MILAN-Biomarkers predicting radiotherapy toxicity were reported by the REQUITE project (Validating Predictive Models and Biomarkers of Radiotherapy Toxicity to Reduce Side-Effects and Improve Quality of Life in Cancer Survivors) at the 2019 European Society for Radiotherapy and Oncology (ESTRO) conference (Abstract OC-0647).
"Prediction of radiotherapy toxicity is advancing very fast. Quite soon we'll be in a position where we can give a personalized prediction [of] how every patient [will] respond to radiation," said Christopher Talbot, PhD, Senior Lecturer in Medical Genetics at the University of Leicester, U.K.
Talbot told Oncology Times how genetic and other biomarkers being assessed by the REQUITE project, together with clinical and other aspects of cancer treatment, could be combined in algorithms computing radiation toxicities to guide therapy decision-making and give each patient the least toxicity consistent with effective therapy.
"There are parallel fields working on chemotherapy and other types of treatment. So we'll be able to personalize the cancer treatment pathway of every patient," he said.
Patients were enrolled from 26 centers in eight countries, with follow-up collected for 2 years ending in September 2018. The primary endpoints were change in breast appearance at 24 months (for patients with breast cancer), rectal bleeding at 24 months (in prostate cancer), and breathlessness at 12 months (for patients with lung cancer). Altogether 4,442 patients were followed up and analyzed for toxicity, clinical parameters, patient-reported outcomes, and predictive biomarkers.
Toxicity Biomarkers
All patients who completed the study were genotyped to identify single nucleotide polymorphisms thought to be relevant to toxicity. Blood samples were tested for radiation-induced lymphocyte apoptosis (RILA) to check for the percentage of apoptosis in cytotoxic T lymphocytes after being irradiated. The RILA test was found to predict acute breast pain and fibrosis in patients with breast cancer and urinary incontinence in those with prostate cancer.
"The RILA assay seems to predict a doubling of risk for [breast] fibrosis, which is significant," said Talbot, noting this could directly affect treatment decisions. "An obvious possibility in breast cancer is [that] patients who are predicted to have a really bad adverse reaction [to radiotherapy], that can severely affect their quality of life, could be offered mastectomy instead of mammoplasty and radiotherapy."
In the case of patients with prostate cancer, there were other possibilities. "One possibility is the insertion of a rectal spacer which increases the space between the rectum and the prostate and, therefore, reduces the amount of damage to the rectum [from] radiotherapy," Talbot said.
It was clear from the REQUITE findings that simple but effective changes to treatment planning were already becoming possible with reference to the various biomarkers they had validated, said Talbot. "We're doing a genome-wide association study to find single nucleotide polymorphisms that associate with toxicity. And they could be combined with the RILA assay."
Creating an Algorithm
But what Talbot described as the "real gold" was to generate an algorithm to include these biomarkers with clinical factors.
"We know that diabetic patients have a higher rate of radiotherapy-related toxicity," he said. Combining such data with established treatment-related toxicities such as radiation dose and field configuration could predict toxicity even more accurately. "All of those could be combined in an algorithm-potentially including the RILA assay, the genetic assay, and circadian rhythm effects [which his group was also investigating]-to come up with the risk of radiotherapy toxicity [for] an individual tumor."
Talbot was confident that patients and their doctors could jointly adapt their treatment plans and reduce side effects. He said that, in the case of toxicities affected by circadian rhythms, for example, there was good potential for reducing side effects by the simplest of maneuvers based on established factors determined by circadian rhythm genotypes.
"The exciting thing about circadian rhythm data is it provides a very easy way of modifying treatment. If we could guide [patients] and suggest morning or afternoon [treatment], it provides a 'free' way of modifying treatment and hopefully reducing side effects without any real clinical change," explained Talbot.
Commenting on the REQUITE findings, Pierfrancesco Franco, MD, PhD, from University of Turin, Italy, and Chair of the ESTRO Young Committee, described the study as an elegant demonstration of how translational research could help clinicians.
"[It] can help oncologists offer patients a personalized approach during the clinical decision-making process that balances the need to control the tumor with the need to minimize the side effects on normal tissues and optimize radiotherapy treatment," he stated.
Peter M. Goodwin is a contributing writer.