Article Impact Level: HIGH Data Quality: STRONG Summary of Journal of Molecular Medicine, 103(6), 725–737. https://doi.org/10.1007/s00109-025-02550-z Dr. Bruna Pereira Sorroche et al.
Points
- A novel four-gene signature, comprising CD24, NFIL3, FN1, and KLRK1, was identified to accurately predict patient resistance to standard immunotherapy for advanced melanoma.
- Patients with high expression of these genes were found to be 230 times more likely not to respond to treatment and had significantly lower five-year survival rates.
- The identified genes are linked to known mechanisms of immune evasion and tumor progression, which explains why some patients fail to benefit from immunotherapy despite being technically eligible.
- The signature’s predictive power was successfully validated in two independent international cohorts, confirming its robustness and potential for broad clinical applicability across different patient populations.
- This discovery aims to develop a cost-effective diagnostic tool that guides treatment decisions, thereby avoiding ineffective therapies and optimizing resource allocation in public health systems.
Summary
A study of 35 advanced melanoma patients treated with anti-PD-1 immunotherapy between 2016 and 2021 aimed to identify a predictive transcriptomic signature for treatment response. Using the NanoString nCounter panel to evaluate 579 immunology-related genes, researchers identified 18 genes upregulated in non-responders and three in responders. Multivariate analysis isolated a four-gene signature (CD24, NFIL3, FN1, KLRK1) whose increased expression was strongly associated with treatment failure. Patients with high expression of this signature were 230 times more likely not to respond to immunotherapy compared to those with low expression.
The resulting predictive score demonstrated high discriminatory power, achieving an area under the curve (AUC) of 0.935 (p < 0.001). This signature was also significantly associated with poorer clinical outcomes, including progression-free survival, overall survival, and survival following immunotherapy (all p < 0.001). Specifically, the five-year overall survival rate was 5.9% for patients with high gene expression, compared to 48.1% for those with low expression. The signature’s prognostic value was not detailed with specific hazard ratios or confidence intervals in the provided source.
To confirm the findings, the signature was validated in two independent international cohorts, demonstrating robust and generalizable performance with AUCs of 0.758 (p = 0.036) and 0.833 (p = 0.004), respectively. The use of NanoString technology presents a more accessible platform for clinical implementation. This research represents a significant step toward personalizing melanoma treatment by identifying patients who are unlikely to benefit from immunotherapy, potentially sparing them from ineffective treatment and helping to optimize healthcare resource allocation.
Link to the article: https://link.springer.com/article/10.1007/s00109-025-02550-z
References Sorroche, B. P., De Jesus Teixeira, R., De Souza, V. G., Tosi, I. C., Tostes, K., Laus, A. C., Santana, I. V. V., De Lima Vazquez, V., & Arantes, L. M. R. B. (2025). CD24, NFIL3, FN1, and KLRK1 signature predicts melanoma immunotherapy response and survival. Journal of Molecular Medicine, 103(6), 725–737. https://doi.org/10.1007/s00109-025-02550-z
