Thomas Herzog, MD, traces the therapeutic evolution of frontline treatment for patients with ovarian cancer.
Thomas Herzog, MD
Thomas Herzog, MD
Clinical trials of intraperitoneal (IP) chemotherapy, hyperthermic intraperitoneal chemotherapy (HIPEC), and dose-dense regimens, while conflicting, have helped evolve the ovarian cancer field in the frontline setting, explained Thomas Herzog, MD.
As the field continues to evolve, physicians hope to augment these developments with a better understanding of biomarkers and patients’ genomic signatures.
The frontline setting most recently saw an advance with the June 2018 FDA approval of bevacizumab for use in combination with carboplatin and paclitaxel, followed by bevacizumab monotherapy, for the treatment of women with advanced ovarian cancer following initial surgical resection.
In an interview during the 2018 OncLive® State of the Science SummitTM on Ovarian Cancer, Herzog, professor of Obstetrics and Gynecology, deputy director, University of Cincinnati Cancer Institute, traced the therapeutic evolution of frontline treatment for patients with ovarian cancer.Herzog: We owe all the progress we've made in ovarian cancer to clinical trials. It was those clinical trials over a decade ago that allowed us to adopt paclitaxel into the care of women with ovarian cancer. That has significantly improved survival. Phase III trials have made a difference in taking a very cumbersome inpatient regimen and making it a very convenient outpatient regimen. Those trials allowed us to substitute cisplatin, which had a lot of toxicities with carboplatin, as well as take the paclitaxel infusion time from 24 hours to 3 hours. That's been extremely helpful.
If we then look at what else clinical trials have done, it was about trying to improve survival beyond that. Over a decade ago, our standard was carboplatin/paclitaxel every 3 weeks. Then people questioned whether giving maintenance therapy improved survival. SWOG-9701/GG-178 showed an improvement in progression-free survival (PFS). The endpoints were set up so that if there was a difference in PFS or overall survival (OS), they stopped the trial. In that case, we never had the OS data because of the number of crossovers and so forth.
GOG-212 was designed with a novel taxane, as well as regular paclitaxel, versus observation in patients who had a complete response after initial treatment. That trial did not show an advantage for giving 1 year of an extra taxane, regardless of whether it was the novel taxane or the traditional taxane. We took that off the table.
The next thing we looked at was this concept of dose-dense therapy. Could we give a higher dose over a shorter period of time and improve outcomes? Japanese researchers came out with a trial that showed an improvement in PFS in 2 widely read papers. Then, they had a final analysis on their OS that showed a significant improvement. We thought that we should be giving dose-dense chemotherapy, so most people moved to that.
Then GOG-262 was an attempt to somewhat mimic that trial. The problem was that we allowed bevacizumab to be used. It was given as an option, and over 80% of patients chose to be on it. Patients were randomized to dose-dense versus no dose-dense [chemotherapy], but they were able to choose whether they went on to bevacizumab. The overall results of that trial absolutely contradicted the results of the Japanese trial, which showed that dose-dense therapy was beneficial.
The small unplanned subgroup analysis showed that if patients did not have bevacizumab, there was an apparent improvement in outcomes if they had dose-dense therapy. That left everyone confused.
We then had the MITO-7 trial, which didn't show any difference in outcome. However, they fractionated the carboplatin, so people didn't know what to [make of those results]. Then, the ICON8 data came out, which showed that there was no difference; it was negative. Bevacizumab was not part of that. Looking at that, most people believed that dose-dense therapy was probably OK to give. Adding in the patient experience of having to come in on a weekly basis, the extra cost, and in many cases, more toxicity, it was decided that for most people dose-dense therapy is no longer as avant-garde as it was. We see more neurotoxicity, for example, which can be a long-term side effect.
People then looked at other ways of improving survival. The biggest area there has been IP chemotherapy. Three phase III trials showed an improvement in outcome with a 20% to 30% improvement in PFS or OS by giving some of the regimen directly into the peritoneal cavity through a catheter. The idea behind that was that you can give a much higher concentration and, therefore, achieve a more favorable pharmacokinetic result. People were impressed with the data; however, use never exceeded 15% in the United States. This was largely due to concerns about toxicity. It took a lot more time without additional reimbursement to compensate for that.
