Monoclonal Antibodies and Small-Molecule Inhibitors Associated Osteonecrosis of Jaw: A Retrospective Pharmacovigilance Study

Background

Osteonecrosis of the jaw (ONJ) is an adverse effect associated with medications such as monoclonal antibodies and small-molecule inhibitors.

Purpose

This study assesses the association between monoclonal antibodies and small-molecule inhibitors with ONJ.

Study Design, Setting, Sample

The study design was a retrospective pharmacovigilance case series. The sample was derived from the United States Food and Drug Administration Adverse Event Reporting System database reporting ONJ from March 2004 to March 2024. The inclusion criteria were the reports relating ONJ occurrence with exposure to monoclonal antibodies (trastuzumab, bevacizumab, denosumab, elotuzumab, isatuximab, pertuzumab, ramucirumab, romosozumab, and ado-trastuzumab emtansine) and small-molecule inhibitors (abemaciclib, alpelisib, axitinib, cabozantinib, lapatinib, lenvatinib, palbociclib, ribociclib, and sunitinib). The exclusion criteria were the reports associating the role of monoclonal antibodies and small molecule inhibitors to ONJ with roles other than primary suspicion.

Predictor Variable

Not applicable.

Main Outcome Variable

The outcome variable is the case status divided into case and noncase. Cases were defined as reports with ONJ associated with monoclonal antibodies or small-molecule inhibitors while noncases were the reports with other adverse events. Secondary outcome variables were death, hospitalization, and disability observed with monoclonal antibodies/small-molecule inhibitors–associated ONJ.

Covariates

Age and gender were the covariates included in this study.

Analyses

The case–noncase approach was applied for signal detection, using frequentist (reporting odds ratio [ROR] and proportional reporting ratio [PRR]) and Bayesian methods (lower limit of 95% CI of Information component [IC025]). The ROR is determined by comparing the odds of ONJ being reported for a given drug to the odds of the same event being reported for all other drugs and PRR is estimated by the ratio of proportion of reports for a specific drug with ONJ over the proportion of reports with ONJ for all other drugs. The information component is defined as defined as the logarithmic ratio of the observed ONJ with the primary suspected drug relative to the expected frequency of the drug-ONJ pair based on overall reporting rates in the database. Outcomes were analyzed with statistical comparisons using χ² tests (χ2) at P ≤ .05.

Results

A total of 7,402 reports were included with median age ranging between 61 and 76 years with female preponderance. Denosumab (ROR: 64.7 [62.5, 67]; PRR: 61.9 [59.9, 64.1]; P < .05; and IC025: 5.1), romosozumab (ROR 4.2 [3, 6]; PRR: 4.2 [3, 6]; P < .05; and IC025: 1.5), and lenvatinib (ROR: 3.1 [2.3, 4.2]; PRR: 3.1 [2.3, 4.1]; P < .05; and IC025: 1.2) showed positive signals for ONJ compared to all other drugs in the database by both frequentist and Bayesian analyses indicating a potential association. Sixty-two point eight percent cases resulted in hospitalization and was no significant differences were observed in the distribution of outcomes between the monoclonal antibodies and small molecule inhibitors (χ²: 12; df: 14; P = .6). The risk was mainly observed in elderly and female patients.

Conclusion and Relevance

Denosumab, romosozumab, and lenvatinib are significantly associated with ONJ, particularly in older and female patients. Further research is needed to understand the mechanisms and improve risk management strategies.