Unexplained Refractory Iron-Deficiency Anemia Associated with Helicobacter pylori Gastric Infection in Children: Further Clinical Evidence

Abstract

Helicobacter pylori infection in children has mostly been associated with recurrent abdominal pain, gastric dyspepsia, or duodenal ulcer. Uncommon clinical features such as protein-losing enteropathy (1) and malabsorption (2) have also been reported. In 1993 we first described a case of iron-deficiency anemia (IDA) not due to any conventional cause, which was associated with H. pylori gastric infection (3). The anemia was refractory to iron therapy and reversed only after bacteria eradication, thus suggesting possible interference of H. pylori in iron metabolism. In this article we describe four consecutive similar cases observed since then. To understand the possible pathogenic role of H. pylori in the development of the anemia, extensive investigations have been undertaken, and the internal cycle of iron metabolism has been studied in all patients. CASE REPORTS Case 1 A 4-year-old boy was admitted to a local hospital for severe IDA (hemoglobin, 5.9 g/dl) without gastrointestinal symptoms. After a packed red cell transfusion, he was discharged with a diagnosis of IDA of unknown origin. One year later (October 1991), he was admitted to our institution because of inefficacy of oral and intramuscular injections of iron. On examination he appeared pale and in poor general health. His weight was 15 kg (-3 standard deviations [SD]), height 102 cm (-3 SD). The hematologic profile showed persistent IDA (Table 1). Findings in an extensive initial work-up (including erythrocyte sedimentation rate; levels of C-reactive protein, total serum proteins and immunoglobulins; D-xylose absorption test; IgA antigliadin and antiendomysial antibody tests; bilirubin, lactate dehydrogenase, folate, B12 vitamin, and hemoglobin electrophoresis; multiple stool examinations for ova, parasites, and occult blood; barium meal and follow-through; technetium-99m pertechnetate imaging scans for ectopic gastric mucosa and occult gastrointestinal blood loss) and iron absorption test (Table 2) were all normal. Oral iron polimaltosate was administered (5 mg/kg daily elemental iron), but after 2 months of therapy, hemoglobin was unchanged, and the patient was readmitted for re-evaluation. IgG antibody titer to H. pylori (enzyme-linked immunosorbent assay; Helori-test, Eurospital, Trieste, Italy; normal value, <15%) was 71%. Findings in upper gastrointestinal endoscopy were normal. A rapid urease test result was positive. Histology showed H. pylori-positive chronic superficial antritis and normal duodenal mucosa. Analysis of bone marrow aspirate indicated a 4:1 myeloid-erythroid ratio with substantially normal morphology. Pearl's staining showed no detectable iron. After an internal iron cycle study (Table 3), a 1-month course of 50 mg/kg amoxicillin daily and 200 mg/m2 colloidal bismuth subcitrate daily was administered without iron supplement. After this therapy, no changes were observed in the patient's hematologic profile and iron stores. The child was then given 500 mg of intramuscular iron polimaltosate over 3 weeks. Three months later, H. pylori was no longer detectable in gastric histology and hematologic and iron profiles were normal (Table 1). They remained so after 3 years, as did the IgG antibody titer to H. pylori. Case 2 Iron-deficiency anemia developed in a boy aged 13 years, 7 months (hemoglobin, 9 g/dl) without gastrointestinal symptoms. After 2 months (September 1992), because there was no response to oral therapy with 100 mg/day of ferrous sulfate administered for 1 month, he was referred to our institution. On examination he appeared pale but in good general health; his weight was 48 kg (50th percentile), height 155 cm (25th percentile), Tanner pubertal stage P5. Iron-deficiency anemia (Table 1) was still present. Extensive initial work-up, similar to that in case 1, was performed, and test results were all normal including iron absorption test (Table 2). IgG antibody titer to H. pylori was positive (60%). Findings in upper gastrointestinal endoscopy were normal. A rapid urease test result was positive. Histology revealed H. pylori-positive chronic superficial antritis and normal duodenal mucosa. After an internal iron cycle study (Table 3), a 1-month course of 2 g amoxicillin and 200 mg/m2 bismuth subcitrate daily was prescribed without iron supplement. Two months after the end of treatment, there was no hematologic improvement. Another endoscopy showed persistence of H. pylori infection. A further 15-day course of amoxicillin plus 1.5 g tinidazole daily and 300 mg ranitidine per day was administered. Three months later, histology showed H. pylori eradication, but hemoglobin was unchanged. The patient took 100 mg oral ferrous sulfate daily for three months. Seventeen months later, hematologic profile, iron stores (Table 1) and IgG antibody titer to H. pylori were normal.TABLE 1: Hematologic and iron balance values before (pre) and after (post) Helicobacter pylori eradicationTABLE 2: Iron absorption testTABLE 3: Internal iron cycle studyCase 3 A girl aged 13 years, 6 months had recurrent epigastric pain and fatigue. Menarche had been at 11 years and 6 months of age and menstrual cycles were regular. Clinical examination findings were normal; weight was 53 kg (50th-75th percentile), height 158 cm (25th-50th percentile). The hematologic profile and iron balance (April 1993) showed IDA (Table 1). The initial work-up was performed as described earlier, and all test results were normal including the iron absorption test (Table 2). IgG antibody titer to H. pylori was positive (48%). Findings in upper gastrointestinal endoscopy were normal. A rapid urease test result was positive. Histologic analysis revealed H. pylori-positive chronic, active, superficial antritis and normal duodenal mucosa. After an internal iron cycle study (Table 3), 2 g amoxicillin daily and 200 mg/m2 bismuth subcitrate were prescribed daily for a month, and the epigastric pain disappeared. One month after the end of treatment, histologic analysis showed H. pylori eradication, but the hematologic profile was unchanged. Oral iron supplement of 100 mg/day ferrous sulfate was prescribed for 2 months. At the end of treatment, hematologic and iron profiles (Table 1) were normal. One