Anemia is an important public health problem in developing countries and very often it is a possible consequence of a common nutritional defect, iron deficiency. The possible role of H. pylori infection in the development of hyposideremic anemia was recently investigated in five Latin America countries, Argentina, Bolivia, Brazil, Cuba, Mexico, and Venezuela , but no evidence was found to confirm the responsibility of such an infection. Brazilian individuals were investigated in greater depth  and, although no significant association was observed between anemia and H. pylori infection, a crude multilevel linear regression showed a reduction of 0.07 g/dL in those who were colonized, after adjusting for sex, skin color, income, age, and smoking. A major problem in those countries, however, is that only approximately 50% of anemia cases can be attributed to iron deficiency; other causes, which include malaria, hookworm infestation, schistosomiasis, inherited conditions such as thalassemia and dietary vitamin deficiency do not always emerge in the clinical history of individuals.
Numerous case reports published in minor journals revealed that the eradication of H. pylori infection resolved iron-deficiency anemia [52–58]. However, a large intervention study carried out on Mexican children showed that the bacteriological healing of the infection led to a significant increase of hemoglobin concentration only if children also received iron supplementation . In this survey, however, it was also demonstrated that most children had an iron deficit without anemia and that 2/3 ca. of children with iron deficiency or anemia were not infected by H. pylori, signifying that other factors may play a role in the development of anemia.
Muhsen et al.  stressed the importance of establishing the CagA status of patients which lacks in most surveys. They found low ferritin levels, respectively, in 14.5% and 8.6% of H. pylori infected and uninfected Israeli Arab children. Despite the fact that low ferritin levels were mostly detected in CagA-positive subjects, it should be considered that the infection by strains expressing CagA enhances the risk of developing peptic ulceration and reduces the levels of gastric ascorbic acid. Both conditions which may concur to cause iron-deficiency anemia through gastrointestinal blood loss and insufficient dietary iron absorption, thus complicating the question even more.
A condition that may lead to a chronic idiopathic iron deficiency is represented by autoimmune atrophic gastritis, which has been shown to be responsible for refractory iron-deficiency anemia in over 20% of patients with no evidence of gastrointestinal blood loss . Such a disease is considered a possible outcome of a long lasting H. pylori infection. Infected subjects, in fact, have circulating antibodies to the H+,K+-ATPase of the gastric parietal cells .
H. pylori infection is a condition in which autoimmunity is exalted; we therefore aligned the amino acid sequence of catalase, an enzyme abundantly expressed by erythrocytes, with peptides expressed by H. pylori J99, to see whether mechanisms of molecular mimicry could account, at least partially, for the development of anemia in infected individuals. We found a linear homology with numerous bacterial proteins, the widest of which was with the bacterial catalase.
In conclusion, to better define the role of H. pylori infection in iron-deficiency anemia, as well as its pathogenic mechanisms, we need larger controlled trials, the definition of the CagA status and exclusion of all the other causes of anemia, including the presence of autoantibodies to erythrocytes.
Idiopathic Thrombocytopenic Purpura (ITP)
The possible role of H. pylori infection in the development of ITP is a subject of extensive investigation. Systematic reviews of past literature [62,63] showed an overall platelet response in more than 50% of the patients successfully treated for the infection and increased response rates in countries with a high prevalence of H. pylori infection in background populations, i.e. in patients with less severe degrees of thrombocytopenia and in those with shorter disease duration. In the meta-analysis performed by Arnold et al. , the cumulative sample size of cases was 282 patients with ITP (pooling 11 studies, eight from Japan), 205 of whom were H. pylori positive and 77 patients H. pylori negative. All patients underwent eradication treatment. The odds of achieving a platelet count response following eradication therapy were increased by 14.5 times (95% CI 4.2–83.0) in patients with H. pylori infection (51.2% vs 8.8%).
Not all studies revealed a beneficial effect of eradication. In a multicentre randomized controlled trial performed in Thailandese children with ITP, no favorable consequence of H. pylori eradication on platelet recovery was observed . It should be underlined that the platelet recovery at 6 months was demonstrated in only one of the seven infected patients in the treatment group and in one of the nine subjects in the control group (55 children with ITP were examined and 16 of them (29.1%) were infected). Compared to other studies on the same subject carried out on children and adults [63,65], the rate of improvement of ITP, consequent to the healing of infection, was very small.
As far as the putative mechanisms that may account for such an association are concerned, an Italian group  suggested that infection by H. pylori strains expressing CagA could play a role in chronic ITP cases. A positive response to the eradication was observed in 43% of patients after 6 months of follow-up. Patients who responded and those who did not respond were comparable for all the principal clinical features but not for anti-CagA serum antibodies, which were present in 83% and 12.5% of cases, respectively (p = .026). These researchers concluded that an antibiotic treatment of the infection should be considered in ITP cases, particularly for patients who harbor CagA antibodies.
Such an observation prompted us to align the amino acid sequence of CagA with human platelet peptides, to verify the existence of a linear homology, but we did not find any similarity. However, a platelet component, a 289 amino acid receptor binding domain for the von Willebrand factor (GP1b platelet), showed a certain structural similarity to a putative partial recombinant VacA paralog of H. pylori J99. It is known that almost all of the CagA-positive H. pylori strains also express the vacuolating toxin; therefore, it is possible that the observation made by Scandellari et al.  actually implicated the existence of a molecular mimicry of some platelet peptides with VacA; the eradication of the infection may lead to the disappearance of immune cross reacting antibodies and therefore to the healing of the platelet disorder. A study has shown a cross-reaction of an H. pylori urease B monoclonal antibody with platelet glycoprotein IIIa and suggested that the immune response to UreB may be involved in the pathogenesis of ITP . The fact that all H. pylori isolates express urease activity does not weaken the importance of the observation that ITP in adult patients may be associated with anti-CagA antibodies  because it is known that the ratio of strains expressing CagA varies in different countries from < 30% to > 80% of H. pylori isolates.
In conclusion, together with George , we believe that the cure of H. pylori infection totally corrects the thrombocytopenia in certain patients. In other patients with ITP, the infection can be considered as an additional disorder which aggravates the main disease, while in a third group of patients the eradication of H. pylori appears to have no effect on the course of thrombocytopenia.