Synta Pharmaceuticals Corp. (SNTA) Investor Meeting and Webcast at the 2013 ASCO Annual Meeting June 3, 2013 8:00 PM ET
Safi R. Bahcall - Co-Founder, Chief Executive Officer, President and Director
Vojo Vukovic - Chief Medical Officer and Senior Vice President of Clinical Research & Regulatory Affairs
Safi R. Bahcall
All right, thank you all for coming. We're going to get started. I've asked -- just a couple of logistics. I've asked the panelists to sit up front. They're going to come up in about 20 minutes for a Q&A panel that we're going to have. I'll do a brief welcome. We're very fortunate to have Dr. Neil Spector from Duke here today. He's going to do an introduction to an overview of ganetespib. Dr. Spector has been in academia, had a very distinguished career before he joined industry. And he has had a terrific career since he has joined industry. And you'll hear from him, he's been involved in Hsp90 for many years. Then we're very happy that he has to put on his tennis shoes and run downstairs to a parallel session that we're having for investigators on our clinical trials. So we apologize in advance. He -- immediately after his presentation, he is going to run downstairs. And then we have Dr. Suresh Ramalingam, who is the principal investigator on our GALAXY-1 clinical trial that just presented just about an hour ago at the oral lung session at ASCO. And he'll take you through an abbreviated form of the slide so we can get quickly to a Q&A panel. And we have some guests, and we are very pleased to have in addition Dr. David Gandara, who is the chair of the Southwest Oncology Group and also a very distinguished lung cancer oncologist who has participated in a number of our advisory boards.
So let me just do the -- as I mentioned, this is a -- we're very happy today to have principal investigators from our 2 principal programs in lung cancer and breast cancer. Before I turn this over to Neil, what I hope you will get out of today and what we decided to do is put on a list the top 5 questions that we've heard today. So I hope today, you will get answers to these questions. If you do not, please ask around. We'll give you George Farmer's home phone number, home address, or just pose the questions to the Internet. I'm sure we'll get to them in time.
So first question, results on the diagnosis, 6 month group. How are those? So what's the maturity? How is that involving with time? How are people supposed to think about that? What about the medians? Why was a Cox regression provided? What about the curves crossing on the right of the graph? And how robust or how commonplace is this disease progression by velocity criteria? I think Dr. Ramalingam will talk about it. And then we'll have a panel. We'll open it up to Q&A from the audience, and we'll go through a number of these questions.
So without further ado, let me turn it over to Dr. Spector.
All right, thank you. Thanks, Safi. It's a pleasure to be here. So I was -- I told my daughter when I made this slide, when she saw 1989, I said that it's amazing what a 10-year-old, what kind of research it could do at the age of 10. So -- but it's true, before I -- when I was a fellow at the Dana-Farber, I was very much involved in he tried heat shock protein research before it was as big as it is today. And then subsequently, I went into industry and helped develop Tykerb HER2 breast cancer drug and some others, and now I'm back in academia. So it's a great -- it's really a thrill to actually be here talking about Hsp-- an Hsp90 inhibitor that actually is demonstrating activity in the clinic because it's for me sort of coming full circle and seeing the realization of a dream.
I want to point this out because this is really a major -- I mean, I still see patients, I have a laboratory. But one of the really the major dilemmas in treating cancer patients and in developing therapeutics is the molecular heterogeneity in cancer. So this is a paper in The New England Journal that essentially show that if you take biopsies from millimeter to millimeter in a primary tumor, that there's tremendous heterogeneity. So if you're a drug developer, what are you going to develop a drug against? This millimeter or this millimeter? And that not even really taking into account the heterogeneity between different sites of disease. So the primary tumor -- this pointer doesn't work -- but the primary tumor where you might get a biopsy and do a genomic analysis and say, well, that's where we're going to treat on versus the disease that might be down there in the liver and one that might be up in the bone [ph].
So it'd be nice to have a drug or might -- it'd be nice to have a therapeutic strategy that didn't worry about the millimeter to millimeter variation and can target hundreds of proteins all at once, that might be involved in that site of disease, that site of disease and that site of disease. Okay?
