Disclosures: Scott reports having intellectual property related to this study (GARD). He also is a consultant for and shareholder in Cvergenx, which is working to make this test available to physicians. Please see the study for all other authors’ relevant financial disclosures.
October 22, 2021
5 min read
Save

Personalized approach to radiation therapy predicts efficacy across cancer types

Disclosures: Scott reports having intellectual property related to this study (GARD). He also is a consultant for and shareholder in Cvergenx, which is working to make this test available to physicians. Please see the study for all other authors’ relevant financial disclosures.
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

A personalized approach to radiation therapy can maximize therapeutic effect across multiple cancer types, according to findings published in The Lancet Oncology.

The genomic-adjusted radiation dose (GARD) — shown to be significantly associated with time to first recurrence and OS — moves away from the traditional one-size-fits-all radiation dosing paradigm and can better predict benefit of radiotherapy, researchers concluded.

Woman receiving radiation therapy.
Source: Adobe Stock.

“The way we currently prescribe radiation is based on rigorous, empiric clinical trials, and requires each patient with a given disease to be prescribed a standard dose. This standard dose is what has been found to benefit the group the most, but it does not necessarily benefit each individual patient optimally,” Jacob G. Scott, MD, DPhil, physician-scientist in the department of translational hematology and oncology research at Cleveland Clinic, told Healio. “Under this paradigm, dose escalation to large unselected cohorts has not been shown to benefit patients.”

However, in analyzing a large cohort, Scott and colleagues demonstrated that considering an individual patient’s tumor genomics yielded information about who would benefit from more dose and who would not, validating their hypothesis that genomics can be integrated into radiation prescription decisions.

Jacob G. Scott, MD, DPhil
Jacob G. Scott

“We’re suggesting the way we prescribe radiation should fundamentally change,” Scott said. “We should start thinking about prescribing radiation based on the effect it has on a patient, not just on the dose delivered. And right now, the only way to predict this is by using units of GARD, which we’ve shown [is associated] with OS and recurrence, while changes in the physical dose are not.”

Scott and colleagues — representing Cleveland Clinic, Case Western Reserve University School of Medicine and Moffitt Cancer Center — conducted a pooled pan-cancer analysis of previously published data from 11 cohorts that included a combined 1,615 patients with one of seven malignancies: breast cancer, head and neck cancer, non-small cell lung cancer, pancreatic cancer, endometrial cancer, melanoma and glioma.

Researchers assessed 1,298 patients (radiotherapy, n = 982; no radiotherapy, n = 316) for first time to recurrence. They assessed 677 patients (radiotherapy, n = 424; no radiotherapy, n = 253) for OS.

Results showed GARD as a continuous variable was associated with time to first recurrence (HR = 0.98; 95% CI, 0.97-0.99) and OS (HR = 0.97; 95% CI, 0.95-0.99), meaning that for each unit increase in GARD, there was a concomitant, statistically significant benefit to outcome.

Interaction tests using the Wald statistic showed the effect of GARD on OS depended on whether a patient received radiotherapy (P = .011). The interaction test for GARD and radiotherapy did not appear significant for time to first recurrence.

Results showed no significant association between physical dose of radiation and time to first recurrence or OS.

Healio spoke with Scott about the GARD approach and how the findings could guide practice.

Healio: Can you describe the rationale for this study?

Scott: In 2009, my co-author — Javier F. Torres-Roca, MD, senior member in the department of radiation oncology at Moffitt Cancer Center and professor of oncologic sciences in the University of South Florida Morsani College of Medicine — came up with a gene signature called the radiosensitivity index (RSI). This did a good job separating patients into two groups based on whether they were sensitive or resistant to radiation. Although this was interesting, it wasn’t clinically useful in that it didn’t really give a treating physician an answer for how to adjust their treatment in response to the test.

