Helicobacter pylori gastritis has been associated with autoimmune diseases, including idiopathic thrombocytopenic purpura (ITP). 1, 2 It is well known that the prevalence of H pylori infection is greatly variable from country to country, has decreased over the last decades in industrialized countries, and that its frequency increases with age. Recently, Veneri et al3 suggested that the HLA class II allele pattern might identify groups of ITP patients with different incidences of H pylori infection and, possibly, with different pathogeneses of thrombocytopenia. Indeed, eradication of H pylori in adults has been associated with platelet recovery in patients from Italy1, 2, 4 and from Japan5 but not in patients from Spain. 6 To our knowledge, no H pylori data have been published from pediatric patients with ITP. In children, ITP differs from that in adults in terms of clinical picture and mechanisms of thrombocytopenia. We therefore studied the prevalence of H pylori infection in a group of 17 children with chronic ITP. The patients had normal or increased megakaryocytosis in the bone marrow despite isolated thrombocytopenia that had lasted at least 6 months. Of the patients, 10 were girls and 7 were boys. Median age at the time of diagnosis was 3.8 years (range, 0.3-14.3 years), and median duration of thrombocytopenia at the time of the study was 3.9 years (range, 0.6-14.5 years). During a routine outpatient visit of the 17 consecutive patients, after informed consent, serum H pylori IgA and IgG class antibodies were measured by a locally validated enzyme immunoassay, 7 and in patients older than 5 years (n 12), a 13C urea breath test (Diabact UBT) also was performed. The 13C enrichment in the expired breath was measured by automated breath 13C analysis by means of continuous flow-isotope ratio mass spectrometry. 8 Delta-over-baseline (DOB) values were analyzed. DOB over 2.2‰ was considered positive. If any abnormalities were noted, H pylori antigen enzyme immunoassay was done in stool specimens.

H pylori infection was not diagnosed in any of the 17 patients. One patient had borderline titers of class IgG antibodies in the serum, but class IgA antibodies were negative and no antigen was detected in the stools. Two other patients had positive breathing tests (DOB 17.9‰ and 9.5‰), but their serum antibodies test as well as stool antigen tests were negative. In recent studies from Finland, an H pylori seroprevalence of 5.6% was found in children younger than 19 years9, and seroprevalence of 30% to 40% was found in adults. 10 The absence of infection signs in our pediatric ITP patients suggests that H pylori may not be enriched in this Finnish subgroup of patients. This is in