So as I said, I've been involved and I've had an interest in the heat shock world for a long time, and we've seen a variety of inhibitors come and go. So the geldanamycin inhibitors came and went. There were some second-generation serinex-sited [ph] inhibitor. Most of these are really [ph] the lousy drugs from a pharmacokinetic standpoint. They just weren't good drugs as far as achieving good drug levels. Or they had toxicity, retinal toxicity and other toxicities, that were just intolerable and were not developable from that standpoint. Synta has been very lucky that ganetespib is not only a good drug from a pharmacokinetic standpoint but also seems to lack the other toxicities that the ones that came before had. So no visual toxicity, no liver toxicity and is much more potent and impressive than the geldanamycin derivatives.
This is again just pointing out the retinal toxicity. It just shows that ganetespib really doesn't get into the layer of the retina where some of the others do, the DMAG and some of the other Hsp90 inhibitors the do cause visual retinal toxicity.
Hsp90 -- heat shock proteins are fairly ubiquitous, okay? They're one of the most highly conserved proteins in nature. So from bacteria all the way on up, they're incredibly homologous, which is why you have to imagine that their function is critical. And so, years ago, in the late 80s when I was working in heat shock and people said, let's block heat shock proteins, I was laughed at because they said, "Well, you're going to -- patients will fall apart if you block heat shock proteins. It turns out that heat shock proteins, especially Hsp90, is overexpressed in tumors. And actually, even more than that, we're doing research now that the localization -- so it's not just whether it's in the cytoplasma or in the nucleus, but it may be at the membrane of the cell and specifically in the membrane of tumor cells versus nonmalignant cells that may also provide for some selectivity and, therefore, a therapeutic window.
And so here's just some data showing that the overexpression of Hsp90 predicts for a worst survival in non-small cell lung cancer and that ganetespib accumulates in tumor. So the differential of normal tissue. So that's really what you'd want. You want a drug that targets a protein that is preferentially overexpressed and predictive for a bad outcome in a variety of tumors, I'll show you some data that we've generated in a minute, but also preferentially accumulates in tumor and not in normal tissue.
We published this data with some colleagues at Duke showing that overexpression of Hsp90 was particularly a bad prognostic feature in triple-negative breast cancer, triple-negative breast cancer being a horrendous disease in younger women, about 15% of breast cancers where we don't have good therapies and we certainly don't have good targeted therapies, we don't have any targeted therapies at this point.
This is a -- actually, so I've been involved in the ENCHANT study where we're treating women up front with triple-negative breast cancer with ganetespib as a monotherapy. There's now an additional cohort where they'll go on ganetespib and docetaxel.
This is the -- one of the first patients at Duke that was treated, a woman who had recurrent triple-negative breast cancer, again when you recur the triple-negative breast cancer. So she had early-stage disease in 2011, received the adjuvant FEC, 5-FU, epirubicin, cyclophosphamide, did okay for 2 years then recurred with multiple bulky disease, lymphadenopathy, significant pulmonary nodules, some visceral disease as well, and went on ganetespib monotherapy and literally within 3 weeks had a near complete remission and has continued to do well with a very impressive -- with a monotherapy, with a single drug that is not considered cytotoxic like the usual drugs where you lose your hair, you wipe out your bone marrow. This woman is back working and has a very functional life compared to what she had before she started therapy.
Okay, so there's about 200 client proteins. So you could take your wish list of what you'd like to knock out if you're an oncologist or a cancer biologist. Those our client proteins for Hsp90. And therefore, that notion of heterogeneity, you don't need to worry about what's this -- at this millimeter, what's at this millimeter, or what's in the bone, what's in the liver. It's likely that it's going to be on that client list.
So we can either -- so the clients really can fall into several categories. One is in tumors that are thoroughly oncogene addictive. So the ALK mutated non-small cell lung cancer, now there's other ALK inflammatory breast cancers. HER2 -- so 15% to 20% of HER2 breast cancers, 15% -- 10% of gastric cancer overexpress HER2. So you could -- ganetespib should target tumor center oncogenic addicted to a client protein.
Okay, these are the client proteins.
