Broad-Spectrum Antibiotic Use and Disease Progression in Early-Stage Melanoma Patients: A Retrospective Cohort Study

Abstract

Animal studies and a few clinical studies have reported mixed findings on the association between antibiotics and cancer incidence. Antibiotics may inhibit tumor cell growth, but could also alter the gut-microbiome-modulated immune system and increase the risk of cancer. Studies that assess how antibiotics affect the progression of cancer are limited. We evaluated the association between broad-spectrum antibiotic use and melanoma progression. We conducted a retrospective cohort study using IQVIA PharMetrics® Plus data (2008–2018). We identified patients with malignant melanoma who underwent wide local excision or Mohs micrographic surgery within 90 days of first diagnosis. Surgery date was the index date. Patients were excluded if they had any other cancer diagnosis or autoimmune disorders in 1 year before the index date (“baseline”). Exposure to broad-spectrum antibiotics was identified in three time windows using three cohorts: 3 months prior to the index date, 1 month after the index date, and 3 months after the index date. The covariates were patients’ demographic and clinical characteristics identified in the 1-year baseline period. The patients were followed from the index date until cancer progression, loss of enrollment, or the end of 2 years after the index date. Progression was defined as: (i) any hospice care after surgery, (ii) a new round of treatment for melanoma (surgery, chemotherapy, immunotherapy, targeted therapy, or radiotherapy) 180 days after prior treatment, or (iii) a metastasis diagnosis or a diagnosis of a new nonmelanoma primary cancer at least 180 days after first melanoma diagnosis or prior treatment. A high-dimensional propensity score approach with inverse weighting was used to adjust for the patients’ baseline differences. Cox proportional hazard regression was used for estimating the association. The final samples included 3930, 3831, and 3587 patients (mean age: 56 years). Exposure to antibiotics was 16% in the prior-3-months, 22% in the post-1-month, and 22% in the post-3-months. In the pre-3-months analysis, 9% of the exposed group and 9% of the unexposed group had progressed. Antibiotic use was not associated with melanoma progression (HR: 0.81; 95% CI: 0.57–1.14). However, antibiotic use in subsequent 1 month and subsequent 3 months was associated with 31% reduction (HR: 0.69; 95% CI: 0.51–0.92) and 32% reduction (HR: 0.68; 95% CI: 0.51–0.91) in progression, respectively. In this cohort of patients with likely early-stage melanoma cancer, antibiotic use in 1 month and 3 months after melanoma surgery was associated with a lower risk of melanoma progression. Future studies are warranted to validate the findings.