Efficacy of Helicobacter pylori Eradication Therapy on Platelet Recovery in Pediatric Immune Thrombocytopenic Purpura-Case Series and a Systematic Review

Evidence relating to the efficacy of Helicobacter pylori eradication therapy for chronic immune thrombocytopenic purpura (cITP) in childhood is inadequate. The aim of this retrospective study was to determine the efficacy of H. pylori eradication therapy for platelet response in pediatric patients with cITP in our hospital, and to perform a systematic review of previous reports about pediatric patients with cITP who were positive for H. pylori infection and were treated with H. pylori eradication therapy. Analysis of the data of pediatric patients with cITP in our hospital and a systematic review of digital literature databases of studies in pediatric patients with cITP were performed. Data of 33 pediatric patients with cITP from our hospital records showed that the prevalence of H. pylori infection and the rate of response to platelet therapy were 15% and 33.3%, respectively. Data of 706 pediatric patients from 18 previous reports showed that the prevalence of H. pylori infection and rate of platelet response were 23% and 43.8%, respectively. Eradication therapy for H. pylori infection in pediatric cITP patients can be expected to result in a response equivalent to that in the adult population, with fewer adverse effects than other treatments for cITP.


Introduction
Helicobacter pylori (H. pylori) is a microaerophilic Gram-negative spiral bacterial pathogen that resides in the stomach. Urease produced by the bacteria protects them from the highly acidic environment. H. pylori infection is generally acquired in childhood and typically persists for life. Chronic infection with H. pylori is a well-known cause of gastric diseases including chronic active gastritis, atrophic gastritis, peptic ulcer diseases, mucosa associated lymphoid tissue lymphoma, and gastric cancer [1]. Many studies have shown conflicting results relating to the association between H. pylori infection and extraintestinal diseases including nutritional deficiencies, such as vitamin B12 deficiency, and hematological pathologies such as iron deficiency anemia and idiopathic or immune thrombocytopenic purpura (ITP) [2,3].

Patients
Medical records of all patients with the term "thrombocytopenic" or "thrombocytopenia" in their diagnosis at the Juntendo University Hospital between January 2002 and December 2018 were reviewed retrospectively. The patients selected for the study met the following inclusion criteria: (1) age <18 years at the time of diagnosis of ITP and (2) diagnosis of ITP according to the American Society of Hematology criteria [5] based on an initial platelet count <100,000/µL. The exclusion criteria were as follows: (1) thrombocytopenia related to autoimmune disorders, drugs, a family history consistent with inherited thrombocytopenia, human immunodeficiency virus (HIV) infection, hepatitis, congenital heart disease, or pseudothrombocytopenia and (2) previous history of H. pylori eradication before the diagnosis of ITP. In this study, we defined cITP as persistent thrombocytopenia for longer than 6 months. Among cITP cases, data of patients who underwent diagnostic tests for H. pylori infection were extracted and the results of eradication therapy were reviewed. All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Institutional Ethics Committee of Juntendo University Hospital (the project identification code: no. 20-184, date of approval: 25 August 2020).

Diagnosis of H. Pylori Infection, and Eradication Therapy
H. pylori infection was diagnosed using the 13C-urea breath test (UBT), H. pylori stool antigen test (HpSA), or serum anti-H. pylori IgG antibody test. Based on previously published reports, pediatric patients were considered positive for UBT if a delta over baseline ≥3.5% was achieved [29]. The gastric biopsy samples for bacterial culture were obtained endoscopically from patients who tested positive for any of the diagnostic tests for H. pylori. Antibiotic sensitivity of H. pylori isolates to clarithromycin (CAM), amoxicillin (AMPC), and metronidazole (MNZ) were examined using the dilution method with Mueller-Hinton agar plates containing 5% horse blood. The minimal inhibitory concentration (MIC) breakpoints were defined as follows; ≥1 µg/mL for CAM, >0.125 µg/mL for AMPC, and >8 µg/mL for MNZ [30].
Patients infected with CAM-sensitive strains were prescribed the CAM-based regimen for 7-14 days, which consisted of a proton pump inhibitor (PPI), AMPC (50 mg/kg per day), and CAM (20 mg/kg per day). Patients infected with CAM-resistant strains were prescribed the MNZ-based regimen for 7-14 days, which consisted of PPI, AMPC (50 mg/kg per day), and MNZ (20 mg/kg per day). H. pylori eradication was confirmed by UBT or H. pylori stool antigen test performed at least 8 weeks after treatment.

Assessment of Treatment Efficacy
We evaluated the treatment efficacy as follows. A response was considered complete if the platelet count increased to at least 100,000/µL within 2 months of initiation of anti-H. pylori treatment. Partial response was defined as an increase in platelets to a minimum of 50,000/µL within 2 months of initiation of treatment regardless of whether the patient was on maintenance therapy, or an increased platelet count to more than twice that during the pretreatment period. No response was defined as a platelet count that remained below 50,000/µL or showed no increase. ). A total of 401 articles were retrieved. For ICHUSHI, we performed a detailed search using (((thrombocytopenia (in Japanese)/TH or thrombocytopenia (in Japanese)/AL) or (thrombocytopenic purpura (in Japanese)/TH or thrombocytopenic purpura (in Japanese)/AL) or (ITP/TA)) and ((Helicobacter/TH or Helicobacter/AL) and pylori/AL) or ([Helicobacter pylori]/TH) or ([Helicobacter infection (in Japanese)]/TH) or ("Helicobacter pylori"/TH and Helicobacter infection (in Japanese)/TH) or (Pylori (in Japanese)/TA) or pylori/TA))) and (PT = except record of proceedings (in Japanese)), and found 527 articles. A total of 928 articles were retrieved from the combined searches. Study references were collected in Endnote x9 (Clarivate Analytics Japan, Minato-ku, Tokyo, Japan).

Study Selection
Two reviewers (TI and TK) independently performed primary screening of the articles based on their titles and abstracts blinded to each other's decisions. Studies were considered eligible for inclusion if they related to the effects of H. pylori eradication on the platelet count in pediatric patients with cITP in randomized-controlled, case-controlled, or retrospective studies. The exclusion criteria were as follows: (1) papers not written in English or Japanese, (2) research not involving humans (e.g., in vitro or animal research), and (3) research including patients with secondary thrombocytopenia due to autoimmune disorders, drugs, family history consistent with inherited thrombocytopenia, HIV infection, hepatitis, or congenital heart disease. Selection or rejection of articles that could not be decided through primary screening were retained for secondary screening. The results of the two reviewers were crosschecked, and a dataset for secondary screening was prepared which included 153 full-text articles. Two reviewers (TI and TK) independently read the full-text articles, and performed the secondary screening. Articles that included evaluation of the effects of H. pylori eradication on the platelet count in pediatric patients with cITP were selected independently, and the results of the two reviewers were crosschecked. Where the two reviewers had different opinions that could not be resolved by discussion, a third person was invited to make a final decision. All decisions were recorded in Microsoft Excel (Microsoft, Redmond, WA, USA).

Data Extraction
Two reviewers (TI and TK) independently extracted data from each study using a data extraction form that was designed in advance. The following information was extracted: first author, year of publication, study design, definition of cITP and platelet response, age of patients, diagnostic tests of H. pylori infection, eradication regimens, duration of eradication therapy, confirmatory test of eradication, and confirmation period of platelet response. Where the two reviewers had different opinions that could not be resolved by discussion, a third person was invited to make a final decision. All decisions were recorded in Microsoft Excel (Microsoft, Redmond, WA, USA).

Study Quality Assessment
Methodological quality of included RCTs was assessed using Jadad score [31] the following items: (1) randomization; (2) blinding; (3) withdrawals and dropouts. Nonrandomized studies (NRCTs) including cohort or case control study were assessed in accordance with the Newcastle-Ottawa scale (NOS) [32]. They were assessed by selection, comparability, and exposure or outcome.

Retrospective Analysis
We evaluated 642 pediatric patients in whom the term "thrombocytopenic" or "thrombocytopenia" was used in the diagnosis. After exclusion of thrombocytopenia due to other causes, 80 patients with ITP between January 2002 and December 2018 were selected. Thirty-three patients showed persistent thrombocytopenia for more than 6 months, meeting the criteria for cITP (Table 1). Diagnostic evaluations for H. pylori were performed in 20 patients, and three of them had at least one positive diagnostic test result. The results of H. pylori culture in all three patients were positive; thus, the prevalence of H. pylori infection among 20 pediatric patients with cITP was at least 15% even in the unlikely event that false-negative cases were included.
All three patients received H. pylori eradication therapy, which was successful in all of them after the first course of antimicrobial treatment. One patient showed a complete response and continues to have a satisfactory platelet count more than 12 months following eradication therapy. No significant increase in the platelet count was seen in the other two patients. Therefore, the response rate in our patients was 33.3%.

