Background
Pharmacovigilance study using the FDA FAERS database (2014-2024) analyzing hematologic and thrombotic adverse events associated with immune checkpoint inhibitors (ICIs); comprehensive toxicity characterization across ICI regimens.
Interventions and follow up
Design: Pharmacovigilance disproportionality analysis of FAERS (2014-2024) characterizing hematologic/thrombotic AEs associated with ICIs
Primary endpoint: Reporting odds ratios and frequencies of hematologic/thrombotic AEs by ICI regimen
mFollow up: Real-world reporting (no prospective follow-up)
Primary endpoint: Reporting odds ratios and frequencies of hematologic/thrombotic AEs by ICI regimen
mFollow up: Real-world reporting (no prospective follow-up)
Results
Total reports: 14,753 ICI-hematologic AE reports
Anemia: 2,963 cases (16.62%)
Thrombocytopenia: 1,961 (11.00%)
Febrile neutropenia: 1,935 (10.85%)
Neutropenia: 1,392 (7.81%)
Myelosuppression: 979 (5.49%)
Pancytopenia: 739 (4.14%)
Time to onset: median 28 days (IQR 11-79); 78.22% within 3 months
Highest mortality proportion: DIC and DVT among top 10 AEs
Sex: men 50.90% vs women 40.11%
Regimen: combination (ICI+ICI, ICI+chemotherapy) show stronger signals than monotherapy
Anemia: 2,963 cases (16.62%)
Thrombocytopenia: 1,961 (11.00%)
Febrile neutropenia: 1,935 (10.85%)
Neutropenia: 1,392 (7.81%)
Myelosuppression: 979 (5.49%)
Pancytopenia: 739 (4.14%)
Time to onset: median 28 days (IQR 11-79); 78.22% within 3 months
Highest mortality proportion: DIC and DVT among top 10 AEs
Sex: men 50.90% vs women 40.11%
Regimen: combination (ICI+ICI, ICI+chemotherapy) show stronger signals than monotherapy
Adverse events
Most life-threatening: thrombocytopenia and febrile neutropenia
Highest mortality: DIC and DVT
Course: early onset and frequent hospitalizations expected
Highest mortality: DIC and DVT
Course: early onset and frequent hospitalizations expected
Conclusions
Hematologic and thrombotic AEs associated with ICI therapy show notable reporting frequencies, with combination regimens demonstrating stronger signals. Early onset and notable mortality proportions underscore the importance of clinical vigilance and proactive monitoring in ICI-treated patients.
Key Limitations
Spontaneous-reporting (FAERS) database: reporting/notoriety bias, no denominator so true incidence cannot be calculated, confounding by indication and concomitant therapy, unverified causality, and incomplete data; disproportionality signals are hypothesis-generating only.
Clinical Context
Characterizes real-world hematologic and thrombotic toxicity of ICIs to support monitoring. Findings reinforce ASCO/ESMO irAE guidance to monitor blood counts during ICI therapy, with heightened vigilance for combination regimens; signals require confirmation in prospective cohorts before changing practice.