Gastroduodenal inflammation is a common finding when performing esophagogastroduodenoscopy (EGD). In many patients, an obvious cause, such as Helicobacter pylori (HP) or non-steroidal anti-inflammatory drug (NSAID)-associated peptic ulcer disease, can be identified (1). First-line treatment consists of proton pump inhibitors (PPIs) or HP eradication (2). However, some patients do not respond, and a broad range of more rare disease entities need to be explored. We report a case of treatment-refractory ulcerative gastritis and duodenitis complicated by pyloric stenosis, in which we evaluated a broad differential diagnosis based on the clinical course and pathological findings.

A 20-year-old man with a previously unremarkable medical history presented to the outpatient clinic with epigastric pain and 10 kg weight loss over the past 6 months. He also reported early satiety during meals, nausea without vomiting, and looser stools for a couple of months. The patient was an active smoker and had regular alcohol consumption, but denied illegal substance use. He worked as a logistic assistant in the shipping industry. He did not recently use NSAIDs or any maintenance medical treatment. Familial history included colorectal cancer (paternal grandfather, at the age of 60) and a perianal fistula (father). Vital parameters were normal, and physical examination showed no abdominal abnormalities. No pathological lymph nodes were palpable.

Initial laboratory testing showed no anemia. White blood cell count was within normal range and there was mild thrombocytosis (394,000 10E3/μL). The serum electrolytes were within normal limits, except for mild hypomagnesemia (0.61 mmol/L, 0.70–1.05 mmol/L). There was mild elevation of aspartate aminotransferase (AST) (84 U/L; reference <37 U/L) and alanine aminotransferase (ALT) (92 U/L; reference <40 U/L) with no alterations in other liver function tests. IgE titer was remarkably elevated (758 kU/L; reference 0–100 kU/L). Fecal calprotectin was slightly elevated (130.9 mg/kg, reference <50 mg/kg).

A few months earlier, the patient had already undergone an EGD for his complaints at a different center. This revealed a bumpy and erosive appearance of the gastric mucosa and bulboduodenal ulcerations with stenosing effect, causing gastric outlet subobstruction. The biopsies showed chronic active, HP-negative gastritis and bulbitis, in the absence of PPI intake. There were no signs of malignancy, and no granulomas were observed. Pantoprazole 40 mg BID was started.

Control EGD after presentation at our center showed similar macroscopic findings despite PPI treatment: diffuse gastritis with gastric outlet stenosis due to large ulcers at the transition between pylorus and bulbus, which could only be passed with a nasogastric endoscope (diameter 5.4 millimeter) (Figure 1). Repeated extensive biopsy sampling confirmed persistent acute bulbitis and HP-negative, chronic active gastritis. Periodic acid-Schiff staining gave no arguments for Whipple’s disease.

An abdominal computed tomography (CT), followed by additional investigations including ileocolonoscopy with ileal and colonic biopsies and a magnetic resonance (MR) enterography, did not show other locations of intestinal inflammation. An additional video capsule endoscopy to screen for more distal small bowel inflammation could not be performed because the capsule could not be advanced beyond the pyloric stenosis despite endoscopic maneuvers.

Screening for Zollinger–Ellison (ZE) syndrome showed a mildly elevated serum gastrin level (339 ng/L) on PPI and 68Ga-DOTA-1-NaI3-octreotide (DOTANOC) positron emission tomography (PET)-CT showed no elevated somatostatin receptor expression. Endoscopic ultrasound (EUS) of the pancreas and MR enterography showed no arguments for a primary neuro-endocrine tumor. There were no histopathological arguments for autoimmune gastritis, and anti-intrinsic factor antibodies and anti-parietal cell antibodies were negative. Intestinal tuberculosis was excluded by a negative interferon-gamma release assay (i.e., QuantiFERON-TB) and normal X-ray of the thorax. Serum calcium and serum angiotensin-converting enzyme (ACE) levels were within normal limits, which made sarcoidosis less likely.

