**Background:** Gastrointestinal stromal tumors (GISTs) are rare mesenchymal neoplasms with an annual incidence of 0.4–2 per 100,000 people. Most cases are driven by activating mutations in KIT (75–80%) or PDGFRA (10–20%) receptor tyrosine kinase genes. The introduction of tyrosine kinase inhibitors (TKIs) revolutionized GIST treatment, starting with imatinib in 2001, followed by sunitinib (2006), regorafenib (2013), avapritinib (2020), and ripretinib (2021). While TKIs differ in their safety profiles, neuropsychiatric adverse drug reactions (ADRs) remain poorly characterized in postmarketing settings, particularly for newer agents. Premarketing studies suggested a possible correlation between avapritinib and cognitive impairment, and the EMA risk management plan lists cognitive effects as an important identified risk for avapritinib. This study aimed to evaluate the reporting frequency of neuropsychiatric ADRs for TKIs approved for GISTs using the European spontaneous reporting system database.
**Methods:** This retrospective, observational pharmacovigilance study analyzed all individual case safety reports (ICSRs) collected in the EudraVigilance database from each drug's approval date up to 31 December 2021. Included TKIs were imatinib (from 1 June 2005), sunitinib (from 1 October 2006), regorafenib (from 1 January 2013), avapritinib (from 1 October 2019), and ripretinib (from 1 December 2020). Premarketing ICSRs with supporting literature, ICSRs for other indications, and ICSRs with more than one suspected drug were excluded. Neuropsychiatric ADRs were identified using MedDRA System Organ Classes "Nervous system disorders" and "Psychiatric disorders." Disproportionality analysis using reporting odds ratio (ROR) with 95% confidence intervals (CI) was performed to assess the frequency of reporting for each TKI compared to all other TKIs. Time to onset (TTO) was calculated for ICSRs reporting both therapy duration and drug discontinuation.
**Key Results:** Of 72,720 TKI-related ICSRs, 8,512 met inclusion criteria: imatinib (n=4,931; 57.9%), sunitinib (n=2,062; 24.2%), avapritinib (n=1,112; 13.1%), regorafenib (n=230; 2.7%), and ripretinib (n=177; 2.1%). Most ICSRs involved male patients (51.7%) and adults aged 18–64 years (44.7%), though avapritinib and ripretinib had higher percentages of elderly patients (55.8% and 50.0%, respectively). Healthcare professionals submitted 67.2% of reports, and 87.5% of all ADRs were serious. Fatal outcomes occurred in 25.5% of total ICSRs, highest for sunitinib (50.8%).
Neuropsychiatric ADRs were reported in 1,511 ICSRs (17.8%), totaling 2,162 ADRs. Notably, 50.1% of avapritinib-related reports (n=557) had at least one neuropsychiatric ADR. Neuropsychiatric ICSRs were mainly in elderly patients (52.7%) and were serious in 72.6% of cases, though avapritinib had a lower proportion of serious neuropsychiatric ADRs (34.3%). Nearly half of neuropsychiatric ADRs were not recovered/not resolved (49.2%), rising to 76.8% for avapritinib. The median TTO was 91 days for avapritinib and 88 days for sunitinib.
Avapritinib showed a significantly higher reporting probability for both Psychiatric disorders (ROR 6.36; 95% CI 5.15–7.84) and Nervous system disorders (ROR 4.63; 95% CI 3.99–5.37). Most neuropsychiatric ADRs with higher reporting probability for avapritinib were known from its SmPC, except for lumbar spinal cord and nerve root disorders/sciatica (ROR 4.46; 95% CI 1.58–12.54), olfactory nerve disorders including anosmia and parosmia (ROR 8.02; 95% CI 2.44–26.33), and hallucinations (excluding sleep-related) (ROR 22.96; 95% CI 8.45–62.36). For regorafenib, central nervous system hemorrhages and cerebrovascular accidents, especially cerebral infarction, were unknown signals (ROR 3.19; 95% CI 1.59–6.38). Ripretinib showed higher reporting of disturbances in consciousness, including somnolence, not previously reported (ROR 3.15; 95% CI 1.52–6.55).
**Clinical Implications:** This is the first study evaluating TKI-related neuropsychiatric ADRs from the European spontaneous reporting database. The findings confirm that avapritinib carries a substantially higher risk of neuropsychiatric events compared to other GIST TKIs, likely due to its type I inhibitor mechanism and blood-brain barrier penetration, possibly involving inhibition of type 2 sodium channels. The identification of previously unreported signals—sciatica, olfactory disorders, and hallucinations with avapritinib; cerebrovascular accidents with regorafenib; and somnolence with ripretinib—warrants clinical attention. Given that 76.8% of avapritinib-related neuropsychiatric ADRs were unresolved at the time of reporting, and the median time to improvement is approximately 16 weeks, close monitoring and patient education are essential. The study is limited by spontaneous reporting system biases including underreporting, lack of denominator data, and potential confounding by disease progression. Further real-world studies and multidisciplinary collaboration among pharmacologists, neurologists, psychiatrists, and oncologists are needed to better characterize and manage these neuropsychiatric effects in GIST patients.