No Evident Causal Effect between Helicobacter pylori and Idiopathic Thrombocytopenic Purpura: A Twosample Mendelian Randomization Study

Abstract

Background: The relationship between Helicobacter pylori (H. pylori) infection and Idiopathic Thrombocytopenic Purpura (ITP) is contentious, with conflicting evidence regarding H. pylori’s impact on ITP. This study aims to clarify this potential causal relationship using a Mendelian Randomization (MR) approach.
Methods: We utilized a two-sample MR study to assess the causal relationship between H. pylori infection (primary GWAS ID: ukb-b-531) and ITP. Inverse Variance Weighted (IVW), Weighted Median (WM), and MR-Egger regression were employed to estimate the causal effects. Additional analyses were conducted using two other H. pylori GWAS datasets and six immunogenic protein molecules of H. pylori.
Results: The MR analysis found no significant causal association between H. pylori and ITP across all methods. Consistent results were obtained from additional datasets. No causal relationship was observed between H. pylori’s six immunogenic proteins and ITP. Funnel plots showed symmetry, Cochran’s Q test indicated no heterogeneity except for UREA, MR-Egger intercept test and MR-PRESSO global tests suggested no horizontal pleiotropy. Leave-one-out analyses show that no single instrumental variable was responsible for the causal effects. Steiger directionality test results confirmed the correctness of the assessment direction.
Conclusions: This MR study provides robust evidence against a causal relationship between H. pylori infection and ITP, challenging the notion that H. pylori eradication could influence ITP treatment outcomes. Further research should focus on alternative mechanisms and confounding factors influencing ITP.

Keywords:Mendelian randomization, Helicobacter pylori, Idiopathic thrombocytopenic purpura, Causality, GWAS, CagA, VacA

Introduction

Idiopathic Thrombocytopenic Purpura (ITP), also known as immune thrombocytopenic purpura, is an autoimmune disorder characterized by a significantly reduced platelet count, which can lead to severe bleeding and bruising. The condition results from the immune system mistakenly targeting and destroying platelets [1]. The pathogenesis of ITP involves complex interactions between auto antibodies and platelets, leading to their destruction and impaired production. Despite extensive research, the precise mechanisms underlying ITP remain unclear, making identifying contributing factors crucial for improving treatment strategies. One such potential factor is Helicobacter pylori (Hp) infection, which has garnered considerable attention due to its possible impact on ITP [2].

Helicobacter pylori is a gram-negative bacterium commonly found in the human stomach, where it can cause chronic gastritis and peptic ulcers [3]. The potential link between H. pylori infection and ITP emerged when some studies reported increased platelet counts in ITP patients following H. pylori eradication therapy [4,5]. This observation led to a surge of interest and subsequent research into the relationship between H. pylori and ITP. However, the findings have been inconsistent and often contradictory. Some studies suggest that eradicating H. pylori can significantly improve platelet counts in ITP patients, while others report no effect or even adverse outcomes [5-8]. This variability has fueled ongoing debates and highlighted the need for more rigorous investigations to establish a clear causal relationship.

The conflicting results from previous studies may stem from several factors, including differences in study design, patient demographics, and the presence of confounding variables that are challenging to control in observational and cross-sectional studies. Observational studies are susceptible to reverse causation and confounding biases, which can obscure true causal relationships [9]. Given these limitations, a more robust methodological approach is required to clarify the link between H. pylori infection and ITP. Mendelian Randomization (MR) offers a powerful alternative for assessing causal relationships by using genetic variants as instrumental variables [10]. Mendelian Randomization (MR) leverages the random assortment of genes from parents to offspring during meiosis, mimicking the randomization process in a controlled trial [11]. By using genetic variants associated with the exposure of interest (in this case, H. pylori infection), MR can provide unbiased estimates of the causal effect of the exposure on the outcome (ITP). This approach helps to overcome confounding and reverse causation biases inherent in traditional observational studies [12]. In the context of H. pylori and ITP, MR can clarify whether H. pylori infection directly influences the development and progression of ITP or if the observed associations are due to other confounding factors [13]. This study aims to employ Mendelian randomization analysis to investigate the causal relationship between H. pylori infection and ITP. By integrating genetic data and advanced statistical techniques, we seek to provide more definitive evidence regarding the role of helicobacter pylori in ITP pathogenesis. This research has the potential to resolve longstanding controversies and inform clinical practice, particularly in managing Hp infection in ITP patients. Establishing a causal link could lead to novel therapeutic strategies, enhancing patient outcomes by targeting Hp eradication as part of ITP treatment protocols.

