Immune therapy with checkpoint inhibitors against PD-L1 and PD-1 is usually considered fairly gentle treatment in the cancer treatment field as far as side effects. But up to 30% of patients receiving them can develop serious toxicity, particularly when used in combination treatments. To help identify those patients ahead of treatment, a team of investigators from UCLA Jonsson Comprehensive Cancer Center have identified a germline biomarker signature that predicts toxicity with 77% accuracy.
“The toxicity is a new type of side effect that we only see with immune therapy,” says Joanne B. Weidhaas of UCLA Jonsson Comprehensive Cancer Center, vice chair Department of Radiation Oncology, and director Division of Molecular and Cellular Oncology at UCLA Health. Her team published their discovery in a paper in the Journal for Immunotherapy of Cancer. “It is a kind of autoimmunity where the immune system attacks the body instead of the tumor.” Currently there is no way to predict which patients will develop from immune-related adverse events (irAEs); thus, the strategy is to watch and wait after treatment initiation. “The immune-related effects can be so serious that 10 percent of patients will die,” she adds.
This stark side effect risk contrasts the promise checkpoint inhibitors have in some cancers, improving the prognosis for patients with several advanced cancers, including melanoma, renal cell carcinoma, non-small cell lung cancer, Hodgkin lymphoma, and head and neck cancer.
Immune therapy works by preventing the body from hiding a tumor. “But everyone’s immune system is different and not every tumor is hiding,” says Weidhaas. “Some people just need a little push to get their immune system in gear to recognize and attack a tumor while other people have an immune system that can quickly become over-productive.” The abundance of non-cancer-related autoimmune diseases are cases in point.
Weidhaas’s technique examined DNA signatures in 99 patients, looking for patterns that would indicate if inherited DNA biomarkers would predict toxicity. “If this signature is found in a patient it indicates that person has a 9-fold risk, or a 900% increased risk of toxicity from immune therapy,” she says. “And this risk will never change because it is something hard-wired in the DNA.
The team focused on mutations in control centers of non-coding RNAs, or mRNA binding sites. “How those spots work is important in terms of what a patient’s body will do in response to immune therapy,” she adds. These inherited germline mutations disrupt mRNA binding.
Weidhaas focused on germline mRNA pathway functional variants found in at least 5% of patients. “In this way we are catching most of the people that will have toxicity,” she says. “At least 10 percent of the people evaluated will fall into the high-risk category for serious toxicity.” And because these are germline mutations, toxicity to immune therapy does not vary between tumor type or if a patient responses to treatment or not. “Toxicity occurs in 25%–30% of patients supporting the hypothesis that toxicity is patient-specific and not tumor type dependent,” Weidhaas adds.
In earlier work, Weidhaas discovered that this same type of biomarker signature predicted toxicity to radiation therapy for prostate cancer. “This class of mutation is recently discovered, newly applied, and each time we find they are very predictive of important endpoints in cancer,” she says.
