Stanley B. Kaye, MD: Hello. I am Stan Kaye, Professor of Medical Oncology at the Royal Marsden Hospital at the University of London. Welcome to this edition of Medscape Oncology Insights on gynecologic cancer.
Today we are going to be discussing presentations from the 2014 annual meeting of the American Society of Clinical Oncology (ASCO). Joining me today is Dr. Rob Coleman, Professor and Vice Chair of the Department of Gynecologic Oncology and Reproductive Medicine at the MD Anderson Cancer Center, in Houston, Texas.
Before we start, we want to send our best regards to our colleague Maurie Markman, who had to miss his first ASCO in 35 years and is on the mend. Maurie, we will soldier on until you are back in the ranks.
It has been a very good meeting for gynecologic cancer. What presentations have impressed you most?
PARP Inhibitors Back on the Agenda
Robert L. Coleman, MD: Several different presentations have been very interesting. The oral presentations were very exciting, particularly in looking at the combination of poly-ADP-ribose polymerase (PARP) inhibitors and angiogenesis inhibitors. Some neat biomarker work has been done in trying to subclassify high-grade serous ovarian cancers into groups of patients who might respond to the agent.
We also saw some new drugs on the scene that are promising for future development; at least some have shown some signal of activity that we would want to pursue potentially going forward.
Dr. Kaye: The PARP inhibitors were the big topic last year and still are. What I take as a step forward, however, is the combination of PARP inhibitors with other agents. The combination of olaparib with cediranib, for example, which has finally been accepted on the basis of the earlier ICON6 study,[1] is a drug that clearly has activity as a vascular endothelial growth factor (VEGF) inhibitor in the ovary. It was a positive trial.[2] There was a substantial difference in patients with platinum-sensitive disease. Because it was not placebo-controlled, what might you think the criticisms might be, especially given that it was so positive?
Dr. Coleman: You have hit on the major criticism. When you are looking at such short-term endpoints as response rate and progression-free survival, there is potential for bias. It was brought up in discussion that differential toxicities potentially could be important at the determined time points when they assess response and progression.
With any of these trials, a larger sample size and better identification of the population are needed. For example, we know the BRCA status of some of the patients. It seemed to me that about half of the patients were known to be mutation carriers, and the others were either negative or unknown. The investigators saw a differential effect in these 2 subpopulations.
The primary thrust of these studies is their endpoint, but many more questions are raised. One question would be, “What is the best way going forward, given those subgroup analyses?” We tend to overinterpret those. But it is interesting. Hopefully, some future work will tease that out a little more.
Dr. Kaye: It was a striking difference in that the patients who had BRCA wild type had a substantial extra benefit from the cediranib. This finding could be part of the equation in future studies, both to compare that nonchemotherapy combination (olaparib with cediranib) with chemotherapy in platinum-sensitive disease and to determine what could be a slightly more complicated maintenance strategy for patients with BRCA-positive disease.
So, we have heard a lot about PARP inhibitors, although there are other PARP inhibitors. Olaparib is probably in the lead at the moment, isn’t it?
Dr. Coleman: Olaparib has probably the biggest dataset, not only for efficacy but also for toxicity. Returning to your point about moving forward, the therapeutic index is important for patients who have recurrent disease. The combination adds toxicity. You have done a great job in synthesizing the data, looking at chemotoxicity vs PARP inhibitor toxicities. Certainly there are some differential effects. But when you add a drug that has a side effect to a drug that generally doesn’t have a tremendous amount of side effects, trying to balance that against chemotherapy is difficult. And the cediranib dose is higher.
Finding the ‘Sweet Spot’ for Bevacizumab Use
Dr. Kaye: The second thing that is going to be quite controversial is the notion that you can put together a molecular piece of information that will tell you whether you should give a specific patient bevacizumab. That has been a very hot topic. What is your take on that?
Dr. Coleman: Ever since the data came out on the angiogenesis inhibitors, they have raised many questions, which all good studies should do. When we had the front-line data on bevacizumab from ICON7[3] and GOG 218,[4] we found that there was a progression-free survival difference. If you looked at the curves, they tended to have a banana shape, and they met up at some point. Of course, in overall survivorship, we didn’t see much difference.
We sensed that there was a sweet spot. One of the hypotheses was that in both trials, maybe we stopped the therapy too soon. The hazard ratios became nonproportional, and as the drug was withdrawn the benefits seemed to wane.
In GOG 262,[5] we essentially treated the patients until progression, and we didn’t see a robust response or difference between the curves. It raises the possibility that there is a group of patients who would do well and a group who might not.
The 2 presentations[6,7] that looked at the genomic signatures of high-grade serous ovarian cancers that might respond vs those that probably will not were very intriguing. In the one set, they broke down these datasets using a learning set and a validation set. They found a 63-gene signature in a set of Scottish patients and applied it to a separate dataset.[6] They showed that they could essentially validate the signature. Then they applied it to the ICON7 population and identified that in the patient population with this immunoreactive phenotype, there wasn’t a tremendous amount of benefit, and in fact it might be detrimental. It supports our bias somewhat about what angiogenesis inhibitors might be doing to the microenvironment from an immunologic standpoint.