Materials and Methods

Study Design

This study employed a two-sample Mendelian randomization approach to explore the causal relationship between Helicobacter pylori infection and idiopathic thrombocytopenic purpura. The study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization (STROBE-MR-Checklist) guidelines. The design of the MR analysis aimed to satisfy the three core assumptions required for a valid MR study: (1) The genetic variants used as instrumental variables (IVs) must be strongly associated with the exposure (Hp infection). (2) The IVs must not be associated with any confounders that influence both the exposure and the outcome (ITP). (3) The IVs affect the outcome solely through the exposure and not through any alternative pathway (no horizontal pleiotropy) [14].

Data Sources

Genetic associations with Hp infection were sourced from the IEU Open GWAS Project. The primary dataset (GWAS ID: ukb-b-531) was used with replication analyses conducted using two additional datasets (GWAS IDs: ebi-a-GCST90006910 and ieu-b-4905). Furthermore, in line with previous Mendelian randomization studies investigating causality between Hp and other disorders, we similarly selected datasets for six Hp-related proteins, including GroEL (GWAS ID: ebi-a-GCST90006913), OMP (GWAS ID: ebi-a-GCST90006914), UREA (GWAS ID: ebi-a-GCST90006915), VacA (GWAS ID: ebi-a-GCST90006916), catalase (GWAS ID: ebi-a- GCST90006912), and CagA (GWAS ID: ebi-a-GCST90006911) [15]. The outcome dataset for ITP (GWAS ID: ebi-a-GCST90018865) also originates from IEU open Gwas. All participants in the GWAS were of European ancestry. Table 1 provides detailed characteristics of all the data included in this study.

Selection of Genetic Instruments

Instrumental variables for Hp infection and related bacterial factors were selected from the respective GWAS datasets. For the primary analysis of Hp (GWAS ID: ukb-b-531) and the GroEL protein dataset, a threshold of p<5x10-5 was used. For the replication analyses involving two addition Hp datasets and five other Hp-related protein (OMP, UREA, VacA, catalase, and CagA) datasets, a more stringent threshold of p<5x10-6 was applied. These thresholds were chosen to ensure a sufficient number of single nucleotide polymorphisms (SNPs) for robust analysis and are consistent with thresholds used in previous Mendelian randomization studies [16-18]. To ensure the independence of these SNPs, we employed a clustering approach based on low linkage disequilibrium (LD, r2 <0.001) and substantial physical distance (>10,000Kb). This method enabled us to identify reliable SNPs. To mitigate potential confounding factors, we used the PhenoScanner database to exclude SNPs associated with the outcome variables. We further assessed the strength of these SNPs as instrumental variables using the F-statistic, adopting a threshold of F>10 to avoid introducing of weak instrument bias. Additionally, potential outliers were identified and removed using the MR-PRESSO outlier test, thereby enhancing the results’ accuracy and reliability.

Statistical and Sensitivity Analysis

The primary analysis was conducted using the Inverse Variance Weighted (IVW) method, widely recognized for providing the most robust results under all conditions when horizontal pleiotropy is absent [19]. Complementary analyses included the Weighted Median (WM) method and MR-Egger regression [20]. Assessment for potential horizontal pleiotropy was conducted through the MR-Egger intercept test; deviations of the intercept from the origin with a significance threshold of p<0.05 suggested potential horizontal pleiotropy [21]. Heterogeneity among studies was evaluated using Cochran’s Q test, with heterogeneity indicated if p<0.05, necessitating using a random-effects model for analysis [22, 23]. To ensure result reliability, a leave-one-out approach (Loo) was employed, sequentially excluding each independent instrumental variable (IV), followed by Bonferroni correction (corrected P value: 0.05/9 = 0.006) [24,25]. Findings were deemed statistically significant if p<0.006; p values between 0.006 and 0.05 suggested a possible significant causal relationship [26,27]. Additionally, funnel plots were utilized to visually represent results [28]. The robustness and reliability of findings were further verified through the MR-PRESSO global best [29]. The Steiger directionality test was conducted to ensure the correct assessment direction of a causal relationship [30]. The two-sample Mendelian randomization study was conducted using the Two Sample MR 0.6.2 package, operating within the R version 4.3.3. environment.

