Machine learning analysis of DNA methylation profiles distinguishes primary lung squamous cell carcinomas from head and neck metastases

Discriminating lung primary tumors and metastases

Pulmonary metastases of head and neck squamous cell carcinoma (HNSC) are currently difficult to distinguish from primary lung squamous cell carcinomas (LUSCs). Differentiating these tumor types has important clinical implications, as whether the lung tumor is primary or has spread can affect the treatment options offered to a patient. Here, Jurmeister et al. developed a machine learning algorithm that exploits the differential DNA methylation observed in primary LUSC and metastasized HNSC tumors in the lung. Their method was able to discriminate between these two tumor types with high accuracy across multiple cohorts, suggesting its potential as a clinical diagnostic tool.