Dr. Kaye: The other study, although it didn’t show a negative effect, was going in the same direction. That was the AGO study by Winterhoff.[7] It made me think that we are getting to the point in ovarian cancer where these molecular subtypes (David Bowtell classified them as C1, C2, C4, and C5[8]) are starting to make a difference in determining what treatment to give a C1 or a C4. Maybe there is a group of patients for whom it will be obvious to give an antiangiogenic agent. And there might be an immunologic agent. Maybe we are just beginning to get to the place in ovarian cancer where we are in lung cancer and breast cancer already — in which the treatments are divided across separate parts of the pie.
Dr. Coleman: It’s going to be multiple silos of different treatments according to how we identify the patient.
Dr. Kaye: Obviously there will be other ways of trying to predict the benefit from bevacizumab, but this feels right. I’m not a molecular biologist, but it seems a more logical way of looking at it than some of the immunohistochemical markers or seromarkers.
Dr. Coleman: It has been a struggle, yes. We have had trouble finding that signature.
Immunologic Agents on the Horizon
Dr. Kaye: Mentioning the immunologic agents brings us to the group of new agents. The big news last year was PD-1 and PD-L1, and quite rightly.[9] But this year, there were a couple of discussions on that topic in melanoma,[10-13] and for the first time we saw maybe a glimpse of the future in ovary. What did you think of the Japanese study[14] with small numbers?
Dr. Coleman: Yes, it was a small phase 1 study. Most of us who do drug development, when we do a trial with a dose escalation in a drug that we think should work, and we see some responses, we are very happy. Of course, if we don’t see a response, we always say, “Well, maybe we were just trying to get to the active dose.” To see responses through the dose escalation scheme is encouraging. It is a completely new strategy for ovarian cancer, but we have a sense that this probably is relevant. Fortunately now, because of the availability of these drugs and their proven efficacy in other tumor types, the idea that there are ovarian tumor types that might benefit is starting to bleed over. In that trial, they saw some responses at the higher dose. The fact that they saw responses at all is encouraging.
Dr. Kaye: Yes, they look genuine. It’s always hard to know. There were 3 responses out of 18, weren’t there?
Dr. Coleman: Yes.
Dr. Kaye: What is impressive about the drug (the PD-1 antibody, nivolumab) is that it seems to be well tolerated. The exciting thing is the notion that we could identify a subgroup of patients who should get this or other PD-1 agents in first-line treatment or maintenance. The world is full of combinations now.
Building the Right Smart Bomb
Dr. Kaye: A drug that has always disappointed me to an extent is trabectedin — a compound from the bottom of the sea — which, in combination with doxorubicin, showed a benefit. We got into that because of its potential in platinum-resistant disease, and it never really panned out. Did you see the data on the successor compound, lurbinectedin?[15] What do you think of those data?
Dr. Coleman: There were some fairly interesting data looking at the activity of that compound in a homologous recombination deficiency state, showing that it might generate a synthetic lethality signal.
We did see some responses, and that is always encouraging. Our thoughts are that in resistant disease — because so many drugs have been evaluated in combination in that cohort of patients — coming up with a good portfolio of what this drug does as a single agent and in different combinations might be a new direction to look at this particular mechanism of action.
Dr. Kaye: The other newer agent was the antibody-drug conjugate, wasn’t it? It is the sodium-dependent phosphate transport protein 2b (NaPi2b) antibody combined with monomethyl auristatin E (MMAE).[16] The idea of antibody-drug conjugates is a very old notion that is beginning to pan out in breast[17] and other cancers. It may be that if you got the right receptor and the right linker, it works.
Dr. Coleman: The whole concept is to get a smart bomb. It’s to use a toxic agent that you can’t deliver systemically and have it honed for the type of cells that you want to target. The fact that there are better ways to identify the cancer cell from the normal cell and achieve good cytotoxicity and reduce the side-effect profile opens up a big opportunity for development.
Can Weekly Paclitaxel Be Improved?
Dr. Kaye: We both use weekly paclitaxel all the time. We want to try and improve on the benefit of that. A study[18] looked at pazopanib, in addition to weekly paclitaxel, which was quite strikingly positive. Pazopanib has been through some ups and downs. What was your take on that Italian study?
Dr. Coleman: You and I both have dealt with this drug as a single agent and in combination treatments. We know that it has some activity, and we know its side-effect profile. It was striking to me that what they saw in this trial mirrored pretty much what we saw with another angiogenesis inhibitor in combination with chemotherapy, in the AURELIA trial.[19]
We have always had this bias that the mechanism of action for weekly paclitaxel in taxane-resistant patients has to do with a potential avenue for angiogenesis that isn’t necessarily targeted through the same mechanisms that we see with the tyrosine kinase inhibitor pazopanib. It was nice to see that the directionality was the same and that it could be taken and tolerated.
Dr. Kaye: There is a lot of scope for pursuing small molecule tyrosine kinase inhibitors and other agents with weekly paclitaxel. It is an exciting time. Thank you, Rob. It has been a pleasure to discuss ASCO 2014 with you. Thank you for joining us for this edition of Medscape Oncology Insights. This is Stan Kaye, reporting from ASCO 2014.