Central nervous system hemorrhage related to protein kinase inhibitors: A real-world pharmacovigilance case-control study and genetic correlation study

Despite the widespread clinical use of protein kinase inhibitors in oncology, evidence regarding their potential to induce central nervous system (CNS) hemorrhage remains limited. This study aims to investigate the occurrence and risk signals of CNS hemorrhagic events associated with different PKI subclasses, identify the most affected patient populations, and examine variations in risk profiles across demographic subgroups. We conducted disproportionality analyses using data from the FDA Adverse Event Reporting System to detect CNS hemorrhage signals linked to 14 PKI classes. Subgroup analyses stratified by age, sex, and body weight were performed to explore differential risk patterns. Furthermore, Mendelian Randomization and genetic correlation analyses were utilized to evaluate potential causal relationships between PKI therapeutic indications and CNS hemorrhages. We analyzed 9215 CNS hemorrhage reports linked to 14 PKI classes. We identified 6 CNS hemorrhage related signal of disproportionate reporting significantly associated with PKI use. Both BCR-ABL tyrosine kinase inhibitors and BTK inhibitors showed consistent positive signals across all 6 signal of disproportionate reportings, with BTK inhibitors presenting analysis the highest risks (e.g., cerebral hemorrhage reporting odds ratio = 3.30 [3.00-3.62], subdural hematoma reporting odds ratio = 5.90 [5.28-6.60]). Mendelian Randomization revealed no causal genetic relationship between PKI indications and CNS hemorrhage, suggesting the hemorrhagic risks are likely attributable to the drugs themselves rather than underlying diseases. Significant disproportionality signals of PKI-related CNS hemorrhage were identified, while causal links between indications and adverse events were excluded. These findings enhance understanding of PKI-associated bleeding risks and support early detection, targeted monitoring, and informed clinical decision-making.