Now you could give, and people have given, an inhibitor specifically to that client, okay? But it only targets that client. It doesn't target any of the other oncogenes or drivers that might be very relevant to that tumor. So here you've got your ALK rosatinib [ph], you've got drugs like Tykerb, which I helped develop, and trastuzumab all the other TDM-1 now, and you got your BRAF. So specifically targeting the oncogene. But tumors are smart and they develop resistance. So I could tell you that there's a lot of furor on TDM-1. Women are progressing on TDM-1. It's not a cure. You get ganetespib. It's like giving an ALK inhibitor, a HER2 inhibitor and a RAF inhibitor all at the same time, okay, without much toxicity.
Here are some of the data. Here ALK-positive mutated non-small cell lung cancer. Here's an example. Here's a response with single-agent ganetespib. Same -- we can through mutant BRAF in non-small cell lung cancer, same thing. And again, HER2 overexpressing gastric cancer, dramatic response with single-agent ganetespib therapy. So you can target those tumors that are oncogene addictive.
But the other clients that are involved are those that promote an aggressive tumor phenotype, okay? So they're not necessarily the mutated proteins that you see without an EGFR but HIF-1, hypoxia-inducible factor, 1alpha, and some on the other VEGF, V-E-G-F, some of the cytokines that might be involved in promoting angiogenesis and metastasis.
There's plenty of data supporting the role of Hsp90 in promoting a metastatic phenotype, what we call EMT, epithelial-mesenchymal transition. In prostate cancer, in breast cancer and any cancer you name, tumor cells that undergo EMT are more migratory, more invasive, more metastatic, and Hsp90 promotes that. And an Hsp90 inhibitor, would be ideal to block that EMT process, that metastatic process. Heat shock protein 90 has been shown to be involved in angiogenesis, I just showed you VEGF preceptor, HIF-1, and is also -- and I'll you that data in a minute.
So here's some preclinical data doing what we in the laboratory call the standard invasion assay where you take cells, you create a space and you see how quickly those cells migrate and fill in that space. And you could see the top where the cells are treated with vehicle 24 hours. They pretty much filled in that gap. If you treat with ganetespib, you could see that there's a significant inhibition of tumor cell migration. This is in a Petri dish. So this has happened in vivo. So these are animals that have been treated with ganetespib, and you can see, importantly, patients die of metastatic disease, right? You can live with a primary tumor. I mean, women with breast cancer, patients with lung cancer can live with a tumor. They can't live with tumors that metastasize to their brain, liver and lungs. That's what people die from. This drug preclinically, and I think you've seen some of the data already, impressively not only can target the primary tumor but targets metastasis. Again, that's what kills patients.
So here's the data. Here's a non-small cell lung cancer model -- or actually, a breast cancer model, a very aggressive triple-negative breast cancer, 4T1. And you could see pretty clearly that if you treat the animals with vehicle, you get a lot of lung metastases. If you treat them with ganetespib, you completely block that. And I could tell you I work with 4T1. That model is notoriously refractory to chemotherapy. So all the cytotoxics, you name it, cyclophosphamide, adriamycin, they don't touch 4T1 tumors.
HIF-1, I mentioned, HIF-1 being an important mediator of the metastatic profile of dissemination. It's a Hsp90 client protein. So it's protected by Hsp90 from degradation. You give it a 90 inhibitor, it releases that complex and HIF-1 is degraded. Why is HIF-1 important? Yes. Okay, so here's a -- here's animals. There's a control animal. These are luciferase models. So the darker red, green, yellow shows increasing tumor growth and metastatic dissemination. You could see that in the HIF-1 knockout, where the animal's HIF-1 has been knocked out, there's really very little to no tumor growth and no metastatic dissemination, okay? That shows HIF-1 is important. We know HIF-1 is important in metastatic dissemination.
So what does ganetespib do to HIF-1, this an important pro-metastatic protein? Well, what it does is here you could see increasing HIF-1 in tissue as you get away from blood vessels. HIF-1 is a hypoxic -- hypoxia-inducing factor, right? So it's inducible where there's low oxygen concentrations, which happens in tumors. And ganetespib blocks the induction of HIF-1. That's important. We'd like to see that because HIF-1 promotes metastatic disease.
How does that translate into patients? So here's data that I think you probably all heard, showing that the time to new lesion is significantly reduced if you combine docetaxel with ganetespib, again blocking progression of disease, blocking metastatic dissemination.
