Related Experiment Videos
Fluoroquinolone Exposure and Cancer Risk in Interstitial Lung Disease: A Propensity-Score-Matched Cohort Study Using
Yi-Fan Sun1, Yu-Ting Chiu1,2, Yung-En Ko1
1Department of Family Medicine, Geriatric Medicine, Chest Medicine and Medical Research, Ditmanson Medical Foundation Chia-Yi Christian Hospital, No. 539, Zhongxiao Rd., East Dist., Chiayi 600566, Taiwan.
Pharmaceuticals (Basel, Switzerland)
|July 28, 2026
Summary
Fluoroquinolone (FQ) antibiotics did not show a protective effect against cancer risk in patients with interstitial lung disease (ILD). The observed increased cancer risk was likely due to confounding factors, not a direct drug effect.
Area of Science:
- Oncology
- Pharmacology
- Epidemiology
Background:
- Investigating the association between fluoroquinolone (FQ) antibiotics and cancer risk.
- Focusing on patients with interstitial lung disease (ILD), a population with high inflammation and infection susceptibility.
Purpose of the Study:
- To comprehensively investigate the complex association between FQ use and cancer risk in ILD patients.
- To address potential biases like immortal time bias in the analysis.
Main Methods:
- Large-scale retrospective cohort study with 7906 matched patients (3953 pairs) using propensity score matching (PSM).
- Application of three statistical models: standard Cox, time-dependent Cox, and Fine-Gray competing-risks models.
- Strict index date definition to eliminate immortal time bias.
Main Results:
- FQ exposure was associated with increased all-cause cancer risk (aHR 1.45; aSHR 1.28) after bias correction.
- Elevated risks observed for specific cancers, including prostate cancer.
- No significant dose-response relationship found between cumulative FQ dose and overall cancer risk.
Conclusions:
- The hypothesized protective effect of FQ on cancer risk was not supported.
- Increased cancer risk in categorical models, without dose-response, suggests confounding by indication and reverse causation.
- Findings indicate that infections or underlying ILD severity, rather than a direct pharmacological effect, likely drive the observed associations.
