Related Experiment Video
Updated: Jun 20, 2026
Multiplexed Immunofluorescence Analysis and Quantification of Intratumoral PD-1+ Tim-3+ CD8+ T Cells
Published on: February 8, 2018
Predicting anti-PD-1 immune checkpoint blockade response in melanoma patients with spatially aware machine learning
Alyssa Pybus1, Raphael Kirchgaessner1,2, Jonathan Nguyen3
1Department of Machine Learning, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, USA.
Abstract:
There is an acute need to accurately identify patients with advanced melanoma who are most likely to respond to anti-PD1 immune checkpoint blockade (ICB) therapy. While anti-PD1 therapy can be highly effective in advanced melanoma patients, only 30-40% of patients respond well. In this study, we apply single-cell spatial proteomics together with statistical and machine learning (ML) methods to successfully predict advanced melanoma patient response to anti-PD1 ICB in a cohort of 12 patients with >8 million cells. While no single molecular feature is sufficient to predict ICB response in our cohort, ML models integrating multiple molecular features accurately predict response in 11 of 12 patients. A recurrent cellular neighborhood analysis revealed a tumor-infiltrating lymphocytes niche that was present in the tumors of most responders. This neighborhood, tumor microenvironment immune cell composition, and levels of nitric oxide synthases were all important features used by our ML models to make accurate predictions. Optimal predictive performance by our ML models-a ROC AUC of 0.76-was achieved when using all molecular features, including cellular spatial relationships, but limiting our analysis to only immune-rich tissue regions. This study demonstrates the feasibility of using machine learning models to accurately predict patient response to anti-PD1 ICB therapy using spatial proteomics datasets.
Insights
Accurately predicting response to anti-PD1 immune checkpoint blockade (ICB) therapy in advanced melanoma is crucial. Machine learning models integrating spatial proteomics data successfully predicted ICB response in 11 of 12 patients.
Area of Science:
- Oncology
- Immunology
- Computational Biology
Background:
- Identifying advanced melanoma patients likely to respond to anti-PD1 immune checkpoint blockade (ICB) therapy is critical, as only 30-40% achieve significant benefit.
- Current predictive biomarkers for anti-PD1 ICB therapy in melanoma are limited, necessitating novel approaches.
Purpose of the Study:
- To develop and validate machine learning (ML) models for predicting patient response to anti-PD1 ICB therapy using single-cell spatial proteomics.
- To identify key molecular and cellular features within the tumor microenvironment that correlate with anti-PD1 ICB response.
Main Methods:
- Applied single-cell spatial proteomics on a cohort of 12 advanced melanoma patients.
- Integrated statistical and machine learning (ML) methods, including recurrent cellular neighborhood analysis.
- Evaluated predictive performance using receiver operating characteristic area under the curve (ROC AUC).
Main Results:
- ML models integrating multiple molecular features accurately predicted anti-PD1 ICB response in 11 out of 12 patients.
- A specific tumor-infiltrating lymphocytes niche, immune cell composition, and nitric oxide synthase levels were identified as important predictive features.
- Optimal prediction (ROC AUC of 0.76) was achieved using all molecular features, including spatial relationships, within immune-rich regions.
Conclusions:
- Machine learning models utilizing spatial proteomics data can accurately predict anti-PD1 ICB therapy response in advanced melanoma.
- Integrating multi-omic spatial data, particularly cellular spatial relationships and immune microenvironment composition, enhances predictive accuracy.
- This approach demonstrates the feasibility of personalized treatment selection for advanced melanoma patients receiving immunotherapy.