Resistance. So the other category of clients are all these proteins that are involved in mediating resistance: DNA repair, anti-apoptosis, cell cycle, you can see it. So combining an Hsp90 inhibitor like ganetespib should enhance the sensitivity of tumors to chemotherapy and radiation therapy.
And here are some of that data. You can combine with traditional chemotherapies. You can combine with some of the targeting angiogenesis inhibitors like Avastin or bevacizumab. You can combine with radiation therapy and so on and so forth, a BRAF inhibitor. So this makes a good combination, low toxicity, doesn't really shouldn't add a lot of toxicity to some of these other therapies.
So in essence, ganetespib has a unique mechanism of action, simultaneously inhibits client proteins involved in multiple oncogenic pathways, has profound effects on aggressive tumor biology, importantly reducing metastatic potential. Again, that's what kills patients. Too many people look at response in the primary tumor. For me, that doesn't matter all that much. It matters if someone metastasizes to their liver, that's what's going to kill them. And I could tell you personally I had single-agent activity in some very aggressive tumors. And I think there's a strong scientific and clinical rationale for using in combination with other therapies.
I guess I'll be back.
Safi R. Bahcall
Yes, thanks a lot, Neil.
So good evening. I'll walk you through the data we just reported at the metastatic lung cancer session a few -- a couple -- 1-1/2 hours ago.
So this is the overall scheme of the GALAXY program. We knew the nice preclinical data that Neil just went over, where you can combine Hsp90 in the [indiscernible] in the preclinical setting and have synergistic effects with taxanes. So how do we test this in the clinic? In the era of molecularly targeted therapies, the question always comes down to: Who's going to benefit? Is this benefit going to be in a broad group? Or are there subsets of patients where you're going to see the benefit?
So in order to learn as much as possible in a Phase II trial before you go to a Phase III, we did a very large randomized Phase II trial. This is 240 patients, which is large by any standards for a Phase II. And the overall goal was to test the efficacy of the combination and to see which subgroups may derive the greatest amount of benefit. So this was for patients with advanced stage non-small cell lung cancer who had received the standard first-line therapy and had documented disease progression. And patients were required to have a performance status of 0 or 1, and they're randomized to the combination of ganetespib and docetaxel versus docetaxel. Docetaxel is the approved agent. Erlotinib and pemetrexed are 2 of the other agents that are approved. And each of these 3 agents appear to have comparable efficacy in this setting. We stratified patients based on performance status, time since diagnose of advanced disease less than 6 months or greater than 6 months, baseline LDH and smoking status.
Now I just want to spend on this less than 6 months versus greater than 6 month population. Why do we need this? Well, when you look at second-line therapy of lung cancer, we learned very early on several years ago when some of these salvage therapy drugs were approved that if a patient doesn't respond to first-line chemotherapy or benefit from it, those who rapidly progress, they don't do very well with subsequent therapies either. So when you design a randomized trial in the salvage therapy setting, you want to make sure that both arms of your trial have an equal subpar -- equal proportion of these rapidly progressive disease so that, that variation does not confound the results of the trial.
Why did we pick 6 months? Well, if you look at contemporary randomized trials, they would have -- they have chosen 3 months from the last chemo. So if you imagine the first-line chemotherapy regimen that last from -- anywhere from 3 to 4 cycles and you add the 3 months, you come up with something around the 6-month ballpark. We felt that this was easy enough for investigators to know, and that's why this was chosen. And we can get to it more if people have questions.
And then we looked at the endpoints. We were interested in the KRAS mutated patient population because downstream proteins in the KRAS pathway are client proteins of Hsp90.
Elevated LDH was another subgroup we were interested in because we know that tumors that have hypoxia tend to have elevated LDH. And tumor hypoxia results in increased release of HIF-1alpha, which is a very sensitive client protein for Hsp90.
So these were 2 of the endpoints that we had thought about going in. But clearly, you have to look at the overall population, and that was a key secondary endpoint of the trial.
Now for the combination arm, patients were allowed to continue the combination or monotherapy until they have progression of disease. But realizing that docetaxel in North America is not given for extended lengths of time because of cumulative fatigue, we allowed the investigators to say after 4 to 6 cycles of docetaxel. If they want to continue just the ganetespib, if patients were still benefiting, they were allowed to do so.
