3 Alarming Cancer Immunotherapy Response Prediction Methods Revealed
Three independent research teams have developed promising new approaches to cancer immunotherapy response prediction, each capable of identifying which melanoma patients are most likely to benefit from immune checkpoint therapies. The findings, published in Nature Medicine and Nature in August 2018, were funded in part by the National Institutes of Health and represent a major step toward personalized oncology. According to the Mayo Clinic, melanoma is the most serious form of skin cancer, with more than 90,000 new diagnoses expected in the United States in 2018 alone.
Why Cancer Immunotherapy Response Prediction Matters
Immune checkpoint therapy works by blocking the proteins that certain cancer cells use to suppress the immune system’s attack against them. The treatment has shown remarkable results in melanoma and other cancers, but only a subset of patients respond. For those who do not respond, the therapy still exposes them to side effects without therapeutic benefit.
A reliable cancer immunotherapy response prediction method would allow clinicians to identify non-responders in advance, sparing them unnecessary treatment while directing responders toward the therapy most likely to help them. As NIH’s Dr. Eytan Ruppin stated, being able to predict who is highly likely to respond and who is not will enable more accurate and precise treatment guidance.
3 Cancer Immunotherapy Response Prediction Methods From Independent Teams
1. Gene Expression Immuno-Predictive Scoring (NCI Team)
A team led by Drs. Eytan Ruppin and Noam Auslander at NIH’s National Cancer Institute developed a cancer immunotherapy response prediction tool based on the gene expression features of tumors that underwent spontaneous immune mediated shrinkage. By analyzing these features, the researchers computed an immuno-predictive score for each tumor sample.
When tested on 297 melanoma tumor samples, the predictor identified nearly all patients who responded to checkpoint therapies and more than half of those who did not. It outperformed all previously published predictors and remained accurate across multiple independent melanoma patient datasets.
2. Computational Tumor Escape Modeling (Harvard Team)
A team led by Drs. Kai Wucherpfennig and X. Shirley Liu at Harvard University developed a different cancer immunotherapy response prediction approach focused on identifying gene expression features that predict immune escape, meaning which tumors are most likely to evade immune destruction.
Their computational model was trained on data from 33,000 tumor samples collected across 189 studies. When tested against publicly available checkpoint therapy outcome data for melanoma patients, it outperformed previously published prediction models and demonstrated the power of large scale gene expression analysis in guiding treatment decisions.
3. Blood Based Exosomal PD-L1 Detection (University of Pennsylvania Team)
The third cancer immunotherapy response prediction approach, developed by Drs. Wei Guo and Xiaowei Xu at the University of Pennsylvania, uses blood samples rather than tumor tissue. This is a significant practical advantage, as blood draws are far less invasive and easier to obtain than tumor biopsies.
The team found that metastatic melanomas release tiny sacs called exosomes that carry programmed death ligand 1, or PD-L1, on their surfaces. These PD-L1 carrying exosomes attach to the PD-1 receptor on T cells, turning off the cancer fighting immune response. Patients with higher levels of exosomal PD-L1 before anti-PD-1 checkpoint therapy were less likely to respond to treatment, establishing exosomal PD-L1 as a viable blood based cancer immunotherapy response prediction biomarker.
What Comes Next for Immunotherapy Prediction
Each of the three cancer immunotherapy response prediction methods has shown strong independent results. Researchers suggest that combining these approaches could produce even more accurate predictions. With further development and validation using larger patient datasets, these tools could become standard components of pre-treatment evaluation protocols for melanoma and potentially other cancers where checkpoint therapy is used.
FOMAT conducts Phase I through Phase IV clinical research across a national network of investigator sites throughout the United States. To learn more about active oncology studies, visit our patient active studies page.