Results

This study initially analyzed the causal relationship between the primary helicobacter pylori dataset (GWAS ID: ukb-b-531) and ITP. The results indicated that assessments using the IVW method (OR: 2.33 x 105, 95% CI: 3.85x10-15-1.41x1025, P = 0.595), WM method (OR: 6.44 x 10-7, 95% CI: 3.47x10-32-1.20 x1019, P: 0.631), and MR-Egger (OR 1.20x10-133, 95%CI: 0.00-4.76x10157, P: 0.380) all showed no causal association between helicobacter pylori and ITP (Figure 1). Repeated MR analyses using two additional summary statistics for Hp also yielded consistently negative results (Figure 1). Furthermore, no causal relationship was found between all six Hp immunogenic protein molecules and ITP (Figures 1&2).

Figure 1:Overview of the study design. SNP, single nucleic polymorphism; ITP, idiopathic/immune thrombocytopenic purpura.

Figure 2:Forest plots for the causality of Hp, GroEL, OMP, UREA, VacA, catalase, and CagA on ITP. ITP, idiopathic/immune thrombocytopenic purpura; OR, odds ratio; SNP, single nucleic polymorphism; Ab, antibody; 95%CI, 95% Confidence interval.

All funnel plots exhibited a symmetric pattern of effect size variability around the point estimates (Figure 3). Cochran’s Q test indicated no heterogeneity in the instrumental variable effects for the primary Hp dataset (Q_Pval: 0.311) and two additional Hp datasets (Q_Pval is 0.064 and 0.051 respectively) (Table 2). Among Hp’s six immunogenic protein molecules, only UREA exhibited heterogeneity (Q_Pval: 0.049), necessitating using a random-effects model for analyzing this particular protein (Table 2). Furthermore, all MR-PRESSO Global test P values exceeding 0.05 suggest the absence of horizontal pleiotropy. All leave-one-out analysis P values above 0.05 indicate that causal effects were not driven by any single instrumental variable (Table 2). The Steiger directionality test results confirmed the correctness of the assessment direction (Table 2).

Table 1:Details of the GWASs included in this study.

Table 2:Heterogeneity and sensitivity analysis.

Discussion

In this study, we employed Mendelian randomization to investigate the causal relationship between Helicobacter pylori (Hp) infection and immune thrombocytopenia. Our findings suggest no causal association between Hp and the risk of developing ITP. To ensure robustness, we conducted two independent validations using additional Hp GWAS datasets, all of which corroborated our initial results. Furthermore, our analysis did not reveal any causal relationship between six immunogenic proteins (GroEL, OMP, VacA, UREA, catalase, and CagA) of Hp and ITP.

Previous studies have reported conflicting results on this topic. For instance, a retrospective study by Dogan et al. in Turkey found no significant difference in the response to first-line ITP treatment between Hp-positive and Hp-negative patients (p=0.283) [5]. Conversely, recent research by Tabata et al. in Japan demonstrated that among ITP patients receiving platelet transfusion therapy, those who were Hp-positive exhibited significantly better treatment outcomes than Hp-negative patients (p=0.015) [6]. A prospective study by Hwang et al. in Korea indicated that Hp eradication was associated with higher platelet counts compared to non-eradicated cases with ITP, suggesting a potential link between Hp infection status and ITP treatment efficacy. These findings underscore the contentious nature between Hp infection and ITP [31].

