAC) with successful maintenance of remission [22]. MMF has also been used successfully in patients with idiopathic retroperitoneal fibrosis (there was no mention of IgG4 testing or staining in this case series) [28]. These patients were initially started on prednisone and MMF, the prednisone was weaned off over 6 months, and MMF was discontinued after 11-96 months (mean 27 months). All 9 patients had regression of disease and improvement in inflammatory markers; there was no recurrence after discontinuation of medical therapy. In vitro lab work has demonstrated that MMF has direct antifibrotic properties by inhibiting type I collagen production, altering the contractile and migratory functions of fibroblasts, and enhancing the production of matrix metalloproteinase (interstitial collagenase) [29]. Thus, MMF seems a logical choice in treating autoimmune fibrosing disorders.

Pediatric Reports
There are 7 other pediatric patients reported in the literature with AIP or likely AIP (Table 3) by JPS or HISORt criteria. Only one had normal imaging [14] (she was unable to be diagnosed as AIP by JPS or HISORt criteria, but did respond to steroid therapy and had an elevated IgG4), and 4 of the 7 patients had histology reported, all 4 consistent with AIP [8][9][10]13]. The only patient with an elevated IgG4 was the patient who did not meet diagnostic criteria for AIP [14]; she also had an elevated total IgG level. One other patient had an elevated IgG level, but normal IgG4 level [12], and one patient did not have any laboratory work reported [13]. The other pediatric patients reported in the literature with AIP or likely AIP (Table 3) were treated with corticosteroids or pancreatoduodenectomy with mixed success. One patient requires daily steroids [12], treatment is not reported in one [10], the surgical patient requires digestive enzymes [13], and one patient requires NPH insulin for diabetes but did not have recurrence of disease [9]. There are many reports of IFP in the literature, but Mannion and Cron Pediatric Rheumatology 2011, 9:1 http://www.ped-rheum.com/content/9/1/1 only 4 have either enough criteria information reported or criteria consistent with AIP (Table 3) [15][16][17][18]. These patients were treated with surgical resection [15,17], observation (following diagnostic exploratory laparotomy) [16], or biliary stenting [18], all with resolution of symptoms. None of these patients had immunoglobulins reported or were given corticosteroids. The extent of AIP and IgG4-associated autoimmunity in children remains unanswered at present.

Conclusions
The patient presented in this report is diagnosable with AIP by both JPS and HISORt (group B or C) diagnostic criteria [3][4][5]. She also had an elevated IgG4 with evidence of extrapancreatic manifestations, consistent with IgG4 related systemic disease (ISD), even though IgG4 staining was not done (Table 1 and 2). Her symptoms improved with corticosteroids, but she was unable to wean off of them without disease flare and did not want to continue due to systemic side effects. She was successfully treated with MMF for one year with resolution of symptoms and evidence of inactive disease by imaging and laboratory studies for a total of 19 months, until she was lost to follow up. Our patient represents the first pediatric patient with elevated IgG4 who met both JPS and HISORt criteria for AIP, and the first pediatric patient with AIP/ISD who was successfully treated with MMF.
The pediatric patients reported in the literature, including the patient reported here, have a disease that is similar to adult AIP/ISD. The imaging reports are similar, histology (when available) is consistent with AIP, and there is typically a good response to corticosteroids. The pediatric patients, however, more often present with abdominal pain than the adult patients (11/12 reported here), and when immunoglobulins are measured, they are more likely to be normal than elevated. Our patient has criteria most similar to the adult patients and criteria reported in the literature, including an elevated IgG4 level which is only reported in one other pediatric patient, a patient that did not meet diagnostic criteria for AIP [14]. This current report demonstrates that while pediatric AIP/ISD is very similar to adult AIP/ISD, it is not identical. As reports of pediatric AIP/ISD increase, the criteria used to diagnose adults may need to be modified to better suit pediatric patients. Since pancreatic malignancy is significantly rarer in pediatric patients than adult patients, the criteria could likely be more inclusive without the risk of delaying treatment for cancer. If AIP/ISD were a more likely diagnosis in pediatric patients with pancreatitis, corticosteroid therapy could be attempted sooner and limit surgical interventions, symptoms, and hopefully disease progression. In the patients who do meet the adult criteria for AIP/ISD who are unable to wean off of corticosteroids or who cannot tolerate the side effects, other immunomodulatory medications should be considered. Our patient responded fully to 1 year of treatment with MMF, was symptom free at 19 months follow up with normal imaging and IgG4 levels.

Consent
Institutional Review Board approval was attained. The patient and her family have been lost to contact. Repeated attempts at contact, including via her gastroenterologist, have failed.