Helicobacter pylori is recognized as a major gastrointestinal pathogen in developing countries. This microorganism infects up to 60% of children less than five years in those countries and is strongly associated with chronic gastritis and peptic ulcer disease in children and adults. The progression of gastritis to atrophy often leads to decreased gastric acid output, which is a well-known risk factor for anemia. Gastric acid is essential for increasing the bioavailability and absorption of non-heme dietary iron, the most important source of iron in developing countries. Numerous reports suggest that iron malabsorption secondary to low gastric acid output is a problem in developing world countries. It has been further observed that iron deficiency anemia is resistant to iron therapy particularly in these countries. In a recently completed study we observed an association of anaemia with H. pylori infection. We hypothesize that the poor bioavailability of iron in these countries could be related to H. pylori -induced low gastric acid output and we propose to investigate the role of H. pylori infection as a cause of anemia and treatment failure of iron supplementation in Bangladesh. A prospective, randomized, double-blind, placebo-controlled field trial is proposed among four groups ( 65 each) of H. Pylori infected children of 2-5 years of age with iron deficiency anemia. The children will be assigned to one of the four therapies: antibiotics alone (for H. Pylori eradication), antibiotic plus iron therapy, iron therapy alone, or placebo. Hemoglobin concentration, serum ferritin concentration, and transferrin receptor will be measured before and at 1 and 3 month after the intervention. We also propose a complementary study in an additional 20 children with H. Pylori infection and iron deficiency anemia to assess iron absorption with application of double stable isotopes. The change in hematological parameters will also be compared among the groups before and after the therapy. The results of this study are expected to have implications in the prevention and treatment of iron deficiency anemia in developing countries.

We propose to complementary study:

1. A prospective, double-blind, placebo-controlled field trial in two hundred sixty children of a peri-urban community near Dhaka City is proposed.

2. The assessment of iron absorption before and after anti-infective therapy in 20 children with H. Pylori infection and iron deficiency anemia

3. Field trial

   1.1 Description of the study population Children will be recruited from Nandipara, a peri-urban community situated 7 miles northeast of Dhaka City. This community has a population of 3,500 in an area of approximately 2.5 square miles. Among the residents, 70% of males are classified as day laborers, 20% are rickshaw pullers, and 5% are carpenters or service holders. Fifteen percent of women are day laborers and 85% are housewives. The average family size is 4.5 members. The slum has a municipal water supply for drinking and cooking and most families live in poorly constructed houses. The great majority of houses are mud walled with thatched dry leaves and bamboo or tin roofs. The population less than five years of age is about 700. The prevelance of H. pylori infection in less than five children in this community is 50%.44 A weekly clinic has been run by the International Centre for Diarrhoeal Diseases Research, Bangladesh (ICDDR,B) for minor case of illness in the population since 1986.

   All children in the community will initially be screened for iron deficiency anemia and H. pylori infection. After explanation of the study and the tests, informed consent will be obtained from the parents or legal guardians. Demographic information including name, date of birth, sex, weight, height, economic status of the parents, nature of feeding, and previous medication in the past 1 month will be recorded.

   Screening for iron deficiency anemia 50 µl capillary blood samples obtained by fingerstick will be assessed for hemoglobin (Hgb) concentration. Measurement of hemoglobin concentration is the most useful test in screening iron deficiency anemia since it directly reflects the quantity of most abundant essential iron compound in the body. Children with hemoglobin concentration below 110g/l will be considered as anemic. In children with low Hgb identified through the initial screening, the diagnosis of iron deficiency anemia will be confirmed by venepuncture Hgb, serum ferritin concentration, and serum transferrin receptor concentration. Haemoglobin concentration can be transiently depressed due to inflammation or infection, and in such a situation does not indicate pathological anemia. Of all the biochemical tests for assessing iron status, a low SF is the most specific test for iron deficiency. The major disadvantage of SF is that it is readily elevated in response to any infector inflammatory condition. One major advantage of sTfR is that unlike SF, sTfR is not significantly effected by infection or the inflammatory process. However, sTfR levels are increased in situations of increased RBC turnover such as hemolytic anemia, i.e. thalassemia, etc. A uniform cutoff value of >8.5 mg/l indicates an elevated sTfR and therefore iron deficiency or hemolytic anaemia.