It is important to note that previous studies have primarily relied on observational studies with limited sample sizes. Observational studies, by their nature, can establish associations but cannot definitively establish causation and may be subject to confounding factors [32]. In contrast, Mendelian randomization offers a robust framework for assessing causality by leveraging genetic variants as instrumental variables, thereby minimizing biases inherent in observational studies. Our study findings support no direct causal relationship between Hp and ITP. However, observational studies have suggested a relationship between Hp eradication and treatment outcomes in ITP, which may be influenced by several factors. Firstly, potential confounding factors in observational studies, such as uncontrolled environmental or lifestyle factors, could create spurious associations between Hp infection and ITP. For instance, Payandeh, et al. studied 41 ITP patients after controlling for confounders and found minimal clinical response following Hp eradication [33]. Secondly, reverse causality might also play a role in observational studies, where ITP patients could be more susceptible to Hp infection due to immune system dysregulation, exacerbated by steroid treatment, which lowers immune function and facilitates Hp colonization [34]. Mendelian randomization, utilizing genetic markers as instrumental variables, helps mitigate the influence of reverse causality. Lastly, the complexity of biological mechanisms could contribute to observed association in observational studies. Hp infection may affect the immune system through various mechanisms, potentially triggering ITP in specific individuals or conditions. These intricate biological pathways, detectable in observational studies, may not be identified in Mendelian randomization due to its stringent causal inference requirements. For example, Takahashi, et al. found cross-reactivity between Platelet-associated immunoglobulin (PAIgG) and Hp CagA protein, suggesting that molecular mimicry of Hp CagA may play a crucial role in the pathogenesis of chronic ITP by mimicking unknown platelet antigens [35].

Our research represents the first MR study worldwide exploring the causal relationship between Hp and ITP. This approach offers distinct advantages by minimizing confounding factors and potential reverse causation, providing robustness to our findings. Furthermore, using multiple Hp protein variants as instrumental variables enhance the comprehensiveness of our investigation. This conclusion holds true across various analyses, including examining causal links between six Hp proteins and ITP. Our results clarify the previous conflicting findings regarding the causal association between helicobacter pylori and ITP risk [5,6,36]. By employing MR, we contribute to the existing literature by providing rigorous evidence that challenges previous observational studies suggesting a causal link.

Several limitations warrant discussion. Firstly, the study samples were drawn exclusively from the European population, whose genetic structure and environmental background may differ from other ethnic groups, thus potentially limiting the generalizability of the study findings across diverse populations. Nevertheless, the inclusion of European samples is advantageous due to their robust genetic diversity and the availability of genome-wide association study data, which enhances the feasibility of conducting Mendelian randomization studies within this population. Additionally, insights into the biological mechanisms and genetic markers possibly linking Hp infection with ITP in European populations hold significant relevance for understanding this relationship more broadly. Therefore, it does not diminish the scientific contribution and broader applicability of this study in elucidating the association between Hp infection and ITP. Secondly, MR is a powerful methodological approach that utilizes genetic variants as instrumental variables to assess causal relationships. However, horizontal pleiotropy is a common issue in MR analysis. If present, horizontal pleiotropy violates the core assumption of MR analysis and may introduce bias into study results. In this study, the authors employed several methods, including the MR-Egger intercept test, funnel plot analysis, Leaveone- out analysis, and MR-PRESSO, to confirm the absence of horizontal pleiotropy. These rigorous analyses bolster the reliability of the study’s conclusions. Finally, the sample sizes of Hp-related proteins used as instrumental variables for exposure factors in this study are relatively small. However, Hp was the primary exposure factor under investigation, with Hp-related proteins mainly used for supplementary analysis. Additionally, although the sample sizes of Hp-related proteins are not very large, we could still obtain a sufficient number of SNPs, thus minimizing the impact on our results.

Conclusion

In conclusion, our study does not support a causal relationship between Hp infection and immune thrombocytopenia. Moving forward, advancements in genomic sequencing technologies and broader GWAS datasets will be instrumental in refining our understanding of the complex interplay between Hp infection and immune-related disorders like ITP. Continued research efforts, including multicenter studies with longitudinal designs, are essential for elucidating the true nature of these relationships and informing clinical practice.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

This study was supported by the 2024 Zhengzhou Municipal Science and Technology Innovation Guidance Program Project in the Medical and Health Field (Grant No.: 2024YLZDJH385) from the Zhengzhou Science and Technology Bureau.

Ethics

All summary statistics included in the GWASs and consortia used in this analysis had received approval from the relevant ethical review boards, and participants provided informed consent. This study represents a secondary analysis of publicly available data and does not require additional ethical approval.

