Does Helicobacter pylori play a role in the pathogenesis of childhood chronic idiopathic thrombocytopenic purpura?

Idiopathic thrombocytopenic purpura (ITP) is an acute self-limited bleeding disorder that can progress to chronic form in 10–15% of the cases. Helicobacter pylori (H. pylori) infection is a possible cause of chronic ITP. We studied 30 children with resistant chronic ITP for H. pylori infection based on the detection of H. pylori fecal antigen. This retrospective study was based on data obtained from medical records of 30 children aged between five and 17 years (median age at ITP diagnosis was ten years). A specially-designed data sheet was used to record information on age, sex, duration of disease, family history of bleeding disorders, previous treatments and median platelet count. In patients with H. pylori infection, antimicrobial treatment consisted of amoxicillin, metronidazol and omeprazol. Response was assessed every month for one year and defined as complete (platelet count >150×109/L) or partial (platelet count between 50 and 150×109/L). We detected H. pylori infection in 5 patients. In 4 of them increased platelet count was seen during one year of follow-up and in one patient the platelet count was acceptable during six months. Although the pathological mechanism of H. pylori-induced thrombocytopenia was unclear in our patient sample, the assessment of H. pylori infection and use of eradication therapy should be attempted in chronic and resistant ITP patients.


Introduction
Helicobacter is a genus of gram-negative bacteria causing peptic ulcer, chronic gastritis and lymphoma. 1 There is a suggestion that H. pylori has a role in the pathogenesis of chronic idiopathic thrombocytopenic purpura (ITP). 2 However, the relationship between H. pylori infection and cITP is less clear. In children, eradication of the infection was report-ed to produce an increase in platelet count in one study, 3 whereas another report failed to demonstrate this beneficial effect. 4 In adults, the connection between H. pylori infection and ITP is also controversial. 5,6 In the light of this controversy, we designed the present study to evaluate the role of H. pylori in patients with chronic and resistant forms of ITP and the effect of H. pylori eradication on platelet count in these patients.

Materials and Methods
This retrospective study involved 30 children with chronic ITP younger than 18  The median age at ITP diagnosis was ten years (range, 5-17 years). The inclusion criteria were thrombocytopenia of more than six months' duration (chronic ITP), lack of response to multidrug treatment with corticosteroids, intravenous immunoglobulin, RhoGAM (anti-D immunoglobulin) and splenectomy (for resistant forms of ITP), and platelet count below 100×10 9 /L ( Table 1). We used a specially designed data sheet to record information on age, sex, duration of the disease, median platelet count, type of previous treatment, and family history of autoimmune disorder. Stool samples were tested for H. pylori antigen with an enzyme-linked immunosorbent assay (ELISA) kit (Meridian BioScience, Cincinnati, OH, USA). In patients with H. pylori antigen an eradication regimen was given as follows: 1) omeprazol 1 mg/kg/day for four weeks, 2) metronidazol 30 mg/kg/day for two weeks, and 3) amoxicillin 60 mg/kg/day for two weeks.
Platelet recovery occurred one month after H. pylori eradication, which was confirmed by the disappearance of H. pylori antigen from stool. Subsequently, platelet count was checked every month for one year and compared with the patient's median platelet count before H. pylori eradication.
During the follow-up period, patients did not receive any medication. Response was defined as complete (platelet count >150×10 9 /L) or partial (platelet count between 100 and 150×10 9 /L ), and relapse was defined as a decrease in platelet count to pretreatment values.
Two statistical methods (paired t-test and Wilcoxon's signed rank test) were used to analyze the results. p less than 0.05 were considered statistically significant.

Results
We detected 5 H. pylori-positive cases according to the presence of the antigen in stool. None of the patients complained of gastrointestinal problems. The prevalence of H. pylori infection in patients with chronic ITP was 16.6%.
After H. pylori was eradicated in patients with the infection, platelet count was rechecked monthly for one year ( Table 2).
Median platelet count in our patients during one year after treatment increased from 18.6×10 9 /L to 79.2×10 9 /L with a p of 0.043 (Wilcoxon's signed rank test) and 0.001 (paired t-test). According to the definition of partial response (platelet count between 50-150×10 9 /L) our data showed an increase in platelet count in our patients in this category. In 4 patients, platelet count remained constant during one year of follow-up, and in one patient it remained stable for six months and then decreased to its pre-treatment value (Table 3). At this time this patient showed evidence of re-infection in the fecal antigen test.
By repeating H. pylori eradication, the patient's platelet count increased one month after treatment, and remained stable during follow-up. None of the other 4 patients showed evidence of H. pylori infection after one year of follow-up. In the H. pylori-negative group, platelet count was checked monthly for one year but did not increase spontaneously.

Discussion
In 1998, Gasbarrini et al. 2 first reported improvements in platelet count after H. pylori eradication in adults with chronic ITP. Several studies have shown that platelet count in patients with ITP returns to normal after eradication. 7,8 Although the investigation and eradication of H. pylori infection in children with chronic ITP is a matter of debate, there are few reports about children with this disease. [9][10][11][12] Although the association between the pathogenesis of ITP and H. pylori is still not well defined, the search for H. pylori infection and attempts to eradicate the bacterium in positive cases seem appropriate in adult patients as well as children. Some authors reported that the pre-treatment factor most consistently associated with platelet response to H. pylori eradication was shorter ITP duration, 13,14 but our study did not confirm this. Others have also suggested that patients with very low platelet counts appear to have fewer chances of responding, but we obtained a better response in this group. Torres et al. 15 found that H. pylori positivity was higher in older children, but we found no significant differences in this parameter between children of different ages. This may be due to earlier H. pylori infection in our country. 16 The rate of positivity for H. pylori infection was lower in our series (16.6%) compared to the Italian pediatric population, in which a prevalence of 22-28% was found. 17,18 Other characteristics such as age, sex, and previous therapies, including corticosteroids and splenectomy, were not useful in predicting platelet response. Our results indicate that eradication of H. pylori is accompanied by a platelet response in some ITP patients, with ample variations in the response rate among different series. Bacterial factors (i.e., the variability of H. pylori strains) may account for these findings. In this connection, Emilia et al. 8 reported a response rate of 50% whereas we obtained a response rate of 100%, which was consistent with the findings of Gasbarrini et al. 2 It seems that H. pylori eradication is greatly favorable to standard ITP therapy due to low cost, its non-invasive nature and much less toxicity. So H. pylori infection eradication should be considered in those patients with chronic and resistant ITP. 19 20×10 9 /L (10-30×10 9 /L) 50×10 9 /L (40-70×10 9 /L) immediately before H. pylori eradication Table 3. Platelet count in H. pylori-positive and -negative patients eradication.