Test predicts response to postoperative radiotherapy in prostate cancer

Researchers have developed a 24-gene molecular signature that predicts how patients with prostate cancer will respond to radiation therapy after radical prostatectomy.

The test — called PORTOS (GenomeDx Biosciences), the acronym for which stands for Post-Operative Radiation Therapy Outcomes Score — is designed to determine which patients likely will respond to postoperative radiation with a significantly reduced risk for metastatic recurrence at 10 years.

Felix Y. Feng, MD, associate professor of radiation oncology, urology and medicine at the University of California, San Francisco, and colleagues retrospectively analyzed pooled data from five published U.S. studies of patients with prostate adenocarcinoma who had radical prostatectomy, with or without postoperative radiotherapy. All patients underwent tumor gene-expression analysis, and they had long-term follow-up and complete clinicopathological data available.

Patients who underwent postoperative radiotherapy were matched with patients who did not undergo radiotherapy based on several factors, including Gleason score, PSA concentration, surgical margin status, lymph node invasion and receipt of androgen deprivation therapy.

Researchers created a matched training cohort of patients from one study, in which they developed a 24-gene PORTOS score. They then used patients from the other four studies to create a pooled matched validation cohort.

Development of distant metastasis served as the primary outcome.

The findings, published in The Lancet Oncology, showed that patients who underwent postoperative radiotherapy and had a higher PORTOS were less likely to develop metastasis at 10 years than those who did not undergo radiotherapy.

The findings suggest postoperative radiotherapy should be considered in this subset of patients, researchers concluded.

PAGE BREAK

HemOnc Today spoke with Feng about the findings and their potential impact on cancer care.

Question: How did the idea for a test like this come about?

Answer: Many of the patient referrals I receive are patients who have had surgery, and I see them in the context of trying to help them decide whether they should receive radiation after surgery. There actually are three clinical trials that have randomly assigned patients to radiation after surgery for prostate cancer. In all trials, a benefit was achieved in PSA recurrence. However, no trials have showed a benefit in metastasis or survival. We wanted to create a genomic test that potentially would help make that decision.

Q: How was the test developed?

A: We took two different, large patient sets that had undergone surgery for high-risk prostate cancer and we matched them based upon whether they had received radiation after surgery. When we did this matching, we accounted for standard clinical pathologic features. We also profiled the surgical samples from all patients for expression of the vast majority of known genes in the genome. We used these first two sets to develop a general classifier of response. In the second two sets, we validated the performance of our genomic classifier. The purpose of our genomic classifier is to try to more specifically predict which patients would benefit from radiation after surgery.

Q: Have there been any challenges or disadvantages with the test?

A: There are some challenges. When any of us try to develop a biomarker for clinical use, you want to create a discovery set and a validation set. Nowadays, many patients show up in my clinic and ask me if PORTOS is ready for prime time use. What we need to answer is whether we believe the genomic test is enough to change clinical practice. There actually are a number of collaborators that contributed to the making of this test who all have different degrees to which they have embraced it into clinical practice. Some are using the test to help with their clinical decision-making, yet others want to see more validation in prospective data sets before they change their decision-making. I think it comes down to the fact that, with any new technology, there are doctors who will use PORTOS right away and there are those who will not. One of the challenges is to identify how much more data we need in order for everyone to be comfortable using this.

Q: The test was specifically tested in patients with prostate cancer. Is there a possibility that the test could work in other patient populations?

A: This is a really good question. The test basically predicts radiation response, so it is likely that it could work in other cancer types. However, at the same time, this is speculative and would need to be evaluated.

Q: Do you have plans for further research?

A: Absolutely. We have actually used similar approaches to develop a postoperative classifier that uses similar approaches to identify who benefits from hormone therapy after surgery. Our results will be presented in February at the Genitourinary Cancers Symposium.

Q: Is there anything else you would like to mention?

A: This is really a groundbreaking test in the sense that genomics have not been previously used to predict who should get radiation and who should not get radiation in any disease site. Part of the reason why we were able to assess the test in prostate cancer is because of the large number of specimens available. I think the era of genomics guiding who will get radiation is beginning. Our paper is the first of many papers that will improve decision-making in the context of radiation therapy for cancer. Studies like ours are definitely needed to help personalize therapy for patients with cancer. – by Jennifer Southall

Reference:

Feng F, et al. Lancet Oncol. 2016;doi:10.1016/S1470-2045(16)30491-0.

For more information:

Felix Y. Feng, MD, can be reached at University of California, San Francisco, Box 3110, Room 450, 1450 3rd St., San Francisco, CA 94158; email: felix.feng@ucsf.edu.

Disclosure: Feng reports no relevant financial disclosures.