Remission Case Report
A 10-year-old girl diagnosed with cITP was admitted with recurrent purpura, nasal bleeding, and thrombocytopenia for more than a year. She also complained of recurrent episodes of epigastric pain about twice a week, for several years. Investigations performed previously at another hospital showed persistent low platelet count below 50,000/µL, elevated platelet associated immunoglobulin G (196.0 ng/10 7 cells), and normal results of white and red blood cell counts and bone marrow examination (nucleated cells: 18.9 × 10 10 /L, megakaryocytes: 150 × 10 6 /L, M/E ratio: 1.46). She had been prescribed several courses of prednisolone, intravenous immunoglobulin, and methylprednisolone when the platelet counts fell below 20,000/µL and was symptomatic. However, no significant response was achieved with these treatments.
At our hospital, after complaining of chronic epigastric pain and giving a positive family history of H. pylori infection, she was investigated for the presence of H. pylori infection. UBT and serum anti-H. pylori IgG antibody tests were positive. Esophagogastroduodenoscopy revealed nodular gastritis on the antral mucosa, and H. pylori was detected on bacterial culture of the biopsy sample. After confirmation of antibiotic sensitivity to AMPC and CAM, eradication therapy comprising AMPC (1500 mg/day), CAM (400 mg/day), and lansoprazole (45 mg/day) was prescribed for 14 days. The patient's platelet count increased dramatically and reached 187,000/µL 4 weeks after H. pylori eradication. She had received no other treatment for cITP for a month before starting H. pylori eradication therapy. She maintained complete remission over a year and was advised no further follow-ups.

Study Selection and Characteristics of Included Studies
A flow diagram of study identification, inclusion, and exclusion is provided in Figure 1. The searches of PubMed and ICHUSHI using the defined terms yielded a total of 928 records. After 10 duplicates were removed, 918 records remained for the title and abstract screen. A total of 695 records were excluded based article type (reviews, case reports, irrelevant experiments, incomplete data). After full-text assessment of the 223 remaining records, we finally identified and reviewed 18 studies [33][34][35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50] consisting of two RCTs and 16 NRCTs (case-controlled, or retrospective studies, or case series) with sufficient data relating to H. pylori infection and the efficacy of H. pylori eradication therapy in pediatric patients with cITP. Characteristics of the 18 studies are summarized in Table 2. They were published between 2003 and 2015. We extracted the data of 706 pediatric patients with cITP persistent thrombocytopenia (platelet count <150,000/µL) for longer than 6 months, including 164 patients with H. pylori infection from these previous studies (Table 2).

Study Selection and Characteristics of Included Studies
A flow diagram of study identification, inclusion, and exclusion is provided in Figure 1. The searches of PubMed and ICHUSHI using the defined terms yielded a total of 928 records. After 10 duplicates were removed, 918 records remained for the title and abstract screen. A total of 695 records were excluded based article type (reviews, case reports, irrelevant experiments, incomplete data). After full-text assessment of the 223 remaining records, we finally identified and reviewed 18 studies [33][34][35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50] consisting of two RCTs and 16 NRCTs (case-controlled, or retrospective studies, or case series) with sufficient data relating to H. pylori infection and the efficacy of H. pylori eradication therapy in pediatric patients with cITP. Characteristics of the 18 studies are summarized in Table 2. They were published between 2003 and 2015. We extracted the data of 706 pediatric patients with cITP persistent thrombocytopenia (platelet count <150,000/µL) for longer than 6 months, including 164 patients with H. pylori infection from these previous studies (Table 2).    CCS: case control study, cITP: chronic immune thrombocytopenia, CO: cohort study, CR: complete response, CS: case series, CSS: cross sectional study, H. pylori: Helicobacter pylori, Jadad: Jadad score [31], N/A: not applicable, N/D: not described, NOS: Newcastle-Ottawa Scale [32], PR: partial response, RCT: randomized control trial. ∆: describing the platelet increase. Response includes both complete response and partial response. † : result of a pooled data analysis.

Risk of Bias
The two RCTs demonstrated an acceptable level of quality according to Jadad score, 3 or higher. The Newcastle-Ottawa Scale scores of the six cohort and two case control studies ranged from 5 to 9, indicating moderate risks of bias.

