Resolution of Unexplained Iron Deciency Following Successful Eradication of Helicobacter Pylori in Children

Background: Evidence is needed to inform whether Helicobacter pylori (HP) treatment is benecial in children with refractory iron deciency. We aimed to assess association between successful HP eradication and resolution of unexplained iron deciency. Methods: Medical records of children diagnosed with HP infection (based on histopathology) and without signicant upper gastrointestinal source of blood loss, were retrospectively reviewed for presence of iron deciency. Among those with non-anemic iron deciency (NAID) or iron deciency anemia (IDA), hemoglobin, ferritin and C-reactive protein (CRP) levels were compared prior and 6-9 months’ post successful HP eradication. Patients with overt bleeding or subsequent iron supplement therapy post HP diagnosis were excluded. Predictors of resolution of iron deciency following HP eradication were assessed. Results: Among 60 included children (median age 14.8, IQR12.3-16; 62% males), symptoms of anemia were observed in 20%. A total of 21 (35%) had IDA while the remainder 65% had NAID. Following successful HP eradication, 60% of these 60 patients normalized their iron status. There were signicant improvements in both hemoglobin and ferritin following HP eradication with hemoglobin increasing from 12.3g/dL to 13.0 g/dL (P<0.001), and ferritin increasing from 6.3μg/l to 15.1 μg/l (P<0.001). In multivariate logistic regression that assessed sex, ethnicity, baseline hemoglobin, anemia or GI symptoms, previous iron therapy, and time from diagnosis to eradication, older age was the only factor associated with resolution of anemia following HP eradication: (OR 1.65, 95% CI 1.16-2.35, P = 0.005). Conclusion: Successful HP eradication could be helpful in improving iron status among children with refractory NAID or IDA. Older age may predict this outcome. Screening for HP should be considered in the workup of refractory IDA or NAID.


Introduction
Iron de ciency anemia (IDA) and non-anemic iron de ciency (NAID) are among the most common micronutrient de ciencies globally. They are particularly worrisome in children due to their harmful effect on growth, motor and cognitive functions (1). Helicobacter pylori (HP) is a major gastric pathogen infecting approximately half of the world population (2,3), It is predominantly acquired during childhood and persists throughout life, unless therapeutically eradicated (4). HP infection causes chronic gastritis which remains asymptomatic in most infected children. Only about 15% of infected persons develop peptic ulcer disease (5)(6)(7). The association between HP infection and IDA has been a focus of attention for more than a decade (8,9). HP seroprevalence is more common in children with refractory IDA (10). Children infected with HP are considered to be at a higher risk of IDA and NAID (11,12). The mechanism by which HP contributes to iron de ciency (ID), in absence of gastric bleeding remains unclear; however, Several mechanisms have been postulated including the alteration of gastric acidity (10,(13)(14)(15)(16), the induction of atrophic gastritis, reduction in ascorbic acid in gastric secretions (17), the development of gastric microerosions, in ammation and an up-regulation of hepcidin levels (17)(18)(19)(20)(21)(22). However, ID and IDA does not develop in all HP infected patients, only in approximately 25% of them (23,24). Unexplained and refractory IDA is considered as one of the indications for testing for HP infection in children and for eradication if the patient is found to be positive (24). Despite this clinical practice, evidence for the association between HP eradication therapy and resolution of childhood IDA is still matter of debate (23,(25)(26)(27)(28)(29). Ethnic dissimilarities as well as differences in diagnostic criteria may possibly explain differences in those studies. Regardless, the Joint ESPGHAN/NASPGHAN guidelines and the British society of gastroenterology recommend considering eradication of HP infection in patients with IDA and unremarkable esophagogastroscopy (19,30). The primary aim of this study was to assess the resolution of NAID and IDA in pediatric patients after the successful eradication of HP. The secondary aim was to explore the contribution of in ammation, measured as CRP levels, to the change of ferritin and Hgb concentration with HP eradication.

Study population:
The study was conducted at Emek Medical Center, Israel. Patients were included if they were aged 2-18 years and ful lled all of the following criteria: (1) con rmed diagnosis of HP infection (30), (2) presence of NAID or IDA based on ferritin and hemoglobin (Hgb) levels and using age and sex-speci c cutoffs (31,32), (3) lack of overt gastrointestinal (GI) bleeding, (4) successful subsequent HP eradication, and (5) without iron supplement therapy following the diagnosis of HP infection. Electronic medical records from one year prior to and one year following the performance of endoscopy were reviewed.
Exclusion criteria included patients with signi cant source of chronic and/or acute GI blood loss identi ed on upper and/or lower endoscopy, clinical or histological diagnosis of celiac disease, in ammatory bowel disease, eosinophilic gastroenteritis, malabsorption disorders, obvious cause of non-GI loss (e.g., epistaxis, signi cant menstrual bleed), known inherited blood disorder (e.g., thalassemia minor), malignancy or those without accessible iron status measurements following eradication or those receiving ongoing iron therapy.

