Iron deficiency is the most common nutritional deficiency in children and adolescents [1,2]. Adolescence is a time when iron needs increase and the risk of developing iron deficiency anemia (IDA) is particularly high. Iron deficiency affects physical endurance and cognitive performance in adolescents . Therefore, accurately diagnosing and treating IDA in adolescents are important.
IDA in adolescents is commonly caused by heavy menstrual blood loss in females, dietary problems such as a vegetarian diet, and malnutrition . However, because IDA can develop due to other uncommon causes, further evaluation is needed if the cause of IDA is unexplained, IDA is very severe, or the patient does not respond to standard treatment .
Collagenous gastritis (CG) is a rare gastrointestinal disorder that was first reported in 1989; only approximately 60 cases of CG in children have been reported in English literature . CG is histologically characterized by the presence of a subepithelial collagen band associated with inflammatory cell infiltration into the lamina propria . CG can be divided into two clinical phenotypes: pediatric and adult. In pediatric-type CG, lesions are mainly confined to the stomach, and the most common clinical symptom is abdominal pain accompanied by IDA [6,8]. Adult-type CG is related to collagenous colitis, and characteristically, profuse diarrhea demonstrates completely different clinical features from pediatric-type CG.
Here, we present three cases of pediatric CG diagnosed during treatment for recurrnet IDA. Through these case studies, we emphasize the need for esophagogastroduodenoscopy (EGD) and appropriate biopsy, considering the possibility of pediatric CG in pediatric patients with recurrent IDA.
This study was approved by Seoul St. Mary’s Hospital Institutional Review Board (no. KC23ZASI0569).
A 10-year-old boy was diagnosed with IDA 4 years ago and was being treated in the hematology department. When he first visited the hospital, the laboratory examination revealed hemoglobin (Hb), mean corpuscular volume (MCV), reticulocyte count, ferritin, and iron levels of 3.9 g/dL, 55.7 fL, 1.41%, 0.9 ng/mL, and 9 mg/dL, respectively. He was referred to the gastrointestinal (GI) clinic for evaluation of recurrent IDA. He had intermittent epigastric pain; however, the symptoms were not severe, and no other gastrointestinal symptoms were observed. On physical examination, there was mild epigastric tenderness, but there were no other unusual findings. He was being treated with oral iron supplements (3 mg/kg elemental iron per day), and laboratory tests revealed levels of Hb, MCV, reticulocyte count, ferritin, and iron of 11.5 g/dL, 76.4 fL, 0.88%, 8.1 ng/mL, and 28 g/dL, respectively, as well as total iron-binding capacity (TIBC) of 429 g/dL. Stool occult blood test was negative. EGD revealed severe mucosal nodularity, mucosal edema, erythema, and bleeding in the gastric body (Fig. 1A). Histological examination revealed chronic active gastritis with no
At that time, the possibility of CG was considered based on the endoscopic findings; however, a biopsy was not performed on the depressed mucosa around the nodule, and Masson’s trichrome staining test was not separately implemented. No CG-related findings were identified in the biopsy results, and the eosinophilic count was higher than the normal range. Therefore, the possibility of eosinophilic gastritis was considered, and a treatment plan was established. The patient was followed up with a proton pump inhibitor (PPI), hypoallergenic diet, and oral iron replacement (3 mg/kg elemental iron per day for 12 months). The Hb level did not decrease and remained the same, although the ferritin level repeatedly improved and worsened from a low of 7 ng/mL to a high of 46 ng/mL. The patient’s intermittent epigastric pain improved.
Approximately one year later, the patient underwent follow-up EGD, and nodular mucosa and mucosal erythema remained in the gastric body, although these demonstrated slight improvement compared to the previous examination, and no easy touch bleeding was observed. A biopsy was performed on several lesions, including the depressed mucosa around the nodule. Histological examination revealed irregular but linear collagen deposition (by Masson’s trichrome) and partial surface epithe-lial detachment, consistent with CG (Fig. 2). At that time, eosinophil infiltration was confirmed as 41/HPF.
