Research on a new way to match the right radiotherapy dose with the right patient

Summary

This research created and tested a new method, called genomic-adjusted radiation dose (GARD), to predict how much radiation a person should get to treat their cancer based on their tumor’s genomic information. Researchers tested GARD to see if it could predict how well radiotherapy will work to shrink a tumor (response to radiotherapy).

  • Researchers created GARD to adjust a patient’s usual radiotherapy dose based on their genetic information
  • Radiotherapy, also called radiation therapy, is a type of cancer treatment that uses high energy (radiation) to destroy cancer cells and shrink tumors
  • Researchers tested GARD scores with 5 different cancer datasets

What were the main results?

The researchers found that GARD could predict how a patient's cancer responds to radiotherapy. They found that patients with breast, lung, brain, or pancreatic cancer who had higher GARD scores had a better response to radiotherapy.

The researchers believe GARD can help match the dose of radiotherapy to a patient’s specific tumor. They suggest more research is needed to test how well GARD scores can predict a tumor’s response to radiotherapy, relapse, and patient survival.

    • Illustration of a magnifying glass representing researchers looking closely at their study.

      What did researchers study?

      Researchers studied genomic data from patients’ tumor samples and medical records. First, they calculated GARD scores for tumor samples from different cancer types. Then, they compared GARD scores for different kinds of cancer and response to radiotherapy.

    • Illustration of two people. One has their arm raised representing volunteering.

      How many patients' data did researchers study?

      Researchers began with 8,271 tumor samples from patients who joined Total Cancer Care® (TCC). Then, researchers tested GARD scores using data from 538 patients with different cancer types from 5 datasets.

    • Illustration of a bar graph with a magnifying glass representing types of research.

      What kind of research was this?

      This was an observational study. Researchers looked at (observed) data from patients and their tumor samples to predict their response to radiation.

The results of this research alone should not be used to make health decisions. It takes many studies for researchers to confirm findings and use them in patient care.

Why was this research needed?

Researchers are looking for a way to tailor radiotherapy to a patient’s specific tumor. Radiotherapy, also called radiation therapy, is a type of cancer treatment that uses high energy (radiation) to destroy cancer cells and shrink tumors. In the U.S., 2 out of 3 patients with cancer get radiotherapy as part of their treatment. It’s estimated that radiotherapy is responsible for 40% of all cancer cures.

While other cancer treatments, such as chemotherapy, have focused on tailoring treatment based on a patient’s genes, radiotherapy is currently given as a one-size-fits-all treatment. Doctors use different but standard (usual) amounts of radiation for patients based on the type and size of the cancer. If lower doses of radiation can still work to shrink a patient’s tumor, this can help limit unwanted side effects during treatment and late effects after completion of treatment.

Previous research found 10 genes in tumors that can affect how well radiotherapy will work to treat a patient’s cancer. The researchers then created a statistical method that could predict how a cancer responds to radiotherapy based on these genes. Researchers believe the benefit of radiotherapy may be highest for patients with certain genomic profiles.

In this study, researchers built on the previous research to create and test a new statistical method, called genomic-adjusted radiation dose (GARD), to predict how much radiation a person should get to treat their cancer based on their tumor’s genomic information. The researchers tested GARD to see if it could predict how well patients’ tumors responded to radiotherapy based on the dose (amount) of radiation used to treat their cancer.

How many patient's data did researchers study?

Researchers began with 8,271 tumor samples from patients who donated their data to Total Cancer Care® (TCC). The samples were from:

    • 20 parts of the body
    • Patients who had been treated between 2010 and 2012 with standard radiotherapy doses

Then, researchers tested GARD scores using data from 538 patients with different cancer types from 5 datasets:

    • 77 patients with breast cancer in the Karolinska University Hospital Radiumhemmet Cohort – all patients had surgery and radiotherapy from 1994 to 1996
    • 263 patients with breast cancer in the Erasmus Breast Cancer Cohort - all patients had surgery and radiotherapy
    • 60 patients with lung cancer in the Moffitt Lung Cancer Cohort – all patients had surgery and radiotherapy
    • 40 patients with pancreatic cancer from the Moffitt Pancreas Cancer Cohort – all patients had surgery and radiotherapy
    • 98 patients with brain cancer (glioblastoma) from The Cancer Genome Atlas Glioblastoma Patient Cohort – all patients had radiotherapy and chemotherapy

  • What kind of research was this?

    This was an observational study. Researchers looked at (observed) data from tumor samples donated by patients to predict response to radiation.

    This kind of study can further cancer research aimed at developing new treatments, tailoring treatments to patients, and identifying a patient’s cancer sooner.

  • What happened during this research?

    Researchers created GARD

    Researchers created GARD by taking the standard method of calculating radiotherapy dose and factoring in data from 10 tumor genes that can predict how well radiotherapy may work for a patient.

    Researchers compared GARD scores across different cancers using tumor samples

    The researchers calculated GARD scores for the 8,271 tumor samples from TCC. Then, they assigned tumor samples to a low, middle, or high GARD score based the total amount
    of radiation used to treat the tumor. They also compared GARD scores for cancers from different parts of the body.

    Researchers tested if GARD scores could predict radiotherapy effects using the 5 datasets

    Then, researchers tested if the GARD score could predict response to radiotherapy using the 5 datasets. They compared low, medium, and high GARD scores and:

    • Types of cancer
    • Dose and type of radiotherapy
    • Response to radiotherapy, if a patient's cancer came back, and how long they lived

  • How researchers designed this study

    Researchers created GARD, and then calculated GARD scores for 8,271 tumor patient samples from Total Cancer Care (TCC). Researchers compared GARD scores for cancers from 20 different parts of the body.

    Researchers then calculated GARD scores for 538 patients’ genomic data from 5 datasets. Researchers compared GARD scores with responses to radiotherapy for patients with:

    • Breast cancer
    • Brain cancer
    • Lung cancer
    • Pancreatic cancer


    See infographic

  • What were the main results?

    When researchers compared GARD scores across different cancers using tumor samples

    Researchers found GARD scores had a large range of scores from 1.66 to 172.4. This was true even when tumor samples got the same dose of radiotherapy. This shows that patients treated with a higher radiotherapy dose didn’t always have a higher GARD score or a better response to radiotherapy.

    GARD scores also identified different types of cancers:

    • Brain (gliomas) and bone (sarcomas) cancers had the lowest GARD scores
    • Cervical cancer and throat (oropharyngeal head and neck) cancer had the highest GARD scores

    When researchers tested if GARD scores could predict radiotherapy effects using the 5 datasets

    The researchers found that GARD predicted the effects of radiotherapy for patients with breast, lung, brain, and pancreatic cancers. They found that patients with higher GARD scores had a better response to radiotherapy.

    Patients with breast cancer and higher GARD scores were more likely to survive over 5 years without their cancer spreading compared to patients with lower GARD scores.

  • How has this research helped?

    The researchers suggest that GARD could be used in the future to tailor radiation based on the genomic features of a patient’s tumor. The researchers believe GARD can help match the dose of radiotherapy to a patient’s specific tumor to give the right dose to the right patient. They suggest more research is needed to test if tumors with higher GARD scores have better responses to radiotherapy.