Biomarkers for Early Detection Essential to Major Progress in Ovarian Cancer

Robert C. Bast, Jr, MD

While surgical advances and new treatment approvals, including olaparib (Lynparza) and bevacizumab (Avastin), have improved the outlook in ovarian cancer, late diagnosis and tumor resistance remain as significant barriers to major progress in treating the disease, according to Robert C. Bast, Jr, MD, vice president of Translational Research at the University of Texas MD Anderson Cancer Center.

In an interview with OncLive, Bast discussed his research using CA-125 levels and p53-autoantibodies as biomarkers for early detection of ovarian cancer, and where he envisions the treatment paradigm for the disease moving next.

OncLive: What is the biggest challenge in the treatment of ovarian cancer?

Dr Bast: Over the last two decades, the 5-year survival rate for patients with ovarian cancer has improved significantly. However, we currently are curing no more than 30% of long-term ovarian cancers. That is related both to the persistence of drug-resistant cancer cells, but also to late diagnosis. If you can detect ovarian cancer in stage I or II, when it is still limited to the ovaries or the pelvis, you can cure 70% to 90% of patients with available surgery and chemotherapy. When the disease has spread to the rest of adnominal cavity or farther, as is the case in stage III or IV, the cure rate slips to less than 20%.

At present time, we are only detecting one-fourth of patients with stage I or II disease. One of the goals over the last 20 years has been to develop better early detection. There is currently a trial at MD Anderson where we are detecting stage I or II disease through CA-125 testing each year, which measures the amount of cancer antigen in the blood. When that goes up, we do an ultrasound or another imaging test and if that appears abnormal, then we do surgery.

There is a larger trial, the United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS), in which 200,000 women are being studied. Of that 200,000, 50,000 are getting CA-125 testing each year and an ultrasound if their levels go up. That trial will turn out later this year and we are very much excited to find out if that test improves survival by improving early detection. At present time, only about 80% of ovarian cancers express CA-125, so we know we will miss 20% with this test.

Is there another biomarker that is more effective that CA-125?

Over the years, our lab has tested over 100 biomarkers to see if we can improve on CA-125 with a combination of different tests. We finally found a biomarker, called anti-p53 autoantibodies, that appears to detect at least a fraction of those patients we missed with CA-125 and detect it at an earlier stage than we can with CA-125 alone. P53 is mutated and consequently often overexpressed in more than half of ovarian cancers and virtually all of the high-grade serous ovarian cancers.

We’ve found that about 25% of patients at the time of diagnosis for both early and late-stage cancers will have autoantibodies. Other people have found that before, but what is different about our study is that we are using blood samples from the UKCTOCS. We can go back and see if these antibodies were elevated before the detection of the cancer or prior to the elevation of CA-125. We found that about 25% of the patients in the UKCTOCS trial will have elevated autoantibodies, whether they are CA-125 positive or negative. With this method, we can detect the cancer anywhere from 13 to 33 months before the elevation of CA-125 or the detection of cancer in patients that are CA-125 negative. Currently, we are looking at other antibodies. We have about 200 other candidates. We will have to find out if any of them will be as good as anti-p53 autoantibodies.

How far are we from applying these findings to oncology practice?

At present time there is not a screening test that is approved for ovarian cancer for women at conventional risk for the disease. If the UKCTOCS trial is strongly positive, a community oncologist may consider using a rising CA-125 to trigger an ultrasound and detect early cancer. However, we need to first find out how that trial is going to turn out. Even if that trial is positive, we can do better. Using autoantibodies may improve the screening with CA-125 even further.

How has the treatment of ovarian cancer changed over the past year?

Care of patients with ovarian cancer really has evolved. For the last 15 years, a combination of carboplatin and paclitaxel after cytoreductive surgery has been the standard of care. We have tried to periodically add other conventional drugs and it has not improved outcomes. At the present time, bevacizumab has improved progression-free survival, but has only improved overall survival in a subset of patients. Bevacizumab is probably the major advance in community care after carboplatin and paclitaxel.

One of the most exciting things in the last few years has been the focus on better surgery. We have known for a while that taking out as much of the tumor as possible improves outcomes for patients, probably by influencing the effect of the chemotherapy. As we look retrospectively at different studies, we have found that if the disease can be completely resected, the results will be better. This is the case for any solid tumor. Our usual resection rate is about 30%, and we can identify a subset of patients with a 90% resection rate.

For the individuals who are not resected upfront, three cycles of chemotherapy reduce the disease enough that, post-operation, an 80% complete resection rate can be achieved. In other studies, upfront complete resection is correlated dramatically with improved outcomes. We don’t know if three cycles of chemotherapy followed by resection will have results as good as upfront resection; however, with this strategy, the tumor can be completely removed for 80% of an individual. That is very promising.

Where do you see treatment going in the future?

About half of high-grade serous ovarian cancers have the BRCA mutation. Only about 20% have germline or somatic mutations of BRCA1 or BRCA2. There are problems with homologous repair in about 30%. Tests are being done to identify those individuals. Olaparib [which is approved for advanced ovarian cancer] and other PARP inhibitors, in combination with other targeted therapies, are currently being evaluated at different centers including MD Anderson.

The other possibility is PI3 kinase activation. About half of high-grade serous ovarian cancers have abnormalities in activation of PI3 kinase. There are combinations of PI3 kinase pathway inhibitors that are being investigated. There are severe toxicities from those combinations at this point, so we are going to have to determine how we can improve the therapy index going forward.