COLD SPRING HARBOR, NY (GenomeWeb) – Researchers led by Elaine Mardis at Washington University in St. Louis are applying genomics to guide cancer immunotherapy, specifically developing and validating personalized cancer vaccines using mutational information gleaned from an individual's tumor.
During a translational genomics session at the Biology of Genomes meeting held here last week, Mardis, director of technology development at WUSTL's McDonnell Genome Institute, described these efforts as well as some of the challenges that remain in genomics-guided immunotherapy.
Mardis touched on findings from a proof-of-principle study she and her team published in Science last year on three individuals with stage III cutaneous melanoma who were treated with dendritic cell-based vaccines. She noted that two more individuals with melanoma have since received similar treatments.
In their initial efforts to find cancer vaccine targets from genomic data, she explained, the researchers looked at whether they could use somatic mutation data to find neoepitopes that would elicit a response from the immune system's T cells without causing severe adverse effects.
In two of the five melanoma patients tested so far, the team was able to prompt a T-cell response for just under half of the predicted antigen peptides tested in each individual.
To design vaccine-targetable neoantigens, Mardis explained, the team typically analyzes exome sequences for matched tumor and normal samples from each patient, predicting peptide sequences from non-synonymous mutations in the tumor and whittling the set of potential neoantigens down based on their anticipated interactions with the immune system.
"Our work combines genomic analysis to identify mutated peptides produced by individual cancers, then evaluates these peptides relative to the patient's [human leukocyte antigen] molecules to identify those most likely to elicit an immune response in the context of a personalized vaccine," Mardis and her co-authors wrote in the abstract for the presentation.
As part of their pVAC-Seq personal neoantigen discovery pipeline, introduced in Genome Medicine earlier this year, the researchers also use RNA sequencing to determine whether a variant is expressed in the tumor and to track isoform-level expression.
Though they have now come up with ways to profile small insertion and deletion mutations that might serve as immune targets, the researchers are interested in finding ways to target other types of somatic mutations such as structural variants with cancer vaccines.
And there are other outstanding questions in the field, as well. Before such therapies can be applied more widely, researchers need to find ways of dealing with patients who have rarer HLA alleles, Mardis explained. And more work is needed to tease apart details on which potentially targetable neoantigens are processed and presented to the immune system's major histocompatibility complex.
Researchers will also need to continue chipping away at the other aspects of cancer vaccine design and applications, Mardis said, which may require studies on everything from the optimal vaccine platform and number of neoantigen targets to the most beneficial timing for vaccination.
For their part, she and her colleagues are continuing to apply the approach to individuals with melanoma, using both dendritic cell vaccines and anti-PD1 therapy. They are also involved in projects that take the strategy to other cancer types, including triple-negative breast cancer, pancreatic cancer, and pediatric cancer.
The team is also interested in exploring the possibility of using cancer vaccines to target DNA damage that has been induced in tumors through past treatment with alkylating chemotherapy, which would theoretically offer additional treatment options to patients who have already undergone one or more prior treatments.