Tenosynovial giant cell tumor (TGCT), historically termed pigmented villonodular synovitis (PVNS) or giant cell tumor of the tendon sheath, is a rare mesenchymal neoplasm originating from synovial membranes of joints and the tendon sheaths (1). For patients with symptomatic TGCT refractory to local therapies or causing significant functional impairment, systemic therapy should be considered if surgery is unlikely to achieve functional outcome improvement. The molecular pathogenesis of TGCT involves genomic alterations at the colony-stimulating factor 1 (CSF1) locus (1p13), resulting in aberrant CSF1 overexpression in a subset of tumor cells (2). This molecular pathology supports CSF1/CSF1 receptor (CSF1R) axis inhibition as a central therapeutic strategy.

Pexidartinib, an oral selective CSF1R inhibitor, represents the only approved systemic treatment for TGCT (3). Its Food and Drug Administration (FDA) approval in August 2019 introduced it as the first systemic treatment for adult patients with symptomatic TGCT associated with severe morbidity or functional limitations and not amenable to improvement with surgery (4, 5). Pre-marketing clinical trials have since validated pexidartinib’s efficacy and safety, reinforcing its pivotal role in TGCT management (6, 7).

Comprehensive monitoring of pexidartinib’s real-world safety profile remains imperative given its expanding clinical utilization following FDA approval. Drug safety assessments advocate intervention strategies, including dose control, to mitigate toxicity (8). According to FDA’s prescribing information, pexidartinib’s common adverse events (AEs) included increased lactate dehydrogenase, increased aspartate aminotransferase, hair color changes, fatigue, increased alanine aminotransferase, decreased neutrophils, increased cholesterol, increased alkaline phosphatase, decreased lymphocytes, periorbital edema, decreased hemoglobin, rash, dysgeusia, and decreased phosphate. However, patients may experience AEs during off-label use due to comorbidities, concomitant medications, or genetic predispositions. Given the limited sample sizes and short observation periods in clinical trials, extensive post-marketing safety research in real-world settings is essential. Nevertheless, comprehensive real-world pharmacovigilance studies specifically addressing its AEs remain notably lacking.

The FDA Adverse Event Reporting System (FAERS) serves as a national post-marketing surveillance tool and one of the world’s largest pharmacovigilance databases, designed to systematically capture spontaneous safety reports for therapeutic agents (9, 10). Functioning as a critical spontaneous reporting system, this continuously updated, quarterly database (11) captures diverse safety signals, including clinician-reported AEs, medication error documentation, and product quality complaints. For the past few years, many drug safety profile studies based on the FAERS database have been published, affirming the reliability of this resource (12, 13). Using disproportionality analysis based on the FAERS database, we identified post-marketing safety signals associated with pexidartinib. This study detected AEs not previously documented in FDA-approved prescribing information. Importantly, these findings provide critical insights for optimizing therapeutic monitoring and enhancing pharmacovigilance strategies in clinical practice.

Our pharmacovigilance study provides the first systematic characterization of pexidartinib’s post-marketing safety profile using real-world evidence from the FAERS database. Analysis of AEs’ temporal trends revealed a progressive increase in reports over time, likely attributable to expanded clinical use and enhanced pharmacovigilance awareness following FDA approval. We identified a higher frequency of AE reports among adult women in FAERS, which may reflect the known higher prevalence of TGCT in women. TGCT is more common in women than men, and the mean age at diagnosis is 35–50 years (1, 16). However, due to the absence of an accurate number of patients using pexidartinib, prospective population-based studies with standardized AE ascertainment remain necessary to establish robust risk stratification models.

The results of our study underscore a clustering of common SOCs around “general disorders and administration site conditions”, “injury, poisoning, and procedural complications”, “skin and subcutaneous tissue disorders”, and investigations, as well as gastrointestinal disorders. Additionally, frequently reported PTs associated with pexidartinib include fatigue, hair color changes, product dose omission issues, nausea, pruritus, increased aspartate aminotransferase, and increased alanine aminotransferase. Notably, these AEs are in accordance with information provided in the FDA’s drug label and previous research on pexidartinib. For example, a randomized phase 3 clinical trial of pexidartinib versus placebo for advanced TGCT identified that hair color changes (67%), fatigue (54%), increased aspartate aminotransferase (39%), nausea (38%), increased alanine aminotransferase (28%), and dysgeusia (25%) were the most frequent pexidartinib-associated AEs (6). A study of long-term outcomes of pexidartinib in TGCT revealed that “hair color changes” was the most frequent AE (17). A. Vaynrub et al. concluded that the most common AEs associated with pexidartinib are mild hypopigmentation of the hair and transient aminotransferase elevation (18).