   Measurement of serum transferrin receptors also becomes a promising new test for the evaluation of iron status particularly in epidemiological assessment of anemia. Performed in conjunction with serum ferritin measurements, the serum transferrin receptor serves to distinguish true iron deficiency from the anemia from the anemia of chronic diseases or inflammatory diseases and, therefore, offers a major advantage in establishing the true prevalence of iron deficiency anemia in population studies. The combination of low hemoglobin, a low serum ferritin, and an elevated serum transferrin receptor are diagnostic of iron deficiency anemia. The diagnosis of iron deficiency will be made when the following criteria are met: transferrin receptor > 8.3 mg/l, serum ferritin concentration <12ng/ml, and hemoglobin concentration(<11 gm/dl).

   Screening of H. Pylori infected children The anemic children will be screened for H. Pylori infection using a urea breath test (UBT) which is considered by many of the reference diagnostic test for H. pylori infection. We have recently established the test at ICDDR,B.44 In brief the procedures are as follows.

   Urea Breath Test A breath sample will be collected for base line 13 CO2 in a vacutainer following a fasting period of 2 hr. A volume of 100 ml of whole milk will then be allowed to be consumed by the children. Ten min later 40 mg of 13 C urea (99% 13C, Tracer Technologies, Boston, Massachusetts) will be given after dissolving in 25 ml of water. After administration of the substrate, breath samples will again be collected in duplicates after 30 min. Samples will be collected through two-way pediatric mask with an attached non-return valve into a vacutainer and will be stored for shipment to the University of Alabama, to measure the 13C concentration of a respiratory CO2 mass spectrometer. The excess over the baseline will be expressed as parts per thousand(delta 0/00). A breath test in which the excess over a baseline is 5.00/00 at 30 min will be regarded as positive for H. pylori infection.

   1.2 Selection of Study children Iron deficiency anemic children with weight for age >60% of National Center for Health Statistics (NCHS), with no evidence of deficiency diseases or systemic infection and whose parents give informed consent will be enrolled in this study.

   Exclusion criteria The children with following problems will be excluded.

   • Acute infection or apparent inflammatory process
   • Signs of vitamins deficiency
   • Severe anemia (Hemoglobin <70 G/l)
   • Severe malnutrition (marasmus, marasmic kwasiorkar or kwasiorkar)
   • Presence of hook worms and /or Giardia lamblia (cyst or vegetative form) in a stool microscopic examination
   • Presence of fat in a stool microscopic examination
   • Presence of occult blood in a stool as demonstrated by Guaiac test
   • History of taking antibiotics or any drugs for any cause in the preceding month

   1.3 Procedure A detailed medical history will be recorded. Body weight and height will be measured upon enrolment into the study. Naked weights will be obtained to the nearest 10 g using a balance scale (Gebrüder Soehnle, Murrhardt, Germany) and the mean of three consecutive measurements recorded. Recumbent length will be measured in children less than two years of age using a slideboard infantometer (Harpenden, St. Albans, England). In older children, standing height is to be determined using a locally constructed instrument in which a metal tape measure is extended between a footplate and head bar that has an attached fluid level to ensure a horizontal position. The mean of two consecutive height measurements to the nearest 0.1 cm is recorded as the observed value. Measurements are compared to the standards according to the National Center for Health Statistic data and the nutritional status assessed by Z-score. , Percent of median equivalents of standard deviation units will be determined and categorized according to the scheme of Waterlow.