Study design
This retrospective cohort study included the medical records of children diagnosed with HP infection, without overt upper gastrointestinal source of blood loss, from 2005 to 2020. Demographic, clinical, laboratory, endoscopic and histological ndings were collected. Hgb, ferritin and C-Reactive Protein (CRP) levels were assessed before and after the eradication of HP. Our primary outcome was the resolution of IDA or NAID following HP eradication, without iron supplementation, de ned by normalization of ferritin and Hgb levels adjusted for age and sex (31)(32)(33). Baseline Hgb and ferritin were obtained at least 3 months prior to HP eradication trial. Post treatment Hgb and ferritin were obtained within 6-9 months' post successful HP eradication.
Determination of HP status HP status was determined by presence of HP organisms on histological evaluation of gastric biopsies, positive rapid urease test or positive gastric culture for HP. In patients with discordant results by positive rapid urease and/or histology, positive HP serology or presence of chronic active gastritis on biopsy were used to determine the presence of active HP infection (30,34). Patients with discordant histology results, who were on proton pump inhibitors (PPIs) and had no serology tests, were not included. Patients in whom diagnosis was based on non-invasive methods (urea breath test, HP fecal antigen or serology tests) without upper gastrointestinal endoscopy were not included.

Determination of NAID and IDA
In this study, we followed the guidelines of WHO and American Academy of pediatrics (AAP), which have de ned ID as ferritin levels below 12 µg/L and 15 µg/L in otherwise healthy children under and above 5 years, respectively (31,32). Anemia was de ned based on age and sex according to the WHO criteria (33). Altitude correction was not needed as all patients lived below 1000m above sea level altitude. Although WHO recommends adjusting ferritin based on CRP levels (35), there was a very weak and non-signi cant correlation between ferritin and CRP at baseline (r = -0.03, P = 0.8157) and after the therapy (r = -0.03, P = 0.8571) in our study. Therefore, we did not apply the BRINDA method to adjust serum ferritin concentration based on in ammation status(36, 37).

Statistical analysis
Data were analyzed using SPSS (version 21.0, SPSS, Inc., Chicago, IL, USA). Continuous variables were presented as either mean ± SD or median with interquartile range (IQR) depending on the data approximation to normal distribution. In order to analyze the factors predictive of ID resolution, Fisher's exact test was used to explore univariate associations between primary outcomes and categorical variables. Association between variables at triple therapy initiation and resolution of ID was evaluated using univariate and multivariate logistic regression. Multiple logistic regression analysis was then performed using variables that were statistically signi cant on univariate analysis. The difference in ferritin before and after the therapy were compared between patients with elevated and non-elevated CRP levels at the baseline (elevated CRP level is de ned as CRP > 5 mg/L) to determine whether in ammation may modify the association between HP eradication and improvement in iron status. All reported Pvalues are two-sided. P-values < 0.05 were considered signi cant.

Effect on Iron status
Prior to HP eradication, mean Hgb and ferritin levels were 12.3 g/dl and 6.3 µg/l respectively. When including only 21 patients with IDA, mean Hgb and ferritin were much lower 11.1 g/dl and 3.9 µg/l respectively. Post HP eradication, Hgb levels increased signi cantly in the whole group to a mean Hgb of 13 as well as among IDA patients to a mean of 12.5 as shown in Fig. 2a and 2b (P < 0.001). In addition, ferritin levels were signi cantly higher following successful HP eradication (Fig. 2c) (P < 0.001). Prior to HP eradication, of 60 patients, 21 (35%) had IDA while the remainder 65% had NAID. Post successful eradication 60% of the total 60 patients normalized their iron status as shown in Fig. 3.

