Heterogeneous shifts in cancer research topics during the COVID-19 era: evidence from topic modeling of oncology journals.

OBJECTIVES: The coronavirus disease 2019 (COVID-19) pandemic disrupted healthcare systems and may have influenced the direction of cancer research, particularly epidemiology and prevention research. However, empirical evidence regarding changes in research topics during the COVID-19 era remains limited. METHODS: We analyzed 8,694 articles published from 2016 to 2025 in 3 journals using PubMed bibliographic data. Latent Dirichlet allocation topic modeling identified 10 thematic clusters. Differences in topic distributions between the pre-pandemic (2016-2019) and post-pandemic-onset (2020-2025) periods were assessed using the Pearson chi-square test and logistic regression. Interaction models assessed whether temporal changes differed among journals. RESULTS: The overall distribution of topics differed significantly between the 2 periods (χ²=21.42, p=0.011). Logistic regression indicated modestly higher odds of classification as Topic 7, the representative topic of interest, during the 2020-2025 period (β=0.168, p=0.013; odds ratio≈1.18). Interaction analysis showed that this temporal change differed by journal: the interaction was significant for Cancer Epidemiology, Biomarkers & Prevention relative to the reference journal (British Journal of Cancer; β=-0.959, p<0.001) but not for European Journal of Cancer Prevention (p=0.269). Collectively, these findings indicate selective thematic reorientation rather than wholesale transformation of cancer research agendas, with changes varying according to journal scope and disciplinary focus. CONCLUSION: Cancer research topics shifted modestly during the COVID-19 era, with heterogeneous patterns among journals. These findings suggest that major global health crises may be associated with shifts in research agendas, although these changes may vary according to journal scope and disciplinary context.