Vaginal yeast infection: an imidazole FAERS study

Zhou and colleagues reported statistical signals in 9,985 selected FAERS adverse-event reports naming four imidazoles. The reports are not patients, and the analysis cannot estimate a treated person's chance of harm or establish an effect specific to a vaginal VVC treatment route.

The paper adds report-based safety-signal context for four imidazole drugs. Its signals are hypothesis-generating and must be read as patterns in a spontaneous-report database, not as clinical event rates or treatment comparisons.

This was a retrospective pharmacovigilance disproportionality analysis of female FAERS reports from 2004 Q1 through 2022 Q3. The analysis initially included 23,169 reports and retained 9,985 positive-signal adverse-event reports: 2,256 naming clotrimazole, 490 econazole, 1,901 ketoconazole, and 5,338 miconazole. These are counts of reports, not a cohort of people. Age was missing or abnormal for 3,491/9,985 reports (34.96%). The paper's prose and flow diagram do not clearly reconcile every exclusion count with the final report total.

The authors queried OpenVigil 2.1 for reports naming clotrimazole, econazole, ketoconazole, or miconazole, restricted the query to female reports, and used MedDRA 25.1 event terms. They used reporting odds ratio (ROR), proportional reporting ratio (PRR), and Bayesian Confidence Propagation Neural Network signal-detection methods. The comparator was other reports in the database; there was no assigned treatment, untreated group, or verified patient-level drug exposure.

Table 1 marked 3,187/9,985 reports (31.92%) as severe. It listed 546 reports with a death outcome, 262 with disability, 1,852 with hospitalization, and 527 with a life-threatening outcome; these report categories may overlap. Serious adverse-event reports were 1,407/2,256 (62%) for clotrimazole, 365/490 (74%) for econazole, 675/1,901 (35%) for ketoconazole, and 740/5,338 (13%) for miconazole. These report denominators are not drug-exposed patient groups. The paper placed kidney/urinary and pregnancy System Organ Class findings among the top ten for all four drugs.

Selected pregnancy/perinatal Table 2 signals included clotrimazole fetal death (5 reports in the reported age band 19–39; ROR 10.28, 95% CI 4.24–24.92), econazole spontaneous abortion (9 reports, age band 19–39; ROR 19.62, 95% CI 9.31–41.34), ketoconazole premature rupture of membranes (8 reports, age band 19–39; ROR 22.91, 95% CI 11.3–46.45), miconazole neonatal sepsis (7 reports, age band 0–18; ROR 123.71, 95% CI 55.7–274.76), and miconazole spontaneous abortion (4 reports, age band 40–59; ROR 5.98, 95% CI 2.24–16.02).

Selected renal/urinary signals included clotrimazole acute kidney injury (19 reports, age band 40–59; ROR 10.45, 95% CI 6.59–16.57) and chronic kidney disease (22 reports, age band 40–59; ROR 19.96, 95% CI 12.97–30.71); econazole acute kidney injury (6 reports, age band 60 or older; ROR 4.41, 95% CI 1.95–9.97); miconazole hematuria (18 reports, age band 60 or older; ROR 19.03, 95% CI 11.88–30.48); and ketoconazole chronic kidney disease (14 reports, age band 40–59; ROR 11.92, 95% CI 6.99–20.32). The RORs describe disproportionality within FAERS reports. They are not patient risks, incidence, comparative safety, causal effects, or VVC-route-specific effects.

These findings describe spontaneous reports naming four imidazoles. The study did not verify a VVC diagnosis, the indication or route for each report, or a patient-level exposure. It does not estimate an individual's likelihood of harm and does not address oral fluconazole, pregnancy-treatment advice, renal monitoring, or a person's treatment choice.

FAERS is a passive reporting system with no denominator of people exposed to each drug. Individual dose, duration, route, indication, timing, adherence, and concomitant medicines were not established. The data can contain underreporting, duplicate or incomplete reports, confounding by underlying conditions or other medicines, and subjective event interpretation. The analysis is not designed to establish causation. Several selected event terms have small report counts; a large ROR does not turn those counts into an event rate. Table 2 often prints confidence-interval bounds in reverse order; the intervals on this page are transcribed in ascending order.

Contextual only. The signals do not establish that an imidazole caused a reported event, do not quantify harm among treated people, and do not verify that a report concerned vaginal treatment for VVC. They therefore neither support nor challenge the current oral-versus-intravaginal harm claim.

No practical change. This analysis does not change the current claim or action and provides no basis for an individual treatment decision, pregnancy advice, or added kidney monitoring.

The study was funded in part by Clinical Toxicology of Chinese Society of Toxicology, grant CST2021CT101. The authors declared no competing interests.

Zhou T, Chen C, Chen X, et al. “Possible adverse events of imidazole antifungal drugs during treatment of vulvovaginal candidiasis: analysis of the FDA Adverse Event Reporting System.” Scientific Reports. 2024;14(1):14560. Published June 24, 2024. Read the original publication.