5 Key Takeaways

• 1

  A machine-learning approach using blood-derived DNA can predict immunotherapy response in melanoma.

• 2

  The study identified nine conserved tumor microenvironment ecotypes linked to biology, survival, and immunotherapy response.

• 3

  The cfDNA-based method outperformed traditional biomarkers like tumor mutational burden and PD-L1 expression.

• 4

  High levels of inflammatory ecotypes SE7 and SE8 correlated with better treatment outcomes, while SE4 indicated resistance.

• 5

  This noninvasive blood test may enable longitudinal monitoring of tumor microenvironment changes during therapy.