After the diagnosis of CG, regular follow-up was performed with intermittent intravenous iron replacement (1,000 mg of elemental iron) and hypoallergenic diet (shellfish and nut restriction). Both Hb and ferritin levels were within the normal range, and the patient had no gastrointestinal symptoms including epigastric pain. EGD performed 6 months later revealed similar findings.
A 13-year-old boy who was being followed up in the hematology department with recurrent IDA was referred for the evaluation of intermittent abdominal pain and nausea. When he first visited the hospital, the laboratory examination revealed Hb, MCV, reticulocyte count, ferritin, and iron levels of 10.5 g/dL, 73.4 fL, 1.0%, 5.5 ng/mL, and 14 mg/dL, respectively. Except for abdominal pain and nausea, there were no other gastrointestinal symptoms or changes in weight. There were no unusual findings in physical examination other than pale appearance. IDA was diagnosed 2 years prior and treated with intermittent intravenous iron supplements; however, the laboratory results repeatedly improved and worsened from 10 g/dL to 12.6 g/dL for Hb and from 3.7 ng/mL to 22.8 ng/mL for ferritin. Stool occult blood tests were performed twice, but both were negative. He had a body mass index of less than the 3rd percentile and no abnormal findings on physical examination. The laboratory examination findings revealed the following: Hb, MCV, reticulocyte count, ferritin, and iron levels of 12.6 g/dL, 81.8 fL, 1.08%, 7.9 ng/mL, and 31 mg/dL, respectively and TIBC of 384 mg/dL. EGD revealed nodular mucosa and mucosal erythema in the gastric body and easy touch bleeding (Fig. 1B). At the first examination, a biopsy was not performed on the depressed mucosa around the nodule, and Masson’s trichrome staining was not requested. Histological examination revealed chronic gastritis (severe) with mild
Based on the test results,
Follow-up EGD was performed 5 months later, and heterogeneous nodular mucosa remained in the gastric body. However, easy touch bleeding improved. Multiple biopsies were performed, including those of the depressed mucosa. Histological examination revealed thickened subepithelial collagen, and the patient was diagnosed with CG. No eosinophilic infiltration was observed.
In this case, the association between milk protein, eggs, and gastrointestinal symptoms was clear; therefore, the intake of dairy products and eggs was restricted, and intermittent oral and intravenous iron replacement was performed. He was regularly followed up, and both Hb and ferritin levels were within normal ranges.
An 11-year-old female patient was diagnosed with IDA in the hematology department 10 months previously. When she first visited the hospital, the laboratory examination revealed Hb, MCV, reticulocyte count, ferritin, and iron levels of 4.2 g/dL, 73.7 fL, 2.1%, 1.3 ng/mL, and 11 mg/dL, respectively. She was referred to the GI department because of recurrent IDA during follow-up and intermittent epigastric pain and vomiting. Physical examination showed tenderness in the epigastric area. Labora-tory examination revealed Hb, MCV, reticulocyte count, ferritin, and iron levels of 10.5 g/dL, 84.3 fL, 1.46%, 7.1 ng/mL, and 25 mg/dL, respectively, and TIBC of 389 mg/dL. Stool occult blood test was negative. EGD revealed nodular mucosa in the antrum and gastric body, and the gastric body nodules were heterogeneous in shape (Fig. 1C). The patient was suspected to have CG, and multiple stomach biopsies were performed. On histological examination, Masson’s trichrome staining was negative, and the diagnosis was CG with erosion, regenerative foveolar epithelium, and without
The patient was diagnosed with eosinophilic gastritis and received PPI treatment and prescribed a hypoallergenic diet; however, no clear association between food intake and epigastric pain was identified. Subsequently, as intermittent abdominal pain and IDA persisted even after oral iron replacement (3 mg/kg elemental iron per day for 3 months), a follow-up endoscopy was performed 3 months later. On follow-up EGD, the nodular mucosa was more severe than before (Fig. 1D), CG with subepithelial deposition of collagen bands was confirmed on histological examination. At that time, eosinophil infiltration was confirmed as 28/HPF.
The patient subsequently discontinued the hypoallergenic diet and was regularly followed up with intermit-tent PPI and intravenous iron supplementation. Afterwards, the patient had no epigastric pain or vomiting, and the IDA findings require further follow-up.