   All children will be subjected to physical examination by one of the investigators. The children will be assigned to one of the four groups according to randomization:

   • Anti Hp group
   • Fe group
   • Anti -Hp group + Fe group
   • Placebo group

   In order to compare the overall prevalence of resistance to iron therapy among iron deficient children, there will be a fifth group comprising children with iron deficiency anemia but without H. pylori infection. This group will be treated with iron therapy.

   Anti Hp therapy: Clarithromycin 15 mg/kg/day in two divided doses for 10 days plus omeprazole 20 mg single dose for 14 days. ,

   Iron therapy: An iron dose of 3 mg/kg/day (or about 30 mg/day) as ferrous sulphate as advocated by Dallman41 will be fed in a single dose in the morning in one half hour before food. The iron therapy will be given for 90 days.

   Placebo Glucose syrup

   1.3.1 Supplementation The intervention syrup(s) will be fed daily in a single dose in morning in an empty stomach (Half an hour before food). The children in the placebo group will receive same amounts of glucose syrup. All therapies will be kept in identical dark brown containers and will be fed by a health worker who will be kept unaware about the nature of the study and of the contents inside the bottle. Twenty four-hour food intakes will be recorded on a precoded form by dietary recall.

   The children will be followed for three months after the end of supplementation. Urea breath test will be repeated at the end and at 1 and 3 months after the supplementation. Venous blood (500 µl) will be taken for determination of serum ferritin, transferrin receptor and hemoglobin concentration at the end and after 1 and 3 months of supplementation. Children in the placebo group will be given a three month course of supervised iron supplementation at the end of study period.

   1.4 Sample size calculation The sample size is calculated on the basis of (i) prevalence of H. Pylori infection, (ii) prevalence of non response to iron therapy.

   Sample size calculation based on the prevalence on H. Pylori infection

   Considering the prevalence of H. Pylori infection 50% and iron deficiency anemia 70%,44 we anticipate H. Pylori related anemia in 50% of infants and children in this community. We expect a reduction in prevalence of H. Pylori related iron deficiency anemia after therapy with antibiotic plus iron therapies 25%. To detect a difference of this magnitude at 5% level and at 80% power the sample size is calculated as follows:

   N =[P1(100-P1) +P2(100-P2)]*8/(P1-P2)2

   Where P1= prevalence of H. Pylori related anemia P2= prevalence of H. Pylori related anemia after therapy

   =[50(50) +25(75)] * 8 / (25)2

   =56 in each group

   Based on a previous study in this community we anticipate a drop out rate of approximately 15%. With a (Akramuzzaman SM, personal communication), the necessary sample size will be 65 children in each group and the total sample of children will be 260.

   Calculation based on prevalence of non-responder to iron therapy

   Observation from an ongoing study with an oral iron supplementation programme in anaemic patients being conducted in Dinajpur (a northern district of Bangladesh) suggests that 70% of the iron supplemented group do not respond ( S. Rahman, personal communication). The prevalence of H. Pylori infection in the proposed study population is approximately 50%,35 and we expect that after oral iron supplementation the non responders will be 40% less among the H. pylori negative group, compared to the H. pylori positive group. To detect a difference of this magnitude at 5% level and at 80% power the sample size is calculated as follows:

   N = [P1(100-P1) +P2(100-P2)]*8/(P1-P2)2

   Where P1= prevalence of anemia after iron supplementation in H. pylori infected children P2= prevalence of anemia after iron supplementation in H. pylori non-infected children

   • [(70)(30) +(42)(58)] * 8 / (35)2 =46 in each group

   With 15% drop out rate, the number of children will be 53 in each group i.e. total sample of 212 children.

   We will take the larger sample size, i.e. 260 children for this study.

