Head and neck cancers: antibiotics associated with poorer survival, without proven causality

Could antibiotics affect treatment outcomes in certain head and neck cancers? A meta-analysis involving 6,293 patients finds an association between their use around the start of cancer treatment and poorer survival. But the data, mainly observational and highly susceptible to bias, do not establish that antibiotics are responsible.

Illustration of the gut and oral microbiota in the context of head and neck cancer treatments.
NutritionEditorial illustration, not study data.

What the researchers compared

The review brings together eleven studies involving patients with head and neck squamous cell carcinoma or nasopharyngeal carcinoma. Treatments included immune checkpoint inhibitors, definitive chemoradiotherapy or chemoimmunotherapy.

The researchers compared outcomes in patients exposed to systemic antibiotics around the start of treatment with those in unexposed patients. Nine studies, involving 4,507 patients, contributed to the survival meta-analyses. The supplied abstract does not specify follow-up durations.

An association found across several settings

Antibiotic exposure was associated with poorer overall survival, with a hazard ratio, or HR, of 1.78 and a 95% confidence interval of 1.52 to 2.08. For progression-free survival, the HR was 1.58, with an interval of 1.28 to 1.96.

These measures indicate neither a loss of lifespan nor an absolute difference in the probability of death. The abstract does not provide the data needed to calculate these differences.

The association was also observed in the groups treated with immunotherapy and chemoradiotherapy. Survival results remained similar when the researchers removed each study from the analysis in turn. This stability strengthens the consistency of the signal, but does not eliminate study biases.

The microbiome is a hypothesis, not a proven explanation

The authors suggest that antibiotics may disrupt the gut and oral microbiota, potentially influencing the effectiveness of cancer treatments. This mechanism is a biological hypothesis: this synthesis does not demonstrate that it explains the observed association.

Most studies had a serious or high risk of bias, mainly related to the indication for antibiotics. In other words, the medical condition prompting their prescription may itself be linked to prognosis. Comparing exposed and unexposed patients is therefore not enough to isolate the drug’s own effect.

What this study cannot determine

These findings do not demonstrate that avoiding an antibiotic would improve survival. Nor do they justify changing a prescription on your own during cancer treatment.

The authors call for prospective studies incorporating the microbiome to better distinguish any potential effect of antibiotics from that of the circumstances prompting their prescription.

Limitations to keep in mind

Only the PubMed abstract supplied in the dossier was consulted, not the full text. The data are mainly observational, and most studies have a serious or high risk of bias. Follow-up durations, absolute survival differences, funding and conflicts of interest cannot be verified from this excerpt.

What we can take away

The signal deserves further study, but it does not establish that antibiotics are to blame: their prescription may also reflect a medical condition associated with a poorer prognosis.

Sources and provenance

Partager cet article

X f in

An error or a source to report? Contact the editorial team.