Comment on “Safety profile and potential clinical risks of xanomeline and trospium chloride” - Confirmatory findings from an independent FAERS analysis

We read with interest the study by Bao et al. describing the early postmarketing safety profile of xanomeline and trospium chloride (X/T) using the FDA Adverse Event Reporting System (FAERS) [1]. Their analysis provides timely, real-world pharmacovigilance data regarding the adverse event (AE) profile of this novel, first-in-class therapy, and complements recent reports [2,3]. Since pharmacovigilance data are sensitive to preprocessing decisions, case definitions, and signal-detection methods, analyses conducted using independently developed methods are important for evaluating the consistency and reproducibility of newly identified reporting signals [4,5].

Before publication of Bao et al., we independently conducted a disproportionality analysis of X/T reports from 2024 Q4 through 2026 Q1. Although our dataset substantially overlaps with that of Bao et al. (2024 Q3–2025 Q4), our analysis extends the reporting period by one quarter. After deduplication by case number and retention of the most recent version of each case, 1758 reports listed X/T as the primary suspect drug. We calculated the proportional reporting ratio (PRR) with χ², reporting odds ratio (ROR) with 95% confidence intervals (CI), and Bayesian metrics (EBGM with EBGM05, and information component with IC025). We evaluated signals using five prespecified criteria: PRR ≥2, χ² ≥ 4, at least three cases, ROR lower 95% CI > 1, and IC025 > 0). Reporting followed READUS-PV recommendations.

Forty-five of the 100 most frequently reported preferred terms (PTs) met all five criteria. Our findings closely corroborated the principal disproportionality signals reported by Bao et al. [1] The strongest signals occurred in gastrointestinal (GI), psychiatric, nervous system, and urinary disorders. In both Bayesian analyses, nausea, vomiting, constipation, urinary retention, dyspepsia, blurred vision, hyperhidrosis, dry mouth, somnolence, gastroesophageal reflux disease (GERD), and drooling were among the most frequently reported positive signals. Nausea, vomiting, dyspepsia, hyperhidrosis, dry mouth, somnolence, and drooling are consistent with the cholinergic effects of xanomeline, whereas constipation, urinary retention, blurred vision, and GERD are consistent with the peripheral antimuscarinic effects of trospium. The reproducibility of these reporting patterns, together with their pharmacologic plausibility and consistency with clinical trial findings, support the need for continued surveillance of these AEs as the use of X/T expands.

A key methodological distinction is important when quantitatively comparing these two analyses. Bao et al. calculated disproportionality using PT records as the unit of analysis (2785 for the drug; 7,111,973 in the background), whereas our primary analysis used case reports (1758 and 2,154,103, respectively). Because X/T reports include approximately 2.0 PTs per case compared to approximately 3.3 PTs in the background, the two conventional analytical approaches are based on different units of analysis and should be interpreted differently. Consequently, our case-based ROR estimates were systematically lower than the estimates calculated using PT records (by a factor of roughly 1.5). For example, the primary case-based ROR for urinary retention was 42.4 (95% CI, 34.9–51.5), compared with 66.76 (95% CI, 55.14–80.82) using the PT-based approach. This difference illustrates how the choice of analytic unit can influence the magnitude of disproportionality.

To facilitate direct comparison, we recalculated our RORs using the PT-based convention employed by Bao et al. [1]. As shown in Table 1, the two independent analyses demonstrated close quantitative agreement across all ten selected terms. RORs in our analysis versus Bao et al. were 66.76 versus 66.89 for urinary retention, 96.47 versus 98.36 for drooling, and 12.70 versus 12.98 for vomiting, with no term differing by more than approximately 10%. Thus, differences in the magnitude of disproportionality were resolved after harmonizing the unit of analysis, further supporting the reproducibility of these reports.

Table 1.

Comparison of selected xanomeline–trospium disproportionality signals reported by Bao et al. and our extended FAERS analysis (2024 Q4–2026 Q1).

Preferred term	Bao et al., cases (n)	Bao et al., ROR (95% CI)	Extended analysis, cases (n)	Extended analysis, ROR (95% CI)
Nausea	313	11.19 (9.95–12.59)	365	10.09 (9.05–11.25)
Vomiting	239	12.98 (11.37–14.83)	297	12.70 (11.27–14.31)
Constipation	106	9.86 (8.12–11.98)	134	10.00 (8.41–11.89)
Urinary retention	88	66.89 (53.96–82.93)	112	66.76 (55.14–80.82)
Dyspepsia	51	11.75 (8.91–15.52)	65	11.86 (9.28–15.17)
Vision blurred	50	9.06 (6.85–11.99)	68	9.89 (7.78–12.58)
Hyperhidrosis	46	10.56 (7.89–14.14)	62	11.72 (9.11–15.07)
Dry mouth	35	9.86 (7.06–13.77)	45	10.30 (7.67–13.83)
Gastrooesophageal reflux disease	32	9.17 (6.46–13.00)	42	9.34 (6.88–12.67)
Drooling	30	98.36 (68.17–141.90)	38	96.47 (69.55–133.80)

RORs for both analyses are calculated using PT records as the unit of analysis. ROR = reporting odds ratio; CI = confidence interval; PT = preferred term.

Additionally, our extended analysis also detected ten statistically significant PT reporting signals not currently included in the FDA Prescribing Information, including suicidal ideation (ROR 2.8; 95% CI, 1.8–4.4), mania (ROR 14.8; 95% CI, 8.9–24.6), anger (ROR 8.3; 95% CI, 5.0–13.9), aggression (ROR 3.7; 95% CI, 2.1–6.7), paranoia (ROR 11.8; 95% CI, 6.8–20.4), dysphagia (ROR 2.2; 95% CI, 1.3–3.5), retching (ROR 3.3; 95% CI, 1.5–7.4), tremor (ROR 2.4; 95% CI, 1.7–3.6), cold sweat (ROR 6.4; 95% CI, 3.2–12.9), and increased appetite (ROR 3.7; 95% CI, 1.8–7.4) [6]. Interpretation of the psychiatric signals warrants particular caution as several of these events may overlap with manifestations of the underlying disease, increasing the potential for confounding by indication.

Spontaneous reporting databases lack reliable exposure denominators, and are subject to underreporting, missing clinical detail, concomitant therapies, and confounding by indication. Accordingly, these findings represent statistical reporting associations rather than confirmed AEs, and, therefore, should be considered hypothesis-generating. These statistical associations do not establish incidence, comparative risk, or causality and should be considered hypothesis generating [4,5]. Further evaluation in controlled and real-world studies is warranted.

Nevertheless, our independently conducted and temporally extended analysis substantially corroborates the principal findings of Bao et al. and extends the known safety profile of X/T demonstrated in controlled, clinical trials [1]. These findings support the predominant GI and urinary reporting patterns, strong signals for drooling and urinary retention, and identification of several unlabeled PTs [[7], [8], [9]]. The additional, unlabeled signals identified in our analysis warrant continued surveillance and require evaluation in data sources better suited to address confounding. Further analyses using non-overlapping FAERS periods, additional pharmacovigilance databases, and exposure-based observational studies will be necessary to determine the clinical significance of these reporting signals.

Author contributions

Leppien, Rola, Guliani, and McCall contributed to the study conception and methodology. McCall, Daly, Gandhi, and Patel performed data curation and formal analysis. Leppien, McCall, Daly, Gandhi, and Patel prepared the original draft. Leppien, McCall, Rola, and Guliani contributed to supervision, interpretation of the findings, review and editing of the manuscript, and project administration. All authors reviewed and approved the final manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.