The Prevalence of H. Pylori Infection and Platelet Response Rate
The total prevalence of H. pylori infection among all 726 cITP patients including our patients was 23.0%, and a wide variability in prevalence from 0% to 43.5% depending on the reports was observed. Among 167 cITP patients with H. pylori infection, 132 patients were prescribed eradication therapy. The platelet response rates after eradication therapy also showed a wide variability from 0% to 100%, and the total response rate was 43.8%; a significant increase in platelet count following H. pylori eradication therapy was observed in 60 patients. When limited to the cases in Japan, the prevalence of H. pylori infection was 11.6%; 11 of 95 cITP patients tested positive for H. pylori, and the platelet response rate was 42.9%; however, it should be noted that the data were based on only three responders among seven cITP patients.

Discussion
ITP is an autoimmune disease that causes thrombocytopenia due to the production of autoantibodies against platelets, and is often triggered by some infection or vaccination in childhood. Although 80-90% of pediatric patients with ITP recover spontaneously, some cases are refractory to the administration of glucocorticoid and immunoglobulin and may experience severe bleeding. There are regional, ethnic, and racial differences; however, there is much evidence in adults for an association between H. pylori infection and ITP. The mechanism by which H. pylori infection causes ITP has been investigated from various perspectives. Takahashi et al. reported that cross-reaction of the platelet membrane glycoprotein (GP) IIb/IIIa, GP Ib/IX, GP Ia/IIa, and GP VI, the antigen corresponding to antiplatelet autoantibodies, with the H. pylori constituent protein CagA caused thrombocytopenia [22]. Asahi et al. showed that H. pylori infection resulted in inhibition of the Fcγ receptor on peripheral blood monocytes, and led to increased production of antiplatelet antibodies, thereby accelerating platelet turnover through decreased expression of FcγRIIB [12]. However, the pathogenesis of ITP and its causal relationship with H. pylori infection is still not clear.
Given that children have a lower prevalence of H. pylori infection and lower incidence of refractory ITP than adults, data from previous reports are limited to a small number of patients. Although Russo et al. [35] in Italy, Treepongkaruna et al. [38] in Thailand, and Brito et al. [34] in Brazil reported the results of prospective, multicenter, and interventional trials investigating the relationship between H. pylori infection and ITP in childhood, it is still difficult to accumulate comprehensive evidence with an adequate sample size to determine the efficacy of H. pylori eradication therapy in children with cITP who test positive for H. pylori.
Thus, these suggestions are based on extremely limited information, although the prevalence of H. pylori infection among children with cITP may be slightly higher than that in the general pediatric population as observed in Italy, Taiwan, and Japan. The prevalence reported in systematic reviews from Japan (11.6%) was about the same as, or slightly higher than, the prevalence in other children of the same age in Japan, in whom it was approximately 5-10%. This trend was also observed in Japanese adults, among whom the prevalence of H. pylori infection in cITP patients (69%) was slightly higher than that in the general population of the same age group (about 60%) [20]. This lack of a significant difference in the infection rate indicates that ITP is a multifactorial disease and that H. pylori infection is just one of the causes of ITP in adults and children.
A wide variability in the rates of platelet increase from 0% to 100% after H. pylori eradication therapy was observed in the review of previous reports. However, when all previous cases were combined, the mean rate of platelet increase was 43.8%. This result proves the existence of a subset of pediatric cITP patients who would benefit from H. pylori eradication therapy, and suggests that a response rate equivalent to that observed in the adult population (40-60%) can be expected [20]. Therefore, H. pylori eradication therapy should be considered as the first treatment for cITP in children with H. pylori infection, given the adverse effects of systemic glucocorticoid administration and splenectomy. The cost of treatment and fewer side effects with H. pylori eradication therapy makes it a favorable first choice.
Our study has several inherent limitations that need to be discussed. First, this systematic review included a limited number of studies with a small sample size indicating moderate risks of bias, and there were only three prospective, multicenter, interventional trials [34,35,38]. Second, the indication for H. pylori diagnostic testing in cITP pediatric patients depended on the decision of the main attending physician in our retrospective analysis. Thus, 13 of 33 pediatric cITP patients in our hospital were not evaluated for H. pylori infection. Further multicenter studies or large randomized controlled trials are necessary to establish an adequate evidence base for this therapy in pediatric patients with cITP.

Conclusions
This study, based on a systematic review of pediatric cITP reports from Japan and abroad, and a retrospective analysis of 33 cITP patients from our hospital, shows that the efficacy of H. pylori eradication therapy in inducing a platelet response among pediatric patients with cITP is equivalent to that seen in adults. The study suggests that H. pylori eradication therapy should be considered as the first-line treatment for cITP in children with H. pylori infection.