   1.5 Randomization A master randomization chart will be prepared by an appropriately trained person who is not connected with the study using random permuted block numbers taking a size of three. Bottles containing anti-Hp therapy, iron therapy or placebo will be identical in appearance and will be arranged in sequence that corresponds to the randomization chart and will be serially numbered. The serial number of the bottles will correspond to the serial number of the patients enrolled in the study. As one group of children would receive anti-Hp therapy and iron therapy simultaneously, there will be two bottles (bottles A or B) for each group of children to make the study double blind. For children belonging to anti-Hp plus Iron therapy, there will be one bottle containing anti-Hp therapy or and the other containing iron therapy. Children belonging to iron therapy or anti-Hp therapy alone, will be getting one bottle containing the medication, while the other will contain glucose syrup. As the duration of anti-Hp therapy is 14 days, the children belonging to this group will be provided glucose syrup for the rest of the month in order to make the duration of therapy equal among the four groups.

   1.6 Laboratory test Hb concentration will be determined by the cyanmethemoglobin method, serum ferritin will be determined by a radio immunometric assay. Serum transferrin receptor will be measured by enzyme-linked immunosorbent assay (ELISA). Stool will be tested for the presence of occult blood by the guaiac reaction (Occultotest, Ames Co.,Elkhart, Inc.,. USA). Serum pepsinogen and gastrin will be measured by radioimmunoassay.

   1.7 Withdrawal from the study - The noncompliance of subject, either because the leaves the study area for other place before the end of the study or

   - because the child requires unscheduled treatment for serious interim infection.

   Treatment of withdrawal during analysis Results of all randomized children will be included in the analysis of the study. Data from the patients withdrawn will be included up to the time of withdrawal. A supplemental analysis in which such patients are excluded will also be made

   1.8 Outcome variables

   Major

   - The incidence of H. Pylori infection following the therapy.

   - Rise of serum hemoglobin, serum ferritin concentration or reduction of level of serum transferrin receptors.

   - The number of children in each group with IDA as primarily defined before and after the intervention

   Minor

   - Change of anthropometric indices

2. Iron absorption study Twenty additional children with a positive Urea breath test and iron deficiency anemia will be selected for the Iron absorption study. The purpose of the study and all procedures related to iron absorption study will be explained to the parents and consents will be obtained. Before commencing the iron absorption study, all children will be given a commercial whey adjusted cows' milk formula with 8 mg Fe/l (Nidina 1; Nestle, Paris, France) for a period of at least three weeks. After 21 days of formula feedings they will be admitted to the Metabolic Ward of Clinical Research and Service Center.

Erythrocyte incorporation of labeled iron (radio stable) has been used successfully to examine absorption from the gut. , The erythrocyte incorporation method involves oral administration of an isotope of iron followed by measurement of the isotope ratio of iron in erythrocytes at a selected time after administration. We will follow the technique for measuring iron absorption in children by a double stable isotope technique, developed recently by Kastenmayer et al.

2.1 Method for Iron absorption study

Preparation of labeled iron sulfate solutions Enriched 57Fe and 58Fe-ferric oxide (Fe2O3) will be purchased from Oak ridge National Laboratory (Oak Ridge, TN, USA). The isotope compositions of iron in the 58Fe preparation (as provided by Oak ridge National Laboratories) are (mg): (weight%) 54 Fe (0.65), 56 Fe (35.6)57 Fe (3.47), and 58 Fe(60.28). The isotope compositions of iron in the 57Fe (weight%) are; 54 Fe(0.65), 56Fe(35.60), 57Fe (3.47), and 58Fe (60.28).

Ferric oxide will be dissolved in concentrated H2SO4 by bubbling carbon monoxide through the solution over heat for approximately four hours. The solution will then be derated by passing purified nitrogen through it for several minutes. The volume will then be adjusted using derated water. Two ml aliquot of the resultants 57 FeSO4 and 58 FeSO4 solutions will be transferred into opaque ampules. Each ampule will be purged with nitrogen and sealed. Ampules will contain either 200 µg/ml of total iron (as ferrous sulphate) and 150 µg/ml of 57Fe or 700 µg/ml of total iron (as ferrous sulphate) and 420 µg/ml of 58Fe. Aliquot from an individual ampule will be used to prepare 57Fe and 58Fe-spiked working solutions.