Predictors of resolution of IDA/NAID
In univariate logistic regression, older age, higher baseline Hgb levels and being NAID rather than IDA prior to triple therapy initiation were associated with resolution of ID at end of follow-up: ( respectively. Multivariate logistic regression demonstrated that only older age was an independent, predictive factor for resolution of ID at end of follow-up: (OR 1.65, 95% CI 1.16-2.35, P = 0.005). Sex, ethnicity, presence of self-reported anemia symptoms, previous iron therapy, time from diagnosis to eradication and the presence of GI symptoms were not found to be associated with the resolution of ID (data not shown)

Correlation between in ammation and the iron status
After the treatment, the mean serum ferritin concentration increased by 8.64 (CI: 4.48-12.8) µg/l among children who had elevated CRP (de ned as CRP > 5.0 mg/L), compared to 8.68 (CI: 6.27-11.1) µg/l among children with normal CRP level as shown in Fig. 4.

Discussion
In this study, we have found that the eradication of HP in children was associated with a signi cant increase in Hgb and ferritin levels. The increase in iron stores was much higher in older children and those with pre-eradication higher Hgb. Older age was the sole independent predictive factor associated with the resolution if ID.
The treatment of IDA in children is a public health priority due to its harmful long-term effects and relative ease and availability of diagnosis and treatment (38). For decades, researchers have been concerned with the relationship between HP and IDA and the importance of HP eradication in IDA treatment. This issue has major implications especially in low resource settings where both conditions are prevalent (39) and HP diagnostic tools are not always accessible (40). Also in a uent settings, concerns over antibiotic resistance are alarming (41).
Our results are an important contribution to the growing body of literature exploring the correlation between HP eradication in children and the resolution of NAID and ID (42)(43)(44). Our study is unique in that it has studied patients not treated with supplementary iron and can thus focus on the sole effect of HP eradication on iron absorption without dietary changes. Kurekci et al (45) have also previously demonstrated an improvement in Hgb and iron stores after the eradication of HP in children, but unlike their study, ours have relied on a histological evidence of HP infection.
HP infection is usually acquired in childhood and its signi cance as a colonizing pathogen is still debateable. NAID and IDA are also very common in childhood. Several risk factors such as lower socioeconomic status, poor nutrition, household crowdedness and being a minority group or from a developing country are common between HP and IDA (25) , (29) , (46) , (47) making exploring the causation a harder task. Nevertheless, several large studies have managed to show a correlation between NAID / IDA and HP infection and an improvement in the iron stores after the eradication of HP (48). It is still unknown how lasting this effect is. In a randomized controlled trial in rural Alaska, Gessner et al did not demonstrate any improvement in Hgb or iron stores 14 months after the eradication of HP (29). When the same cohort was examined at 40 months post eradication, a modest improvement in iron stores was noticed in the HP negative group (49). Sarker et al found that HP was not related to IDA or iron treatment failure in Bengali children (25). Our results suggest that HP eradication without iron supplement is associated with the resolution of ID/IDA at 6 months' post eradication.
Older age as a predictive factor of the resolution of IDA was observed in several studies, Duque et al have observed that Mexican children with HP infection and IDA were older than their schoolmates with IDA without HP infection (27). Baggett et al also observed that in Alaskan children, the relationship between HP infection and IDA was age dependent and reached signi cance in older children. This was explained by a longer duration of infection causing a larger loss in iron stores (29). This might explain the nding that in our cohort older children had a better improvement in iron stores compared to younger children, since longer duration of HP infection in those cases might have contributed initially to their anemia.
CRP is an accurate marker of in ammation. It has been found to be higher in adult patients with HP infection, also those with peptic ulcer disease and asymptomatic carriers compared to healthy adult controls (50). In children on the other hand, the ndings in the literature are unequivocal. Azab  These results align well with our results, for in our cohort there was no correlation between CRP and the iron status in all children before and after HP eradication. This nding contributes to the growing debate regarding the role of systemic in ammation in HP induced IDA in children.
Limitation of our study include those inherent to retrospective studies like the absence of randomization and of a control group. Other limitations include being a single centre study. Though our sample is similar in Hgb levels to larger samples and can be considered representative. Some of the strengths of our study is the use of biopsy and gastric culture and not serological testing as an indicator of HP infection.

Conclusion
In conclusion, we have found a clear correlation between HP eradication and improvement in Hgb levels and iron stores in children 6 months' post eradication. The improvement was much more signi cant in older children. Larger prospective studies are needed in order to better understand the complex relationship between HP and ID/IDA and the potential bene t compared to the harms of HP eradication in children.   Proportions of patients with normal iron status, iron de ciency or iron de ciency anemia before triple therapy and after successful eradication of Helicobacter pylori infection.