From the Journals

A case is building for personalized, genome-based radiation dosing



A team of researchers from the Cleveland Clinic, the Moffitt Cancer Center in Tampa, and Case Western Reserve University in Cleveland is zeroing in on a way to personalize radiation therapy for cancer patients based on genomic profile, much as genomics is used to tailor oncologic drug therapy.

It’s called “genomic-adjusted radiation dose” (GARD), a dose tailored to a person’s radiosensitivity as determined by the expression of 10 genes, known as the radiosensitivity index (RSI), combined with a linear quadratic model to yield GARD, a prediction of risk and benefit at various radiation doses for a particular patient.

A recent report in The Lancet Oncology validated GARD in 1,615 patients with seven cancer types from 11 study cohorts. If it holds up in clinical trials set to start later this year, GARD should “allow us to predict the benefit of radiation for an individual patient and adjust their treatment strategy,” wrote the authors of an editorial that accompanied the study. “The efforts need to be applauded worldwide, because radiotherapy is considerably lagging, compared with the enormous progress done in the field of personalized medicine,” Orit Kaidar-Person, MD, a radiation oncologist at Sheba Medical Center in Ramat Gan, Israel, and colleagues wrote.

GARD was associated with time to first recurrence and overall survival for patients receiving radiotherapy and predicted radiotherapy benefit, while physical dose did not. The team found a relative 2% reduction in risk of first recurrence for each unit increase of GARD (P = .0017) and a relative 3% increase in overall survival for each unit increase in GARD (P = .0007), among those who got radiotherapy. Values of GARD run from 0 to over 100, with higher scores meaning more radiation benefit.

The radiosensitivity index, which was derived from genomic studies of cancer cell lines exposed to radiation, was previously validated by the team and other groups across several tumor types.

Currently, radiation dosing is generally uniform for a given disease site and stage, based on the assumption that a given dose of radiation results in the same clinical effect across patients. In fact, the biological effect of a given dose varies widely between individual patients. “Patients we treat uniformly do not have a uniform response” which is why a more personalized approach would help, said lead investigator and Cleveland Clinic radiation oncologist Jacob Scott, MD, DPhil.

One patient with a given tumor might benefit from 2 extra fractions, while the next might need an extra 15 for the same benefit. “You need to know about [a patient’s] tumor genomics to know how hard you have to work,” he said.

Dr. Scott and colleagues are working with a genomics company to commercialize the approach. The vision for now is that physicians would ship in biopsy samples to be analyzed; RSI and GARD would be calculated, and then a decision support report would be sent back to the treatment team outlining the risks and benefits of various doses for the patient.

Dr. Scott, who holds proprietary rights on the approach, is bullish. When asked if he anticipates GARD dosing to be standard of care in 10 years, he said that “I can’t imagine another world. Everything else in cancer is personalized. Why aren’t we? It just makes sense. I know there’s a better way” to prescribe radiation, “and I’m excited for the future when I can use it.”

When asked for comment, Brian Marples, PhD, a radiation oncology professor at the University of Rochester (N.Y.), said the data so far for GARD “seem very solid. I’m very excited by the concept.”

It’s been “the holy grail” of radiation researchers to find a biologic marker that predicts what dosages patients need and what can be given safely. “This strategy is a good way of doing that. Other groups are proposing similar strategies, but I think this group is ahead. I can see [GARD] being readily applied to the clinic because patients are [already] getting their tumors genomically characterized as part of care,” Dr. Marples said.

But many questions remain. For instance, the editorial writers questioned how GARD is “affected by tumor heterogeneity, response to systemic therapy, and changes in the tumor microenvironment.” Also, the approach is based on conventional 2 Gy fractions, but other fractionation regimens are becoming more common.

For Dr. Marples, the big caveat is that most cancer patients are treated with both radiation and chemotherapy. He said he would like to see GARD validated in patients who receive both.

They seven tumor types in the study included breast cancer, head and neck cancer, non–small cell lung cancer, pancreatic cancer, endometrial cancer, melanoma, and glioma. The majority of the subjects were treated with radiation, and each had the genomic data needed to calculate GARD.

Dr. Scott, senior author and Moffitt Center radiation oncologist Javier Torres-Roca, MD, and a third author hold intellectual property rights on RSI, GARD, and prescription dose base on RSI, plus equity in Cvergenx, a company that seeks to commercialize the approach. Dr. Torres-Roca and another author are cofounders. The editorial writers and Dr. Marples did not have any relevant disclosures.

Next Article: