Predicting Tumor Microenvironment and Immunotherapy Response in Lung Cancer
Author Information
Author(s): Patkar Sushant, Chen Alex, Basnet Alina, Bixby Amber, Rajendran Rahul, Chernet Rachel, Faso Susan, Kumar Prashant A., Desai Devashish, El-Zammar Ola, Curtiss Christopher, Carello Saverio J., Nasr Michel R., Choyke Peter, Harmon Stephanie, Turkbey Baris, Jamaspishvili Tamara
Primary Institution: SUNY Upstate Medical University
Hypothesis
Can a deep learning approach predict tumor microenvironment composition and immunotherapy response from histopathology images in non-small cell lung cancer?
Conclusion
The HistoTME model effectively predicts tumor microenvironment signatures and improves prognostication for lung cancer patients receiving immunotherapy.
Supporting Evidence
- HistoTME predicted microenvironment signatures improve prognostication of lung cancer patients receiving immunotherapy.
- The model achieved an AUROC of 0.75 for predicting treatment responses.
- HistoTME complements PD-L1 expression in identifying patients likely to respond to immunotherapy.
- Pearson correlation of 0.5 was achieved with ground truth on independent tumor cohorts.
- Significant differences in cellular abundances of T cells, B cells, and macrophages were observed across predicted TME subtypes.
Takeaway
Researchers created a computer program that looks at images of lung cancer to help doctors understand how the tumor environment affects treatment responses.
Methodology
The study used a weakly supervised deep learning model called HistoTME to analyze histopathology images and predict tumor microenvironment signatures.
Potential Biases
Potential biases due to the retrospective nature of the study and reliance on specific patient cohorts.
Limitations
The study was retrospective and lacked external validation datasets.
Participant Demographics
The study included 652 patients with non-small cell lung cancer, primarily treated at SUNY Upstate Medical University.
Statistical Information
P-Value
0.0012
Confidence Interval
[95% CI: 0.61-0.88]
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website