Acknowledgments

Genetic instruments for Hp and associated proteins were obtained from IEU open GWAS. The authors are thankful to all the participants and investigators for sharing their data.

Data Availability Statement

Publicly available datasets were analyzed in this study. This data can be found at IEU Open GWAS (https://gwas.mrcieu.ac.uk/). The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

1. Rodeghiero F, Stasi R, Gernsheimer T, Michel M, Provan D, et al. (2009) Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. Blood 113(11): 2386-93.
2. Muhammad JS, Zaidi SF, Saeed SA, Ishaq M (2017) Current status of Helicobacter pylori association with haematological and cardiovascular diseases: A mini review. JPMA The Journal of the Pakistan Medical Association 67(6): 907-11.
3. Camilo V, Sugiyama T, Touati E (2017) Pathogenesis of Helicobacter pylori infection. Helicobacter 22 Suppl 1.
4. Sultan S, Irfan SM, Kaker J, Hasan M (2016) Efficacy of helicobacter pylori eradication as an upfront treatment of secondary immune thrombocytopenia: an experience from Pakistan. The Medical journal of Malaysia 71(2): 53-6.
5. Dogan A, Ekinci O, Ebinc S (2022) Effect of Helicobacter pylori infection on the first-line treatment outcomes in patients with immune thrombocytopenic purpura. European review for medical and pharmacological sciences 26(11): 3995-4000.
6. Tabata S, Hosoi H, Wan K, Sakaki A, Mushino T, et al. (2024) Comparison of platelet transfusion effectiveness between Helicobacter pylori-positive and -negative immune thrombocytopenia. British journal of haematology 204(5): 1953-7.
7. Suetsugu R, Sakuma H, Maruyama K, Matsuki M, Ogawa Y, et al. (2024) Successful treatment of coexisting membranous nephropathy and immune thrombocytopenia by eradicating gastric Helicobacter pylori infection: a case report. CEN case reports 13(2): 98-103.
8. Han B, Kim HJ, Yhim HY, Oh D, Bae SH, et al. (2022) Sequential eradication of Helicobacter pylori as a treatment for immune thrombocytopenia in patients with moderate thrombocytopenia: a multicenter prospective randomized phase 3 study. Annals of hematology 101(7): 1435-45.
9. Jin T, Huang W, Cao F, Yu X, Guo S, et al. (2022) Causal association between systemic lupus erythematosus and the risk of dementia: A Mendelian randomization study. Frontiers in immunology 13: 1063110.
10. Ou Q, Pan D (2023) Sarcopenia and risk of cardio-cerebrovascular disease: A two-sample Mendelian randomization study. Bioscience trends 17(5): 413-4.
11. Chen C, Wang P, Zhang RD, Fang Y, Jiang LQ, et al. (2022) Mendelian randomization as a tool to gain insights into the mosaic causes of autoimmune diseases. Autoimmunity reviews 21(12): 103210.
12. Qu Z, Yang F, Yan Y, Huang J, Zhao J, et al. (2021) A Mendelian randomization study on the role of serum parathyroid hormone and 25-hydroxyvitamin D in osteoarthritis. Osteoarthritis and cartilage 29(9): 1282-90.
13. Baurecht H, Freuer D, Welker C, Tsoi LC, Elder JT, et al. (2022) Relationship between periodontitis and psoriasis: A two-sample Mendelian randomization study. Journal of clinical periodontology 49(6): 573-9.
14. Skrivankova VW, Richmond RC, Woolf BAR, Yarmolinsky J, Davies NM, et al. (2021) Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement. Jama 326(16): 1614-21.
15. Sun M, Zhang Z, Zhang J, Zhang J, Jia Z, et al. (2024) Causal relationships of Helicobacter pylori and related gastrointestinal diseases on Type 2 diabetes: Univariable and Multivariable Mendelian randomization. PloS one 19(4): e0300835.
16. Nounu A, Kar SP, Relton CL, Richmond RC (2022) Sex steroid hormones and risk of breast cancer: a two-sample Mendelian randomization study. Breast cancer research: BCR 24(1): 66.
