Primary immune thrombocytopenia (ITP), an immune-mediated bleeding disorder marked by accelerated platelet destruction and impaired thrombopoiesis, manifests clinically as thrombocytopenia (platelet count <100 × 10⁹/L) with consequent bleeding susceptibility (1). Studies conducted in Western countries report an annual incidence of 2–10 per 100,000 individuals (2, 3). Complete platelet normalization remains an elusive clinical endpoint (4). Current therapeutic interventions prioritize achieving hemostatic platelet thresholds (>50 × 10⁹/L) to prevent severe bleeding, particularly below 30 × 10⁹/L (5, 6). First-line management employs glucocorticoids and intravenous immunoglobulin for rapid platelet elevation, whereas second-line strategies encompass thrombopoiesis-stimulating agents, rituximab, and splenectomy (6). However, there are still many challenges. Despite initial glucocorticoid efficacy in 70–80% of newly diagnosed patients, prolonged exposure heightens risks of metabolic complications (steroid-induced diabetes, osteoporosis), avascular necrosis, and thromboembolic events (7). Splenectomy, though achieving durable remission in ≈66% of cases (4), confers increased susceptibility to sepsis and venous thromboembolism (8). Recombinant human thrombopoietin (rh-TPO) demonstrates limited clinical utility due to suboptimal response rates and relapse propensity (9). Thrombopoietin receptor agonists (TPO-RAs: eltrombopag, avatrombopag, hetrombopag, romiplostim) emerge as promising alternatives, offering enhanced safety profiles and sustained efficacy compared to conventional therapies (4, 10), and TPO-RAs have also been found to be safe and effective in the treatment of ITP in pediatric patients (11). Furthermore, TPO-RAs have also demonstrated a positive impact on health-related quality of life (HRQoL) (12). By effectively reducing the risk of bleeding episodes and the dependency on rescue therapies, TPO-RAs therapy can significantly enhance patients’ physical function. Consequently, TPO-RAs is an essential consideration in the long-term management strategy for ITP.

Randomized controlled trials (RCTs) have validated the capacity of TPO-RAs to elevate platelet counts to hemostatic thresholds (≥50 × 10⁹/L), reduce hemorrhage incidence, and maintain favorable tolerability profiles within short-term therapeutic windows (≤6 months) (13, 14). However, the constrained observational frameworks of RCTs—typically limited to ≤6-month follow-up durations—restrict robust evaluation of longitudinal efficacy and safety outcomes. This methodological limitation underscores the complementary role of real-world studies (RWS) (i.e., non-interventional, observational studies such as prospective and retrospective cohort studies), which employ extended surveillance periods (>6 months) to assess sustained therapeutic performance (15, 16).

To bridge the gaps between the short-term efficacy and safety data from RCTs and the need for long-term evidence in clinical practice, this meta-analysis introduces a novel approach by comprehensively synthesizing data from both RCTs and RWS. Unlike previous meta-analyses that were exclusively based on short-term RCT data (typically ≤6 months), this research uniquely integrates long-term evidence derived from RWS. This methodology not only enhances the external validity of the findings but also provides pivotal insights into the sustained effectiveness and safety of TPO-RAs in the management of ITP, thereby generating more actionable and clinically relevant evidence to guide long-term therapeutic decision-making.

The chronic nature of ITP often necessitates long-term treatment with TPO-RAs, making the evaluation of their sustained therapeutic performance important (32). While RCTs establish robust short-term efficacy and safety profiles, lack of long-term follow-up data restricts their ability to inform clinical practice regarding the sustained efficacy and safety of TPO-RAs. To address this critical gap, our meta-analysis innovatively synthesizes RWE with RCT data, revealing two pivotal insights: (1) TPO-RAs demonstrate superior platelet response rates in both short-term (≤6 months) and more long-term treatment periods, and (2) prolonged administration correlates with increased rescue therapy requirement (12 to 32%) and SAE incidence (8 to 27%). These findings highlight the necessity for risk-adapted monitoring protocols that optimize the benefit-to-risk calculus during extended TPO-RA therapy, particularly in patients requiring indefinite thrombopoietic support.

Previous meta-analyses and systematic reviews on TPO-RAs in ITP have predominantly relied on RCT data (18, 57, 58), which are often limited by short follow-up durations and highly controlled patient populations. These limitations leave a critical knowledge gap regarding the long-term durability of platelet responses and the cumulative risk of adverse events, thereby restricting the generalizability of findings to real-world clinical settings. RWS provide valuable insights into the effectiveness and safety of TPO-RAs over extended periods. Unlike previous meta-analyses, this study incorporates both prospective and retrospective RWS to supplement the short-term outcomes reported in RCTs. This approach not only enhances the external validity of our findings but also provides critical insights into the long-term efficacy and safety of TPO-RAs. By bridging the gap between RCTs and RWS, our study offers a more nuanced understanding of TPO-RAs in ITP management, which is essential for guiding clinical decision-making.