Beyond confirming established safety signals, our pharmacovigilance analysis identified 16 novel adverse drug reactions not currently documented in pexidartinib’s FDA labeling. These include photosensitivity reaction, dysmenorrhea, oligomenorrhea, increased gamma-glutamyltransferase, decreased blood iron, increased blood calcium, prolonged prothrombin time, increased red cell distribution width, hepatitis A, seasonal allergy, vanishing bile duct syndrome, cholecystitis, hunger, soft feces, eye color change, and blepharospasm. Sorbarikor Piawah et al. (19) reported a case of severe drug-induced liver injury requiring liver transplantation due to vanishing bile duct syndrome after exposure to pexidartinib. Beyond vanishing bile duct syndrome previously reported in the literature, most AEs associated with pexidartinib identified in our study represent novel safety findings. These newly detected AEs warrant further validation to ensure the safe clinical use of pexidartinib. Importantly, the mechanisms underlying most unexpected AEs remain unstudied, necessitating dedicated mechanistic investigations.

Occurrences such as therapy cessation, therapy change, tumor excision, product dose omission issue, product dose omission in error, wrong dose, no adverse event, and treatment non-compliance were not classified as pexidartinib-induced AEs. Pexidartinib was often used as therapy for TGCT preoperation and postoperation (20, 21), and a few patients have reported drug dosage reduction from the initial dose (20). Therefore, we posit that treatment discontinuation, therapy modification, and tumor resection are integral components of managing the primary disease. Additionally, intentional dose omissions and unintentional dosing errors were predominantly associated with supply chain disruptions. Recognizing these distinctions is essential for holistic patient assessment and optimized therapeutic management.

Gamma-glutamyltransferase is one of the key enzymes involved in the transformation function of hepatocytes, and it is also an important link in glutathione metabolism (22). Therefore, we postulate that elevated gamma-glutamyltransferase levels may reflect pexidartinib-induced hepatotoxicity. Notably, decreased hemoglobin is a documented adverse reaction in the drug’s prescribing information. Significantly, we identified reduced serum iron and increased red cell distribution width as previously unreported AEs potentially causally linked to hemoglobin reduction. Increased blood calcium, an off-label AE of pexidartinib, may be associated with decreased phosphate mentioned in pexidartinib’s prescribing information (23, 24). The relationship between these clinical manifestations should attract the attention of patients and physicians, and the mechanism warrants further research.

Our analysis identified dysmenorrhea as a previously unreported adverse drug reaction associated with pexidartinib. This finding holds particular significance given the established association between anticoagulant-induced platelet dysfunction and menstrual abnormalities. A multicenter, single-arm, open-label, phase 2a, proof-of-concept trial (25) revealed that menorrhagia and dysmenorrhea are adverse drug reactions to rivaroxaban. Moreover, prolonged prothrombin time is detected as a kind of AE of pexidartinib. In a number of studies, prolonged prothrombin time is a common adverse reaction to tigecycline (26, 27). Therefore, the coagulation dysfunction induced by pexidartinib is worthy of attention and needs further exploration.

While leveraging that the FAERS database enables large-scale pharmacoepidemiologic surveillance, our study design inherits limitations inherent to spontaneous reporting systems that warrant cautious interpretation. First, because of the voluntary nature of the FAERS database, it has some inherent selection bias, such as the ethnicity and geography of the reported cases, the timing of approval and market penetration of different drugs, the level of public awareness of specific adverse reactions, and the fact that not all reports of serious adverse reactions occurring are being collected (28). Second, polypharmacy, comorbidities, and underlying disease severity pose a challenge in addressing confounding factors. The absence of comprehensive clinical information, such as interventions following AEs and the health status of the reporting patient, prevents the impact of confounding factors on the determination of causality between AEs and pexidartinib from being mitigated (29). Third, due to the absence of an accurate number of patients using pexidartinib, it remains impossible to calculate the true incidence rates for each AE (30).

Notwithstanding these limitations, the FAERS database remains a cornerstone resource in global pharmacovigilance, providing critical post-marketing surveillance insights through its unparalleled scale (30). It is crucial to acknowledge that although data mining techniques cannot compensate for the inherent limitations of a spontaneous reporting system, the combined utilization of the large-scale database and case reports remains an effective approach for delving into adverse drug reactions (31). The insights gleaned from the FAERS database provide valuable preliminary information for further investigation and prospective studies. Although the findings should be interpreted with caution, they represent contributions to a broader understanding of the safety profile of pexidartinib and its potential implications in clinical practice.

Through a comprehensive and systematic pharmacoepidemiologic study based on the FAERS database, we characterized the post-marketing safety profile linked to pexidartinib. Most of the AEs we identified closely align with the information provided in the FDA’s official prescribing guidelines. Notably, our study unveiled 16 unexpected AEs, expanding upon the existing knowledge derived from pre-marketing clinical trials. Although the limitations of the FAERS database are difficult to overcome, these findings are particularly valuable for ensuring drug safety. Further prospective clinical trials are warranted to establish a definitive connection between pexidartinib and these newly identified AEs. This study demonstrates the utility of pharmacovigilance databases in complementing clinical trial data for comprehensive drug safety assessment.