So as the trial was designed, we wanted to have 240 patients with lung adenocarcinoma. And among these 240, we were hoping to have 80 patients with mutated KRAS and 120 patients with elevated LDH. But the protocol specified that if in the 240, you don't have those 2 subgroups to 80 and 120 patients each, then the trial would remain open just for those 2 subgroups of patients so that you can get to the 80 and 120.
So at the time this was completed, we still were not done with the elevated LDH and mutated KRAS cohorts, so enrollment for just those 2 groups were kept open until May, less than 4 weeks ago when the last patients were -- was enrolled.
So the data for this group of patients, elevated LDH, are not mature enough to report at this point, and we hope to report that at a meeting later this year. And here, the primary population that this report is based on had 252 patients. The interim analyses were planned for 6 months from the accrual of last patient and 12 months from the accrual of last patients. So that would have been May 26 based on what we had, but because we had to submit the slides to ASCO ahead of time, we did the analysis about 11 days earlier than what the interim analysis called for.
And you have the statistical parameters. Given the scent [ph] of the sample size, we had 88% power to detect a hazard ratio of 0.67 for BFS and 0.71 for overall survival.
252 patients. Patient baseline characteristics were very similar between the 2 arms of the trial. 56% of the patients were men. 75% were current or former smokers. 17% of the patients enrolled to the trial came from North America, but 22% came from Western Europe, 60% came from Eastern Europe.
Overall, what did we see in terms of safety of the combination? It was very well tolerated. As one would expect certain toxicities are seen more frequently with the 2-drug combination, and that included mild to moderate diarrhea. And this diarrhea is very different from what you see with an agent like erlotinib or [indiscernible] where it's seen throughout the course of therapy. Here, it's about 2 to 3 days after the infusion, the diarrhea, that's when you see them. By giving prophylactic anti-motility agents, it has not been a major problem. Grade 3 diarrhea was seen in only 3% of the patients.
The other toxicities, we certainly saw a higher number of fever with neutropenia episodes, 11% versus 2%. Even though it's high, I think 11% is still a range that we would be willing to tolerate with an effective regimen. We did see more fatigue, nausea and anemia, though most of them are Grade 1 and 2 and not Grade 3 or 4. So overall, we felt that the combination is tolerated well.
And here is the PFS score for the overall patient population. Median was 3.2 months for docetaxel, which is what one would have predicted; 4.5 months for the combination. And if you look at the hazard ratio, it's 0.83 on the Cox regression and 0.84 overall. So the PFS was in favor of using the combination, or the combination was associated with a favorable PFS.
Here, you see the overall survival for adenocarcinoma patients, 9.8-month median versus 7.4 months. And the hazard ratio was 0.82. And then when you did a Cox regression to adjust for variables, the hazard ratio was 0.73, and the P-value for both were significant 0.04 for the Cox regression.
One thing I want to point out here is, this came up as a question, is why are these crossing over? The last 3 months of the trial, we saw a significant accrual rate. So many of the patients are still on treatment and events are still followed, so you see some sensoring [ph] here. But this part of the curve is relatively mature. And therefore, when the data matures, we anticipate that this is unlikely to be seen.
Now Here are the forest plots for baseline variables. If you look at ECOG performance status, gender, geographic region, LDH population, the overall trends seem to not have a major or favoring the treatment arm. Again, small subgroups, large confidence intervals. The biggest factor that stood out was the greater than 6 month population. And this population accounted for 70% of the patients enrolled to the trial. So 176 patients. And when you apply the test for interaction for this factor with treatment, the P-value here was highly significant, 0.0064, which obviously calls for further evaluation of that population in greater detail.
When we looked at the mutated KRAS, as you can see there are only 63 patients at the point of this -- at the time of this analysis, and it doesn't seem like 1 group mutated KRAS or wild type had a differential effect. So this is something we we're closely watching. We need to see what happens after the data matures for the 80 patients that had been enrolled.
PFS for the advanced disease greater than 6 month population shown here. You can see here there are -- the data are very mature for this kind of an analysis for PFS, 5.4 months versus 3.4 months, and the hazard ratio is 0.62 with a highly significant P-value.
