Immune Checkpoint Inhibitor-Associated Autoimmune Hemolytic Anemia and Immune Thrombocytopenia: Cross-Database Consistency in FAERS and JADER

Background/Objectives: Autoimmune hemolytic anemia (AIHA) and immune thrombocytopenia (ITP) are uncommon hematologic immune-related adverse events of immune checkpoint inhibitors (ICIs). We evaluated whether strict reporting signals were consistent across two national spontaneous-reporting systems and whether this consistency persisted after reviewer-requested harmonization analyses. Methods: We performed a cross-database disproportionality study using cumulative FAERS data through 2025 Q3 and JADER 202608. MedDRA 29.0 preferred-term phenotype definitions and analytic rules were defined before final analysis. Primary analyses used primary/secondary-suspect ICI reports in FAERS and suspected-drug reports in JADER, with oncology-restricted, role-harmonized, common-calendar, interaction, and time-to-onset sensitivity analyses. Results: FAERS contained 207,459 ICI-exposed reports, including 338 AIHA reports (ROR 8.13, 95% CI 7.26–9.09) and 602 ITP reports (ROR 6.38, 95% CI 5.87–6.93). JADER contained 86,270 ICI-exposed reports, including 128 AIHA reports (ROR 4.27, 95% CI 3.48–5.24) and 410 ITP reports (ROR 3.56, 95% CI 3.18–3.98). Among 20 adequately supported shared comparisons, all were directionally concordant and 19 were positive in both databases. Role-harmonized analysis yielded 18/18 positive comparisons and a common 2025 Q3 cutoff yielded 19/19. Exact-day time to onset was available for 43.8% of JADER AIHA and 51.0% of ITP reports and was sensitive to drug-event interval selection, particularly for AIHA. Conclusions: Strict ICI-associated AIHA and ITP definitions show robust cross-database directional consistency. These data quantify reporting disproportionality, rather than incidence or clinical risk; cancer-context and temporal findings remain exploratory.