# Disproportionality analysis of drug-associated progressive multifocal leukoencephalopathy using spontaneous reports: A 20-year signal detection study based on the FAERS database

## TL;DR

This study analyzed 20 years of drug safety reports to identify medications linked to a rare brain disease called PML, finding 72 drugs that showed consistent signals.

## Contribution

The study is the first to use four complementary algorithms on FAERS data to detect PML drug associations over two decades.

## Key findings

- 72 drugs showed consistent PML signals, mostly immunomodulators and biologics like natalizumab and rituximab.
- PML reporting increased significantly in non-HIV populations during the study period.
- Over one-third of PML reports were associated with life-threatening outcomes or death.

## Abstract

Progressive multifocal leukoencephalopathy (PML) is a rare, often fatal demyelinating disease caused by JC virus reactivation in immunocompromised patients. With the increasing use of immunosuppressants and biologics, PML reporting in non-HIV populations is rising. This study aimed to evaluate drug-associated PML reporting signals using real-world pharmacovigilance data.

We analyzed FAERS database from 2004Q1 to 2024Q4. We identified PML reports via MedDRA Terms and manual validation. Four algorithms (ROR, PRR, BCPNN, MGPS) were jointly applied, with drugs showing signals across all four defined as high-risk.

7,244 PML reports involving 298 drugs were identified; 72 drugs showed consistent signals, predominantly immunomodulators (e.g., natalizumab, rituximab), antineoplastics, and biologics. High-risk indications included multiple sclerosis, lymphoma, autoimmune diseases, and organ transplantation. PML reporting increased substantially in non-HIV populations. Time-to -reporting varied widely (49–1343days). Over one-third of reports were associated with life-threatening outcomes or death.

This analysis identified 72 drugs with consistent PML reporting signals. However, these findings represent statistical associations in spontaneous reports, not causal relationships or true incidence rates. Inherent limitations—including underreporting, incomplete medication histories, and lack of exposure denominators—require cautious interpretation. Prospective validation studies are essential to establish causality and quantify absolute risks.