An investigative network analysis mapping global cancer epidemiology

Area of Science:

• Epidemiology
• Genetics
• Network analysis

Background:

• Cancer incidence is shaped by genetic and environmental factors.
• Similar incidence patterns across cancers may indicate shared underlying causes.
• Understanding these relationships is crucial for public health interventions.

Purpose of the Study:

• To investigate if cancers with similar global incidence patterns share etiological mechanisms.
• To identify clusters of cancers based on age-standardized incidence rates using network analysis.
• To explore the biological coherence and potential shared mutational underpinnings of these clusters.

Main Methods:

• Utilized age-standardized incidence data for 36 cancer types from GLOBOCAN 2022 across 185 countries.
• Applied network clustering using six algorithms (Leiden, Surprise, Walktrap, Girvan-Newman, Infomap, spectral clustering) based on Spearman's correlation.
• Analyzed mutational signatures and recurrent mutation overlap within identified clusters.

Main Results:

• A dominant cluster emerged, including kidney, pancreatic, colorectal, thyroid cancers, and hematological malignancies.
• A second cluster comprised lung cancer, mesothelioma, melanoma, non-melanoma skin cancer, and head and neck cancers.
• The dominant cluster exhibited significant enrichment of shared mutational signatures and recurrent mutations (p < 0.05).
• Eigenvector centrality correlated positively with global cancer incidence rates.

Conclusions:

• Unsupervised network analysis successfully identified biologically coherent cancer clusters.
• These clusters suggest shared etiological mechanisms influencing cancer incidence globally.
• Findings can inform targeted public health interventions and future research into cancer causes.