Adverse drug reaction profiles of histone deacetylase inhibitors

TL;DR

This study examines the side effects of different histone deacetylase inhibitors and links them to their chemical and pharmacological properties to improve drug safety.

Contribution

The study identifies how the unique physicochemical and pharmacological properties of HDACIs correlate with their distinct adverse drug reaction profiles.

Findings

Vorinostat's high rate of cardiac ADRs is linked to hERG ion-channel inhibition.

Panobinostat's gastrointestinal ADRs are associated with HDAC3 inhibition.

Entinostat's lack of HDAC9 inhibition may explain its absence of depression-related ADRs.

Abstract

The safety and tolerability of histone deacetylase inhibitors (HDACIs) were always a matter of concern. This study aimed to explore the global observational adverse drug reaction (ADR) profiles of the HDACIs: vorinostat, belinostat, panobinostat, pracinostat, entinostat, romidepsin, bufexamac and sodium phenylbutyrate. This study focussed on the investigation whether associations between HDACIs’ ADR profiles and their physicochemical and pharmacological properties exist and how these associations may be harnessed to contribute to safer clinical use. The ADRs of HDACIs were curated from the World Health Organisation (WHO) VigiAccess database (1976–2024). Pharmacology data was curated from the chemical database of bioactive molecules with drug-like properties, European Molecular Biology Laboratory (ChEMBL). Physiochemical and pharmacokinetic properties were curated from PubChem, Drug Bank, ChemDraw and FDA new drug application documents for the HDACIs studied. In total, n = 12,779 ADRs were reported for the selected HDACIs, with n = 15/27 of the system organ class (SOC)-related ADRs showing statistical significance (X2, P < .05). Vorinostat accounted for n = 4,225 of the total ADRs, which may be explained by (1) its extensive use in the clinic and (2) it possessed the most clinically achievable off-target pharmacological interactions (n = 9). This included potent inhibition of the human-ether-à-go-go-related gene (hERG) ion-channel (IC50 = 322 nM vs. cmax = 1,200 nM), which is likely a contributing factor to the highest rate of cardiac ADRs observed (n = 146, P < .05). Musculoskeletal ADRs associated with vorinostat (n = 58, P < .05) can be ascribed to its unique inhibition of HDAC4 (IC50 = 540 nM vs. cmax = 1,200 nM). The gastrointestinal (GI) ADRs of panobinostat (n = 488, P < .05) can partly be related to HDAC3 inhibition (IC50 = 2.0 nM vs. Cmax = 8.0 nM). The absence of HDAC9 inhibition was unique to entinostat and may explain its beneficial drug reaction in depression (n = 0 cases). Thrombocytopenia observed with vorinostat, panobinostat and romidepsin may stem from dual inhibition of HDAC1 and HDAC2 and their higher volume of distributions (Vd). Despite all HDACIs having a similar mechanism of action, their unique pharmacology and differing physicochemical and pharmacokinetic properties landscape results in varying ADR profiles.

The online version contains supplementary material available at 10.1038/s41598-025-19717-w.

Figures