Pediatric CG is rare. To date, most of the literature comprises case reports, small case series, and histopatho-logical reviews. Because patients with pediatric CG are few, studies related to the treatment and prognosis of the disease are lacking . Therefore, no standard guidelines for treating pediatric CG have been established, and treating patients involves many difficulties .
Recently, the number of studies on pediatric CG has increased. Cases of three pediatric patients with CG in another hospital in South Korea have been reported a few years ago, and a national cohort study of pediatric patients with CG has been reported in France a few months ago [10,11]. In our hospital, two more patients were diagnosed within one year after the first pediatric patient with CG was diagnosed. Pediatric and adult types of CG demonstrate completely different features. Adult type CG has more diffuse involvement in the gastrointestinal tract, and chronic diarrhea is the main symp-tom. It is also often associated with autoimmune disease. On the other hand, pediatric type CG is mainly involved in the stomach, and abdominal pain and anemia are the main symptoms. Although the pathophysiology of IDA in pediatric type CG is unclear, the response to iron supplements suggests that microscopic blood loss rather than iron malabsorption may be the cause. Because pediatric type CG shows completely different aspects, research separate from adult type CG is needed in this regard. Diagnosis is difficult unless CG is suspected and appropriate tests are not performed . Therefore, CG may be underdiagnosed. Alternatively, the incidence of pediatric-type CG may increase for unknown reasons. Unlike the previous perception that pediatric CG is a very rare disease, it may be a disease that can be often detected in clinical practice.
The most representative feature of pediatric CG is gastric mucosal nodularity on EGD [7,12,13]. Findings of nodular gastritis in children are most often associated with
When diagnosing pediatric CG, it should be noted that the depressed mucosa around the nodule creates a typical nodular appearance. Therefore, a biopsy of the depressed mucosa needs to be performed for diagnosis. Several mucosal biopsies may be needed to accurately diagnose and differentiate other diseases such as
Eosinophilic gastroenteritis is a rare disease characterized by eosinophilic infiltration of the gastrointestinal tract . Although eosinophilic gastroenteritis can be diagnosed only when secondary causes of eosinophilic infiltration are excluded, it is generally suspected if >30 eosinophils per HPF are observed in the stomach on biopsy. Previous case reports have also mentioned eosinophilic infiltration in some patients, and some patients have attempted adhering to hypoallergenic diets as a treatment. However, in our hospital, all three patients had >30 eosinophils/HPF in the stomach on biopsy, which is different from that of previous reports [6,11]. Although common secondary causes of eosinophilic infiltration such as parasitic infection, drug-induced eosinophilia, inflammatory bowel disease, and leukemia have been ruled out, diagnosing eosinophilic gastritis is difficult because CG-induced inflammation itself may cause eosinophilic infiltration. However, in two patients, hypoallergenic diets restricting certain foods were associated with GI symptoms, and the administration of PPIs was associated with symptomatic improvement and macroscopic findings of EGD, suggesting a possible relationship between pediatric CG and eosinophilic gastritis.
As pediatric CG has no standard treatment, various treatments such as PPI administration, H2-receptor antagonists, iron supplementation, hypoallergenic diets, steroids, and azathioprine have been used [6,9]. In our study, PPI and hypoallergenic diets were administered to all children. Although the improvement in GI symptoms was helpful, the EGD findings were only partially improved or no significant difference was noted. Intravenous iron or oral supplementation was also provided in all children, and all the patients demonstrated improved Hb and ferritin levels only after intravenous iron supplementation. Blood transfusion was not required after the diagnosis of pediatric CG.
IDA is the most common nutritional deficiency in children and adolescents and is clinically important because it significantly impacts growth and development . Although pediatric CG has been previously known as a very rare disease, it may not be rare considering the recent literature and cases of patients in our hospital. Therefore, we recommend EGD for pediatric patients with recurrent IDA, especially if accompanied by GI symptoms, even if the symptoms are not severe. When performing EGD, the possibility of pediatric CG needs to be considered, and in case of a suspicious lesion, diagnosing pediatric CG should be attempted through performing multiple biopsies, including depressed mucosa and Masson’s trichrome staining test.
The authors have no conflict of interest to declare.