This analysis reveals several important findings that have significant implications for clinical practice. First, TPO-RAs demonstrate a high overall platelet response rate in both short-term and long-term treatment periods. In RCTs, the short-term (≤6 months) response rate was 70%, which is consistent with the 76% response rate observed in RWS (38). Notably, the response rate increased to 85% at 6–12 months and reached 91% beyond 12 months. This suggests that TPO-RAs can rapidly induce a high platelet response rate and maintain efficacy over extended periods. However, the increasing need for rescue therapy (23% at 6–12 months and 32% beyond 12 months) highlight the challenges of sustaining long-term responses in some patients. These findings underscore the importance of individualized treatment strategies, regular monitoring, and timely adjustments to therapy to optimize long-term outcomes in ITP management (6).

Second, the meta-analysis confirms that TPO-RAs significantly reduce bleeding risk versus placebo, with lower rates of both any bleeding (OR = 0.43, 95%CI: 0.30–0.63) and significant bleeding (OR = 0.40, 95%CI: 0.26–0.61) during short-term therapy. This is a critical benefit, as bleeding complications are a major cause of morbidity and mortality in ITP patients (28, 58). However, heterogeneity definitions across observational studies precludes quantitative synthesis of real-world bleeding rates, underscoring the imperative for standardized bleeding criteria in longitudinal TPO-RA safety surveillance protocols.

Third, while TPO-RAs were well-tolerated in the short term, with no significant increase in any AEs or SAEs compared to placebo, the risk of SAEs potentially increased with prolonged treatment. The incidence of SAEs rose from 8% in RCTs to 19% at 6–12 months and 27% beyond 12 months in RWS. This trend may reflect the toxicity of long-term TPO-RAs use in real-world populations, emphasizing the importance of regular monitoring and individualized risk–benefit assessments when prescribing TPO-RAs for extended periods.

Beyond these clinical and laboratory endpoints, the impact of TPO-RAs on HRQoL is an important consideration for patients with chronic ITP. A qualitative synthesis of the included RCTs indicates that TPO-RA therapy is associated with meaningful improvements in quality of life. For instance, one study (NCT00102739) reported a significant improvement from baseline in emotional-role scores (p = 0.02) for patients receiving eltrombopag (23). Furthermore, another study (NCT00370331) demonstrated that improvements in HRQoL were significantly associated not only with eltrombopag-mediated increases in platelet counts (p = 0.034), but also with decreases in WHO bleeding grades (p = 0.002) (12). These findings underscore that the benefits of TPO-RAs extend beyond platelet elevation to encompass enhanced overall well-being and daily functioning, which are paramount from the patient’s perspective.

Looking to the future, novel therapeutic modalities such as CAR-T cell therapy have advanced to clinical trials for application in selected patients with treatment-refractory ITP and hold the promise of achieving complete remission or even a cure for ITP (59). Nevertheless, TPO-RAs continue to play an essential role in ITP management due to their well-established efficacy, manageable safety profile, and cost-effectiveness. Consequently, the long-term effectiveness and safety data synthesized in this meta-analysis, which are uniquely derived from RWS, provides critical evidence to inform the sustained and secure use of TPO-RAs in clinical practice.

However, this study has some limitations. First, the heterogeneity in study designs, patient populations, and outcome definitions across RWS posed challenges for data synthesis. Although we used random-effects models to account for this heterogeneity, it may still have influenced our pooled estimates. Second, the lack of long-term data from RCTs limited our ability to directly compare short-term and long-term outcomes. Third, the retrospective nature of some RWS may have introduced bias, particularly in the reporting of adverse event and bleeding events. Fourth, it was not feasible to perform a valid pooled analysis of bleeding events in RWS due to the highly heterogeneous definitions and grading scales employed across the included studies.

Fifth, although a quantitative meta-analysis of HRQoL outcomes was not conducted, we have qualitatively summarized the reported improvements in HRQoL in the Discussion section. Finally, the potential competing risk of death was not addressed in the statistical analysis. However, given that the overall mortality in the included studies was low (see Supplementary Table S6 for details), we believe its impact on the pooled results is likely minimal. Future research should address these limitations by conducting long-term RCTs, standardizing outcome report in RWS.

In conclusion, this meta-analysis establishes that TPO-RAs are highly effective in achieving and maintaining platelet responses in ITP patients, with significant reductions in bleeding risk and a favorable short-term safety profile. However, the increasing need for rescue therapy and risk of SAEs with prolonged treatment underscores the importance of careful monitoring and individualized treatment strategies. Future research could aim to identify factors influencing durable response and late-emerging adverse events to optimize long-term, personalized patient management.