In 2016, Dr. Torres-Roca and I came up with the concept of GARD, which incorporates RSI with a mathematical model to not just understand whether patients are sensitive or resistant, but how they will respond to a specific dose of radiation. In a paper published in 2017 in The Lancet Oncology, we showed this relationship matched our clinical experience across disease sites and that there were differences within various cohorts that could explain the heterogeneity in response.

Our most recent paper takes a meta-analysis-type approach to determine whether these observations we made in 2017 about the effect of radiation for a given patient translate into clinical outcomes. Our new results show that they do — ie, that GARD predicts the clinical benefit of radiation therapy.

Healio: Can you explain why physical radiation therapy dose is not the optimal approach for predicting therapeutic effect?

Scott: The way we administer radiation therapy, there is personalization anatomically — which is to say that we can shape the dose based on each patient’s anatomy based on 3D imaging from CT scans and also MRIs or PET. However, the amount of radiation we prescribe — the physical dose of radiation — is, for the most part, the same for each patient. That approach is based on decades of clinical trials and clinical experience. We know the doses we’re giving are safe and effective for populations of patients with a given cancer, but they don’t take into consideration the important differences between those patients.

Every radiation oncologist knows that all patients respond differently to radiation, but we have been blinded to where each patient lies on this distribution of response. Some who are extremely sensitive to radiation may only need a small dose, whereas others who are super-resistant may need a much higher dose. Unable to see those differences — whether in clinical trials or everyday practice — we have had no choice but to administer the same dose to everyone. Similar to when we treated patients before 3D imaging, we had to make assumptions about field sizes that were most likely to encompass most patient’s tumors. Also, like the era of 3D planning, once we can ‘see’ the difference between patients when it comes to radiation benefit per unit, it makes sense that outcomes can be improved through rational prescription modification.

Healio: So GARD, which uses the science behind genomics, allows you to understand and take advantage of those differences between patients to better measure the effect?

Scott: Exactly. No physician would guess that every man’s prostate is the same shape or in the same exact location. Before the CT scan was invented, we had to deliver a big volume of radiation to ensure the prostate was included. If you do that, you’re hitting an enormous extra area beyond the prostate. The CT scan allows us to see exactly where the prostate is and provides a new dimension for treatment planning. GARD is a similar concept in that we can now see one patient doesn’t need as much radiation as another patient.

Healio: Is this technology ready for widespread adoption?

Scott: That is a great question. I am involved in many conversations with people who want to assess this in the context of clinical trials. However, in this most recent paper, we established GARD as level one evidence from archival tissue studies. For several malignancies — including breast, lung or prostate cancers — we already have safe dose ranges for treating physicians to use. We believe we can use this right now as a clinical decision support tool to adjust dose within those established safe ranges. With a few exceptions, there aren’t rigorous rationales for giving one dose vs. another. With our findings, there now is.

Healio: What do you consider the biggest implication of this research?

Scott: This is the first genomics-based advance that allows us to do our craft better, and it is proof we’re not doing as well as we could be. We are not practicing personalized biological medicine, and we’re not taking advantage of the genomics era. We’re lagging behind our counterparts in medical oncology. This advance clearly allows for us to do better with the technology available to us, without the need for costly novel radiation delivery methods or new drugs. This approach uses a genomic test that will be available commercially soon. We just have to learn to use this new tool.

Healio: What are the next steps?

Scott: I’m working to get this incorporated into clinical trials at my institution. We also are working with a large genomics company to make this commercially available, and we’re hoping that will happen either later this year or early next year. We’re hoping treating physicians are going to be able to order this test very soon from a commercial, trusted genomics vendor.

For more information:

Jacob G. Scott, MD, DPhil, can be reached at Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, 10201 Carnegie Ave., Cleveland, OH 44195; email: scottj10@ccf.org.

References:

Scott JG, et al. Lancet Oncol. 2021;doi:10.1016/S1470-2045(21)00347-8.
Scott JG, et al. Lancet Oncol. 2016;doi:10.1016/S1470-2045(16)3064809
.