Overall survival, for this patient population when you look at the hazard ratio, it's 0.55. P-value, again 0.0036. Medians of 10.7 versus 6.4 months, also favoring the combination in this group.
Now if you look at the forest plot just for the greater than 6 month patient population, you can see that just about every group appears to have favorable outcome with combination therapy as opposed to monotherapy with docetaxel, telling you that this is not just driven by anyone small subgroup of patients, the benefit is broadly seen, so indicative of a biological effect across patients or populations that favors the combination.
So these are the conclusions of our trial. Ganetespib in combination with docetaxel extends PFS and overall survival compared to docetaxel alone in this patient population. And based on a prespecified subset, the overall survival and progression-free survival is substantially enhanced in the greater than 6 month population.
The early evidence indicates that these improved outcomes were not driven by EGFR status or KRAS status, and the combination is tolerated well. And this patient population is basically what's being tested in the GALAXY-2 study that has started enrollment. So this will be for patients with advanced stage lung adenocarcinoma, one prior regimen, ECOG performance status of 0 or 1, greater than 6 months since the time of diagnosis and randomization to ganetespib docetaxel versus docetaxel with overall survival being the primary endpoint.
So some of the questions that have come up both in the press meet that ASCO had especially this morning and in the general session were, "Could any -- could this just be a fluke that you saw, this improved outcome just for the subpopulation of greater than 6 months and the overall population? And my response to that is by doing a Cox regression and a log-rank test, you're collecting for as many known variables as you can possibly think of. And these are sequential tests. So you look for one factor to the other, you choose, you're to remove the ones don't that play a role. And then by doing this kind of an analysis, you are able to correct for these imbalances. So given the relatively large subset of patients that were in the greater than 6 month population, nearly 175, we feel that this correct for any potential imbalances that we're not thinking about, or at least things that we can think off.
Now post-study therapy could -- treatments patients receive after they went off the trial have influenced the outcomes. We saw that 44% of the patients in the combination arm got some form of post-study therapy and 38% in the control arm. But I would also remind you that about 20% -- 20 patients, not 20%, are still on treatment in the combination arm as of this analysis, and 10 patients in the control arm still on treatment. So the final data for post-study therapy numbers may change, but overall we don't think that there's a big imbalance in post-study therapy that could account for the absorbed differences in survival.
Now the other thing is what I just talked about, consistency of benefit we can see from the forest plot.
So that's basically our presentation at this time. I would be happy to take questions as part of the panel. Thank you.
Safi R. Bahcall
All right. Oh, yes, watch your step.
I ask the panelists to come up in front to the panel, Dr. Heymach. Some of you who are at our ESMO panel, remember we had representatives from Eastern Europe, Western Europe and North America. Unfortunately, our Western Europe panelists couldn't be here today, but maybe you could go just starting from Dr. Heymach. Could you just mention your name and your position and institution? Do you have a mic there, John?
I don't have, but...
Safi R. Bahcall
Can we get him a hand?
I'm John Heymach. I'm the chair of Thoracic/Head and Neck Medical Oncology at MD Anderson Cancer Center.
Hi, I'm Dr. Saric [ph], one of the co-investigators at GALAXY-1 trial from Eastern Europe, from Serbia.
Phil Bonomi. I'm Director of Hematology, Oncology at Rush University Medical Center here in Chicago.
I'm Vojo Vukovic, Chief Medical Officer at Synta.
David Gandara. I'm the director of Thoracic Oncology at the University of California Davis. And For your purposes also, I'm Chairman of the Lung Committee for the Southwest Oncology Group and the Chair of the Correlative Science committee for NCI for Lung.
I'm Suresh Ramalingam. I'm a medical oncologist at the Emory University in Atlanta, Georgia. And I also have the fortune of being the Chair of the Lung Committee for the Eastern Cooperative Oncology Group. Actually, we worked together on many projects.
Safi R. Bahcall
You work [indiscernible]?
No actually, we work together on many projects.