year later, the patient was well, with hematologic profile, iron balance and IgG antibody titer to H. pylori within the normal range. Case 4 A 13-year, 6-month-old boy had iron-deficiency anemia (hemoglobin, 6.3 g/dl) without gastrointestinal symptoms. Findings in upper gastrointestinal endoscopy performed in a local hospital showed antral-erosive gastritis; H. pylori infection was not sought. Packed red cells were transfused, 300 mg ranitidine was administered daily for 1 month, and iron in the form of 800 mg iron polimaltosate was administered intramuscularly over 8 weeks. Two months later, another endoscopy showed healing of gastric mucosa, but IDA was still present. Antral histology revealed H. pylori infection; antimicrobial therapy (2 g amoxicillin daily for 15 days and 200 mg/m2 bismuth subcitrate twice daily for 1 month) and intramuscular iron supplement (400 mg iron polimaltosate over a month) were concomitantly administered. Because of the persistence of IDA without intestinal bleeding, 7 months later a further intravenous dose of 1 g elemental iron was administered over a month. Precise data regarding hemoglobin levels and iron stores were not available until 22 months later (hemoglobin, 8.7 g/dl, mean corpuscular volume [MCV] 75 fl, mean corpuscular hemoglobin [MCH] 23 pg, serum iron 22 µg/dl, and ferritin 7 ng/ml) when 2 g of elemental iron was again administered intravenously over a month. At the end of this four course of iron his hematologic profile was as follows: hemoglobin 12.9 g/dl, MCV 81 fl, and MCH 27 pg. Six months later (October 1994), at the age of 16 years, the boy was admitted to our institution because of persisting IDA (Table 1). He was in good general health: his weight was 59 kg, height 174 cm, and Tanner pubertal stage P5. The initial work-up was repeated, and all test results were normal, including iron absorption results (Table 2). IgG antibody titer to H. pylori was positive (52%). Upper gastrointestinal endoscopy showed micronodular antritis. A rapid urease test result was positive. Histologic analysis revealed chronic, H. pylori-positive, superficial antritis and normal duodenal mucosa. After an internal iron cycle study (Table 3) 20 mg omeprazole per day, 2 g amoxicillin twice daily, and 1 g tinidazole daily was administered for 15 days. A month after the end of treatment, histologic analysis showed H. pylori eradication and hemoglobin was 9.9 g/dl and MCV 68. Oral ferrous sulfate at 100 mg per day was prescribed for 3 months, and 2 months later hematologic profile and iron stores (Table 1) normalized. Eleven months later, levels were still normal, and IgG antibody titer to H. pylori was negative. DISCUSSION After the first case of unexplained refractory IDA associated with H. pylori gastritis, in which anemia was cured only after eradication of the infection (3), two similar cases have been reported (4,5). Here we present four more patients with an unexplained form of IDA in whom poor nutritional iron intake, gastrointestinal bleeding, malabsorption or chronic inflammation or infection, other than H. pylori gastritis, were excluded on the basis of specific test results. Cases 1, 2, and 4 were refractory to iron treatment, and IDA was reversed only by H. pylori eradication, which resulted in a consistently normal hematologic and iron profile. The outcome of all four patients demonstrates a causal role of H. pylori in the pathogenesis of the anemia. The results of the ferrokinetic study were not the ones expected in classic IDA, typically defined by a lower half-life and higher 59Fe red blood cell use than with normal iron status (6). In all our cases the 59Fe red blood cell use decreased, which may suggest that iron was diverted from the bone marrow to some extramedullary focus, possibly the H. pylori gastric infection. It is well known that microorganisms need the host's iron to grow (7). H. pylori acquires iron through outer membrane proteins from human lactoferrin (8), an iron-binding glycoprotein produced by inflammatory sites that is able to subtract iron from circulating transferrin. This prevents bacterial iron uptake from the transferrin, eventually inhibiting bacterial growth. Enhanced lactoferrin concentration in gastric mucosa correlates well with the presence of H. pylori infection and gastric inflammation (9). Moreover, it has been shown in vitro that human lactoferrin can support full H. pylori growth through specific bacterial receptors that are thought to play a major role not only in the virulence of the infection but also in its strict human specificity (10). From our data, including the results of the ferrokinetic studies, it may be hypothesized that in our patients H. pylori gastritis acted as a sequestering focus for iron through the following possible mechanism: H. pylori infection enhances gastric lactoferrin which in turn substracts iron bound to transferrin. This allows the bacterium higher iron uptake from lactoferrin, resulting in a vicious circle in which iron supply is unable to increase hemoglobin but enhances H. pylori growth. Our patient 4 received an immense amount of iron which might have saturated the hypothetical diverting extramedullary focus (i.e., infected gastric mucosa) and reached the bone marrow. This could explain his relatively high percentage of 59Fe use in red blood cells. In all patients the eradication of H. pylori may have caused a fecal loss of the bacterial iron stores together with the dead bacteria, making it no longer available for hemopoiesis. This is in keeping with the fact that in none of our cases did eradication of the bacterium per se normalize iron metabolism, and further iron supplement was necessary. In conclusion, we think that a diagnosis of H. pylori infection should be taken into consideration in children with unexplained refractory IDA. This seems particularly useful when considering the long clinical history of our patients and the useless and repeated iron supplementation administered. Although intriguing, the pathogenic mechanism by which H. pylori would induce this sort of anemia only remains speculative in that we did not seek 59Fe nor did we look for lactoferrin in the gastric tissue of our patients. Additional clinical observation and studies on the bacterium strains are needed to explain why IDA seems to be only a rare consequence of such a frequent infection.