Malignant melanoma, a cancer that originates in melanin-producing cells, is the deadliest variety of skin cancer. In the past 11 years, immunotherapy has increased the median survival rate of advanced melanoma from nine months to six years. However, it is still a developing treatment.
Farida Zakariya, a masters student in McGill’s Division of Experimental Medicine, explored the use of artificial intelligence (AI) in immunotherapy of melanoma in a new review paper.
“One major problem of immunotherapy is the diversity in response rate across different cancer types,” Zakariya wrote in an email to The Tribune. “Some cancers like melanoma are very responsive to immunotherapy while others are not or become resistant over time.”
Currently, the most common immunotherapy for melanoma stimulates the production of antibodies that bind to immune checkpoint inhibitor proteins. These proteins are usually found on the surface of healthy cells and inhibit cytotoxic T-cells—an immune system “weapon” that recognizes and induces cell death in infected or cancerous cells. When a T-cell comes into contact with a checkpoint inhibitor protein, it receives a signal to turn off and leave the cell alone.
Cancer cells sometimes produce these same checkpoint inhibitors to avoid detection. When a patient undergoes immune checkpoint inhibitor therapy, checkpoint inhibitors become covered with antibodies, T-cells are inhibited from turning off, and the cancer is recognized and killed.
However, not all cancers are responsive to treatment, and some patients experience a recurrence of their cancer with newfound resistance. Thus, developing different therapies is crucial.
The potential role of mutanome in developing immunotherapies is a new subject of research.
“Mutanomes are the entire mutations that underlie a particular type of cancer. Every cancer is different and within the same cancer type, the underlying mutations differ from person to person,” Zakariya wrote.
Via sophisticated whole genome profiling and next-generation sequencing, researchers can now obtain the complete genomic profile of the mutanome of each patient and an understanding of its spatial folding in the genome for more precise treatment.
Vaccine immunotherapies take data from mutanome sequencing to re-create patient-specific neoantigens—unique proteins that cancer cells present on their surface. If a range of expected neoantigens are injected into the patient, T-cells learn to diversify their targets, and the tumor is destroyed.
In reality, however, cancer treatments are never this straightforward. Therapies are altered by differences in the absorption and metabolism of treatments. Furthermore, there is a real lack of research into how these responses vary by ethnicity, sex, and disease stage, an information gap that causes groups of patients to slip through the cracks.
Genetic variations between patient’s genomes, along with the mutations of the cancers themselves, create an overload of data for researchers to consider. This is where AI comes in.
From identifying neoantigen DNA in the mutanome to developing personalized vaccines and predicting metastatic risk—the chances that the cancer will spread—AI’s ability to detect patterns in large data sets has the potential to revolutionize individualized treatment.
“[AI] will greatly eliminate most of the barriers that the large scale sequencing required for the application of mutanome [sequencing] in melanoma immunotherapy,” Zakariya wrote.
While AI is already changing the landscape of treatment and drug discovery, it has limitations: The dynamic relationship between the mutanome, immune system, signaling pathways, and the patient’s biological makeup and environment are challenging for current static models.
“This [problem] is not something AI can solve for now, but it can increase the turnaround time of the drug discovery process,” Zakariya wrote, “[This] enabl[es] more hits to be identified within a short period of time and increas[es] sequencing and testing capacities. Thus, ensuring that more medications will be available for cancers in the near future.”
Zakariya is currently researching the effects of inhibiting a protein linked to cancer resilience to investigate its role in the tumor’s response to immunotherapy. “I am hoping that the paper will unlock new pathways that can be explored by researchers in the pursuit of a cure for cancer.”