After this profound work-up, the patient was referred to the IBD unit of our hospital for further investigation. In order to establish a diagnosis, a repeated EGD with biopsies showed erosions in the stomach with the known stenosis of the pylorus, a large ulcer at the transition from the pylorus to the bulbus and multiple punched-out ulcerations in the duodenum. The gastric biopsies now showed a few small non-caseating granulomas, suggestive of Crohn’s disease (CD), since other causes of granulomatous gastritis (GG), i.e., sarcoidosis, malignancy, and infectious diseases such as tuberculosis or Whipple’s disease, had already been thoroughly ruled out (Figure 1). In addition, because of persistent ulcerative gastritis with stenosis, despite high doses of PPI, immunoglobulin G4 (IgG4) staining was performed to rule out IgG4-related disease, and the number of IgG4-positive plasmocytes was found to be significantly elevated on both gastric and duodenal biopsies. In the corpus of the stomach, the number of IgG4-positive plasma cells was highest, with more than 50 IgG4-positive plasma cells per high power field (HPF) and an IgG4/IgG ratio above 40%. An elevated IgG4 was seen in serum (194 mg/dL, reference values 8–140 mg/dL). Based on these clinical, serological, radiological, and histopathological findings, the presumptive diagnosis of an overlap of gastroduodenal CD and IgG4 disease was performed, as granulomas are unlikely in IgG4-related disease. Initially, systemic steroids were refused by the patient. Budesonide (9 milligram/day) was started without any effect on the complaints. Consequentially, a methylprednisolone 32 mg/day tapering course was started, and because of the suspected upper gastrointestinal (UGI) CD, anti-tumor-necrosis factor alpha (anti-TNFα) treatment with adalimumab was initiated early in the disease course. This approach led to rapid symptomatic improvement.

A control EGD (performed when the patient was still under CS) confirmed a clear endoscopic response with resolution of large ulcerations but still persistent diffuse gastroduodenitis with erosions. Because of the positive IgG4 staining, after a review of the literature, it was decided to switch adalimumab to infliximab in combination with azathioprine. After induction with Infliximab, EGD now showed an increased response, with only some residual erythema in the antrum and now a normal pyloric opening that could be passed with a gastric endoscope (diameter 9.9 millimeter). No abnormalities were visualized in the duodenal bulbus. However, active chronic inflammation with increased IgG4-positive plasma cells (>100 IgG4-positive plasma cells/HPF) remained present in the duodenal biopsies. Gastric biopsies showed inactive chronic gastritis with up to 14 IgG4-positive plasma cells/HPF. Control EGD 6 months after induction showed no macroscopic abnormalities, with only mild chronic and inactive antritis in the biopsies without granulomas (Picture 1). IgG4-plasmocyte count was also normalized (only 7/HPF in the duodenum and 2/HPF in the gastric biopsies).

In this case report, we suggest an overlap between gastric CD and IgG4-related disease (IgG4-RD). A review of literature was performed below to summarize the diagnosis and treatment of both entities.

A thorough work-up remains necessary in patients presenting with refractory, ulcerative gastroduodenal inflammation, and differentiating between GG, upper GI IgG4-RD, and upper GI CD remains challenging. We report a case in which IgG4-RD gastroduodenitis and upper GI CD were overlapping. Granulomas are typical but not necessary in diagnosing CD after eliminating other causes such as sarcoidosis and tuberculosis. As was the case in our patient, GG-associated CD is more prevalent in male and young patients. On the other hand, granulomas are uncommon in IgG4-RD. However, due to persisting ulcerative gastroduodenitis with stenosis despite high doses of pantoprazole, an IgG4 stain and serum IgG4 were requested, which was positive. As shown in the literature, IgG4-RD should be ruled out not only in ulcerative disease but also in patients with a submucosal tumor or suspicion of GIST to avoid unnecessary surgery. First-line treatment for both CD and IgG4-RD consists of corticosteroids with tapering. Our patient received anti-TNF and azathioprine as a maintenance treatment based on the existing literature for CD, which is in favor of anti-TNF and/or azathioprine, and IgG4-RD, which favors an immunomodulator such as azathioprine. This approach led to an excellent response clinically, endoscopically, and histologically.

Physicians should be aware of the potential existence of IgG4-RD in the upper GI tract in refractory disease. Nonetheless, literature on diagnosis, prognosis, and treatment of upper GI IgG4-RD is scarce and the meaning of increased IgG4 in serum or gastric/duodenal biopsies remains the subject of further research.

Especially in IBD patients, mucosal IgG4-infiltration may be a marker and risk factor for aggressive disease rather than a separate disease entity, which is to date mostly suggested in UC. More research in this field is necessary to unravel this question and further explore the actual role of IgG4-elevation of IBD patients.

Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

VD: Writing – original draft. JG: Writing – review & editing. AH: Writing – review & editing. TL: Writing – review & editing.