Stability and safety of parenteral 57 Fe solutions Prior to i.v. administration to the infants the 57FeSO4 solution will be tested for pyrogenicity, sterility, and stability. Limulus amoebocyte lysate (LAL) testing and repeated culture for bacterial and fungal growth to confirm that the 57FeSO4 solutions are pyrogen-free and sterile at the concentrations to be used in the study. The nutrient content of the parenteral solutions has been previously described. The total iron content of the 57FeSO4 -nutrient solution will be determined by atomic absorption spectrometry before and after filtration (0.22 µ filters), and over time at room temperature.

2.2 Study design of iron absorption study Infants enrolled in the study will receive a combination of parenteral and enteral feeds. The 57FeSO4 solution will be provided parenterally while the 58FeSO4 will be given via the enteral route. Erythrocyte incorporation of 57Fe and 58Fe will then be measured from blood samples taken two weeks after initial infusions. The percentage of the I.V administered dose presents in the circulating erythrocytes will be used as a reference for the percentage of absorbed iron (from the orally administered dose) present in the circulating erythrocytes. The "corrected" percent erythrocyte 58Fe and 57Fe incorporation (58 Fe inc-cor/ 57Fe inc-cor) will be calculated mathematically as described below.

Prior to iv administration to the children, 0.7-0.8 ml of the 57FeSO4 solution(0.2mg/ml) will be added to a burytrol containing 30 ml of the nutrient solution used for parenteral feeding, using sterile techniques. The iron-nutrient mixture, which has been used safely previously, will be infused slowly into a peripheral vein of the children over a 12-24 hour period with the aid of an infusion pump.

Soon after the start of the parenteral infusion of the 57FeSO4 solution, the enteral 58FeSO4 will be given to the children. During this time, parenteral infusion of 57FeSO4 will be continued. After checking to ensure that the nasogastric feeding tube is properly located in the stomach, the tube will be aspirated to check stomach contents. If the stomach is empty, 1.5 ml of the 58 FeSO4 will be delivered directly via a syringe into the feeding tube. The tube will then be flushed with 0.5 ml of sterile water. The dosage of parenteral 57FeSO4 will be 0.15mg/kg (57Fe-enriched ferrous sulfate solution), which would provide 0.12 mg/kg of 57Fe. The dosage of enteral will be 1.5mg/kg (58Fe-enriched ferrous sulfate solution), which will provide 0.8mg/kg of 58Fe. The children will be discharged with therapy for Helicobacter pylori infection (Klarithromycin plus omeprazole) for 10 days.

After one month of therapy the mothers will be requested to bring their children at the CRSC. A repeat UBT and iron absorption study will be performed.

2.3 Blood sampling Blood samples will be taken before administration of the stable isotopes (baseline) and 14 days later in both occasions. Blood sample size will be 200-300 µl up per sample.

2.4 Calculation of Erythrocyte Iron Incorporation The quantity administered 58Fe and 57Fe (mg) incorporated into erythrocytes at specified time 't' after administration of the dose of labeled iron will be calculated in a manner similar to that used by Zlotkin et al 1995.

58 Fe inc. =Rt58/54 -Ro58/54 xFe circ x 0.00322 Ro58/54

57Feinc =R157/54-Ro57/54 x Fe circ x 0.0218 Ro57/54

In this equation, 58 Feinc and 57Feinc will be expressed in mg. R158/54, R157/54 and R057/ 54 are the isotope ratios expressed on weight basis for 58Fe/54Fe and 57Fe/54Fe, which will be determined on digested blood samples at time "t" after dosing and at a baseline (pre 58Fe and pre 57Fe doses) respectively; Fe circ is the quantity of the circulating iron. Fe circ at the time "t"; 0.0032 and 0.0218 are the natural abundances (weight fractions) of 58 Fe and 57 Fe. The quantity of total circulating iron, Fe circ, expressed in mg, will be estimated as

Fe circ. =BV x Hb X 3.47

Where BV is the blood volume in liters, assumed to be 0.085l/kg of the body weight; Hb is the hemoglobin concentration in g/l; 3.47 is the concentration of iron in Hb (mg/g).