Medical oncologists struggle with what to do if there is a catheter problem. It's a problem even if you're the gynecologic oncologist who put the catheter in; however, it's an even bigger problem if you're not the person who put it in and have to rely on other services. It never had the uptake that you would expect. Then, GOG-252 was designed by researchers who were “champions” of IP chemotherapy. The primary endpoint was PFS; there was no difference in whether these patients had the chemotherapy administered directly into the peritoneal cavity or intravenously.
Bevacizumab was part of that protocol. The concern was whether bevacizumab confounded the results by allowing bevacizumab into the control arm. The other thought was that the dose used in GOG-172 was significantly less. The idea behind GOG-252 was to make it more tolerable in the real world. A lot of people have written small reports saying, “You can bring down the cisplatin dose significantly, you can drop the taxane dose, and you can substitute the taxane.” They said they’re achieving about the same results in GOG-172 from selected data but nonrandomized. Here we were in a randomized setting that showed [the purpose of] a randomized trial; it showed no difference.
Then, people championed HIPEC. A big HIPEC trial came out by Willemien J. van Driel, MD, PhD and was published in the New England Journal of Medicine in 2018. The trial touted an almost 1-year advantage in OS by giving neoadjuvant carboplatin/paclitaxel, cytoreductive surgery with or without HIPEC at a single dose of 100 mg/m2 with cisplatin. Those women had a little less than a 4-month advantage in PFS, which was the primary endpoint of the trial. Then they had immature data on OS that showed almost 1 year improvement. It was very impressive. The toxicity looked good. It was much less than what you see with other HIPEC trials. The question is, “Why?” If you look at the control arm, it underperformed significantly.
The thought was the control arm was going to be about 18 months for PFS, and it came out to be 10.7 months. If you look at the experimental arm in the [HIPEC trial] the PFS was 14.2 months. That also underperformed with what you would have expected. Are we looking at a type 1 error here or is there a real effect we don't know about? It completely contradicts the phase III HIPEC data that came out at the 2017 ASCO Annual Meeting, which showed no advantage with HIPEC and more toxicity. We need more randomized data in that area.
Biologics have probably been the only area where we have seen some successes. The biggest success story has been antiangiogenics. Of that, the biggest success has been bevacizumab where we now have a new approval in frontline ovarian cancer based on the data from GOG-218 and ICON7. Now we have an indication in the frontline, platinum-sensitive, and platinum- recurrent settings. The question is, “Where do you most leverage this? Can you use bevacizumab after bevacizumab?”
[The future will be] about combinations, biomarkers, and understanding how to bucket patients correctly. [This will be important in] minimizing toxicity and maximizing outcomes. Additionally, biomarkers, genomic profiles, and signatures will be important, so that we can put the patients who are going to benefit from angiogenesis inhibitors in one bucket, immuno-oncology drugs in another, PARP inhibitors in a third, and maybe some type of novel combination between those 3 or other agents as we move forward.I’ve had patients who have had plural effusions and large volume ascites. To me, it's a no-brainer because it does such a good job of getting those patients off of having to be tapped—–either through arthrocentesis or paracentesis. We also have some subgroup analysis from ICON7 and so forth that looks at patients with high-volume disease and high risk of relapse. Those patients have an improvement that's even greater than the other group and perhaps experience an OS advantage. Those are patients I strongly consider bevacizumab in, as well. Should it be used in all-comers? Many think so. It's a change in behavior for many of us because insurance is now likely able to pick that up. It's an evolving landscape right now, but you'll probably see more use moving forward.We don't know what's behind the curtain, but a press release on the SOLO-1 trial said it was positive. It's important to know that SOLO-1 had an integral biomarker: BRCA-positive. We would like to put our best therapies upfront to enhance the chance for cure and not just disease control. It makes all the sense in the world. We have to see those data, but it seems that, at least for those patients, we’ll see PARP inhibitors in that setting. Likely those with homologous recombination deficiency will benefit almost as much as those with platinum-sensitive disease in the maintenance trials. I'm not sure about all-comers.The whole concept of angiogenesis with PARP inhibitors or angiogenesis with immunotherapy, in terms of synergy, is exciting. Preclinical evidence has been compelling in demonstrating that PARP inhibitors can increase the immunogenic load and make immunotherapy more effective. It will be exciting to see if we can leverage those types of drugs together.Clinical trials remain our platform. Even though I spoke about a number of trials that haven't moved the needle forward, it's a process. If we look at all of the things that have moved the needle forward, it's always been due to clinical trials. We need to keep patients enrolling on these trials to get the best care and the best outcomes for not just one but thousands of patients at a time.