17. Lin S, Mao X, He W (2023) Causal association of circulating cytokines with sepsis: a Mendelian randomization study. Frontiers in immunology 14: 1281845.
18. Wootton RE, Lawn RB, Millard LAC, Davies NM, Taylor AE, et al. (2018) Evaluation of the causal effects between subjective wellbeing and cardiometabolic health: mendelian randomisation study. BMJ (Clinical research ed) 362: k3788.
19. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. (2018) The MR-Base platform supports systematic causal inference across the human phenome. Elife 7: e34408.
20. Sanderson E (2021) Multivariable Mendelian Randomization and Mediation. Cold Spring Harbor perspectives in medicine 11(2): a038984.
21. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG (2017) Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants. Epidemiology 28(1): 30-42.
22. Bowden J, Del Greco MF, Minelli C, Davey Smith G, Sheehan N, et al. (2017) A framework for the investigation of pleiotropy in two-sample summary data Mendelian randomization. Statistics in medicine 36(11): 1783-1802.
23. He Q, Bennett AN, Fan B, Han X, Liu J, Wu KCH, et al. (2022) Assessment of Bidirectional Relationships between Leisure Sedentary Behaviors and Neuropsychiatric Disorders: A Two-Sample Mendelian Randomization Study. Genes (Basel) 13(6): 962.
24. Wu F, Huang Y, Hu J, Shao Z (2020) Mendelian randomization study of inflammatory bowel disease and bone mineral density. BMC Med 18(1): 312.
25. Jiang Z, Mou Y, Wang H, Li L, Jin T, Wang H, et al. (2023) Causal effect between gut microbiota and pancreatic cancer: a two-sample Mendelian randomization study. BMC Cancer 23(1): 1091.
26. Ke B, Li C, Shang H (2023) Sex hormones and risk of epilepsy: A bidirectional Mendelian randomization study. Front Mol Neurosci 16: 1153907.
27. Qin P, Qin T, Liang L, Li X, Jiang B, Wang X, et al. (2024) The role of mitochondrial DNA copy number in cardiometabolic disease: a bidirectional two-sample mendelian randomization study. Cardiovasc Diabetol 23(1): 45.
28. Bowden J, Davey Smith G, Burgess S (2015) Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 44(2): 512-525.
29. Cortez Cardoso Penha R, Smith-Byrne K, Atkins JR, Haycock PC, Kar S, Codd V, et al. (2023) Common genetic variations in telomere length genes and lung cancer: a Mendelian randomisation study and its novel application in lung tumour transcriptome. ELife 12: e83118.
30. Huang J, Gui Y, Qin H, Xie Y. (2023) Causal association between adiposity and hemorrhoids: a Mendelian randomization study. Front Med (Lausanne) 10: 1229925.
31. Hwang JJ, Lee DH, Yoon H, Shin CM, Park YS, et al. (2016) The Effects of Helicobacter pylori Eradication Therapy for Chronic IdiopathicThrombocytopenic Purpura. Gut Liver 10(3): 356-361.
32. Bull CJ, Hazelwood E, Bell JA, Tan VY, Constantinescu AE, Borges MC, et al. (2023) Identifying metabolic features of colorectal cancer liability using Mendelian randomization. Elife 12: RP87894.
33. Payandeh M, Raeisi D, Sohrabi N, Zare ME, Kansestani AN, et al. (2013) Poor platelet Count Response to Helicobacter Pylori Eradication in Patients with Severe Idiopathic Thrombocytopenic Purpura. Int J Hematol Oncol Stem Cell Res 7(3): 9-14.
34. Yazdanbakhsh K, Zhong H, Bao W (2013) Immune dysregulation in immune thrombocytopenia. Semin Hematol 50 Suppl 1(01): S63-67.
35. Takahashi T, Yujiri T, Shinohara K, Inoue Y, Sato Y, Fujii Y, et al. (2004) Molecular mimicry by Helicobacter pylori CagA protein may be involved in the pathogenesis of H. pylori-associated chronic idiopathic thrombocytopenic purpura. Br J Haematol 124(1): 91-96.
36. Lee A, Hong J, Chung H, Koh Y, Cho SJ, Byun JM, et al. (2020) Helicobacter pylori eradication affects platelet count recovery in immune thrombocytopenia. Sci Rep 10(1): 9370.