Safi R. Bahcall
All right. And I should mention before we get started we've had, over the last 2 or 3 months, of course all under confidentiality, reviewed this -- a lot of the results that were being presented today with a number of advisory boards in the United States and Europe. We formed an independent data review committee given the kind of results that we're seeing of 4 medical oncologists and a biostatistician from one of the leading cooperative groups to review -- open all the data, and they did review all the data. And one of our panelists was a member of that independent data review committee. So I'm very glad we could have you all here today.
And what I thought I would do is just start by putting a couple of the questions that have come up frequently over the course of this trial and then maybe have the panel address some of them. So number one, one of the questions that has come up fairly often, the results in the diagnosis greater than 6 month group reported today was around a hazard ratio of approximately 0.6 at ASCO. And that represents about 52% of events, 52 out of the 176 final size of the cohort. That seems to be somehow worse that what we -- what was reported at ESMO last year, 9 months ago, where the overall survival hazard ratio was around 0.4, which represented 25 -- or 108 or, in other words, 23% of survival events. And someone phrased the question as I don't remember exactly what he said today. "Well, I'd rather be up in the fourth quarter by 10 points than up by 15 points in the first quarter," and that's how they phrase it. But I wonder if anyone on the panel, maybe someone could speak to this. I think Dr. Ahman [ph] has spoken to this before. Is there anyone else who would like to speak to this?
Safi R. Bahcall
Go ahead, just the numbers. This one.
Well, I mean, I think that analogy about being up -- you'd rather be up by 10 points late in the fourth quarter, and obviously the later it is, the less your margin makes a difference. Obviously, the data firms up the more mature it is. But also, the impact of time line treatment -- when you're talking about overall survival endpoints, the impact of time on treatment as opposed to post regression, there's a slightly different waiting as more of the patients spend more of their times off progression. So I think this is a natural trend. I wouldn't expect it to continue in the same direction. I would expect it to be coming closer and closer to the final value, the actuarial value, right now. So I think that the -- whatever the final actuarial analysis, I would expect to be much closer to this final number.
I agree with that. I think it's more mature, and 0.6 is still very good -- 0.6 is very good in a lung cancer study.
Safi R. Bahcall
Does anyone in the audience have any questions about this topic?
Okay. So the second question has come up a few times, the median overall survival for the control arm, the docetaxel control arm in this study. In the ITT population, the intent-to-treat all patients was 7.4 months numerically. The confidence intervals are provided there. So that's the median. Whereas in the diagnosis 6-month population, it was 6.4 months. So that appears on the face of it to be, if you just looked at the median numbers, paradoxical. Shouldn't these patients in some sense be healthier? They survived a little bit longer. Why is their median survival somehow shorter? I think when you put the confidence intervals there, actually the confidence intervals are indistinguishable. So I think the confidence intervals may answer the question already. You just -- you don't really have the maturity. And when you looked at the survival graphs, there's a giant bolus of patients, as Dr. Ramalingam said, that were enrolled in September to November, that all those dots on the 5- to 8-month, that's right around the median. So you just need to wait a few more months, and then those medians will, as John just said, Dr. Heymach just said, will approach much more close to the final value. Is there anything else that anyone in the panel -- this question has come up quite a few times, so I just wanted to make sure we put it out there and address it. Maybe Vojo, did you have a thought or David?
Yes, I just wanted to really maybe iterate the point. I heard your point, and I think Suresh mentioned that as well. But literally, in the last 2 months of the trial, we recruited close to 80 patients. So that was -- the last patient came in 6 months ago. First patient of this large group came in 8 months ago. And these are -- and the patients who are still sensitive [ph], these are the patient with the better survival, they're just now approaching the median. It will be at least a couple of months, as Safi mentioned, before we have a more precisely outlined median. So these median -- medians that we had come -- we had at this interview are likely to change.
Safi R. Bahcall
Does anyone else have anything?
I just might add that I'm sure that many of you look at data like this and you'd like to micro-dissect it and you'd like to look for red flags. But in reality, in a study like this, I think the explanation that has been given is more than adequate. And in fact, the more we micro-dissect the data, the more likely we are to come up with false hypotheses. So the proof of the pudding is really going to be the next GALAXY trial.
Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.
THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.
If you have any additional questions about our online transcripts, please contact us at: firstname.lastname@example.org. Thank you!