The quantity of administered 58Fe or 57Fe incorporated into erythrocytes, % Fe inc, will be expressed as a percentage of the administered doses of 58Fe or 57Fe. Feinc.

• 58 Fe inc = 58Fe inc x 100 or Fe inc
• 57 Fe inc = 57 Fe inc x 100 57 Fe inc

The corrected percent erythrocyte for analysis of the stable isotope ratios in erythrocytes will be calculated using the percentage of erythrocyte incorporation of i.v infused 57 Fe.

• 58Fe inc-cor = % 58Fe inc X 100.
• 57 Feinc

Mass spectrometric analysis Preparation of blood samples for mass spectrometric analysis and preparation of enriched standards will be followed as per the procedures described by Kastenmayer et al. Samples of heparinized whole blood will wet ashed with concentrated HNO3 at a ratio of 1:3 (blood: acid)(vol/vol) on a hot plate over low heat for 20-30 min. Digested solutions will be adjusted with deionized water to yield an iron concentration of 10 µg/ml. Isotope-spiked standards for an instrument calibration will be prepared in the same manner as with the unknown samples. Isotope ratios for 58Fe/54Fe and 57Fe/54Fe in blood before and after concurrent enteral (58Fe) and parenteral (57Fe) administration of isotopes will be determined by a computer-controlled thermal ionization mass spectrometer using an Elan Model 250 ICP/MS system, (SCIEX, Inc., Thornhill, Ontario, Canada), operated in the isotope ratio mode, as followed by Zlotkin et al. in the Department of Pediatric and Nutritional Science, University of Tronoto, Canada.

Intubation test A standard intubation test as described earlier14 will be performed in this subset of children after completion of iron absorption study. In brief, as soft gastric tube (18G) will be placed in most dependent part of stomach. Gastric acid samples will be collected from children lying in the left recumbent position using manual suction. The position of the tube will be confirmed by auscultation, aspiration of gastric juice, and nearly complete recovery of water injected in the tube. Samples of gastric juice will be discarded during the first 10 min after intubation. Thereafter, samples will be collected and saved in 15- min aliquot for 90 min. After 90 min of collection (basal sample), pentagastrin (peptavlon, ICI) will be injected subcutaneously at a dose of 0.6 ug/kg. The sample following pentagastrin stimulation for 60 min will be termed as "stimulated sample". The volime of gastric secretion in each of the 15 min collections will be rcorded and expressed as mililitre per hour. The concentrations of acid in each will be measured by titration of 1 mL gastric juice with 0.01N sodium hydroxide to pH 7.4, using an automated titrator (Metrohm, Herisau, Switzerland). Total acid output was calculated for each time pint by multiplying volume of gastric juice and acid concentration. Basal (BAO) and stimulated acid output (SAO) were calculated based on 2 samples collected for baseline measurement and the 4 samples collected after administration of pentagastrin respectively, and were expressed as mMol/h.

Inclusion Criteria:

• Iron deficiency anemic children
• with weight for age >60% of National Center for Health Statistics
• no evidence of deficiency diseases or systemic infection
• Informed consent of the parents

Exclusion Criteria:

• Acute infection or apparent inflammatory process
• Signs of vitamins deficiency
• Severe anemia (Hemoglobin <70 G/l)
• Severe malnutrition (marasmus, marasmic kwasiorkar or kwasiorkar)
• Presence of hook worms and /or Giardia lamblia (cyst or vegetative form) in a stool microscopic examination
• Presence of fat in a stool microscopic examination
• Presence of occult blood in a stool as demonstrated by Guaiac test
• History of taking antibiotics or any drugs for any cause in the preceding month