Madrigal Pharmaceuticals, Inc. (MDGL) Q3 2012 Earnings Report, Transcript and Summary

Good day, and welcome to the Synta Pharmaceuticals Third Quarter 2012 Financial Results Conference Call. Today's conference is being recorded and webcast. At this time, for opening remarks, I would like to turn the call over to George Farmer, Vice President of Corporate Development at Synta Pharmaceuticals. Please go ahead, sir.

Hello, and thank you, all for taking the time to join us today. With me are Dr. Safi Bahcall, our Chief Executive Officer; Keith Ehrlich, our Chief Financial Officer; and Iman El-Hariry, our Vice President of Clinical Research. This morning we issued a press release that reported results for the third quarter of 2012. This release can be found on our website at syntapharma.com. Before we begin, I would like to point out that we will be making forward-looking statements based on our current intent, belief and expectations, which are subject to certain risks and uncertainties. Additional detail can be found in related SEC filings, also available through our website. I will now turn the call over to Dr. Bahcall, after which we will open the call for questions. Safi?

Thanks, George, and thank you, all for joining us this morning. Today it's going to be me, Keith, and instead of Vojo, our Chief Medical Officer, we have Dr. Iman El-Hariry, our VP Clinical Research. Vojo is traveling. This has been a transformational year for us here at Synta. We've made great progress with ganestespib, our lead Hsp90 inhibitor on 3 dimensions. First, single-agent activity has been clearly established in certain target indications, such as ALK+ lung, HER2+ breast, and triple negative breast cancers. Some patients have had responses for over 2 years while on drug, which is an exciting confirmation of both clinical activity and long-term tolerability. Second, results across 20 trials with over 600 patients treated to date have confirmed ganestespib is the first Hsp90 inhibitor to demonstrate clinical activity without the serious liver or common ocular toxicities seen with other drugs in this class. And finally, third, the randomized data presented at ESMO last September, showed an encouraging survival advantage in second line lung cancer. The first positive randomized data for an Hsp90 inhibitor. We're very pleased to announce today that we have recently completed 2 key milestones in the GALAXY program: enrollment of the planned 240 patients in the Phase IIb portion of the trial has completed. We have seen a very strong increase in awareness, enrollment and investigator and site request for participation, following ESMO. We enrolled close to 50 patients last month alone from less than 50 sites. We will have over twice as many sites, which will participate in the Phase III. And second, we completed our end of Phase II FDA meeting, converged on a Phase III protocol and have initiated the Phase III trial. We are enormously excited about this Phase III trial. It is very rare to be in Phase III with so much randomized data in exactly the same design, same dose, same schedule, same population, and same multi-national setting as in Phase II. This reduces many of the common Phase II to Phase III transition risks. As was described in our release this morning, the Phase III trial will be a larger version of the Phase IIb trial with 2 adjustments: first, the primary endpoint will be overall survival; second, the trial will be enriched for patients who show the greatest likelihood of deriving benefit from ganestespib, based on analysis of the Phase IIb data. One note about the Phase IIb. The protocol specified enrollment in the primary sub populations of mutant KRAS and elevated LVH may continue until a prespecified maximum number of patients in each group has been at reach. We expect up to an additional 60 patients will be enrolled over the next couple of months. Enrollment of these additional patients is not expected to change our timelines of adding final PFS data and updated survival data in the first half of 2013. It also does not change our primary focus, which is winning in 2014. Meaning, a positive outcome on Phase III trial in the all adenocarcinoma group. Patients with mutant KRAS and elevated LVH, however, have particularly high unmet needs. There are limited treatment options in the first case, and a very poor prognosis in the second case, with a median survival of about 4 months. The additional patients will not affect our longer term all-adno [indiscernible] strategy. The results in these populations, however, may create interesting opportunities next year for early filing discussions with health authorities, based on the high need, should outcomes be positive. So finally, I'd like to summarize our expected timeline for the GALAXY program over the next 24 months. In the first half of next year, we'll have final PFS and updated survival data from the Phase IIb, and that should about 300 patients. In the second half of next year, we'll have final survival data on the Phase IIb. In the first half of 2014, there'll be an interim survival analysis from the Phase III trial. The second half of 2014, we'll have final survival data from the Phase III trial. We see GALAXY as the fastest and most efficient path towards registration. We do have, however, several other trials, including the CHIARA trial in ALK+ lung cancer, and the ENCHANT trial in metastatic breast cancer. We've seen very interesting activity with ganestespib in these indications. We expect to present interim results from these trials in the first half of 2013. I'd like to share one patient example we received recently. One of our investigators was treating an ALK+ lung cancer patient. She had relapsed after treatment with crizotinib, and had a large central tumor mass that led to complete collapse of the left lung. She also had extensive bone and brain metastasis, an extensive leptomeningeal disease. The patient was treated on a compassionate-use basis with ganetespib. The 6-week PET/CT scan showed a complete metabolic response, with the left lung reinflated and regained function. In addition, there was partial metabolic response in both the bone mets as well as the brain mets. As you all know, quite exciting to see brain mets responding. As another example, we recently received the first report from the patient in our ENCHANT trial, a HER2+ patient who received ganestespib monotherapy. We're excited to learn that our first planned PET/CT scan, just 3 weeks after the start of ganestespib treatment, showed both metabolic activity and tumor shrinkage. To summarize, it's an incredibly exciting time for this program, a big inflection point with lots of data coming. Lung and breast cancer are the 2 biggest cancers in the world, and we have 6 trials across each, have demonstrated clinical activity with ganestespib across both, and are positioning Synta to make a big difference to patients across both these tumor types. I will know to turn it over to Keith who will review our third quarter financial results. Keith?

Thank you, Safi, and good morning, everyone. There were no revenues in the third quarter of 2012 as compared to $1.7 million of revenues in the same period of 2011, which consisted of amortization of an upfront payment under the Roche agreement that was terminated at the end of 2011 and grant revenues. In the third quarter of 2012, research and development expenses were $11.7 million as compared to $10.8 million for the same period of 2011. Third quarter general and administrative expenses were $2.8 million, as compared to $3.1 million for the comparable period in 2011. In the third quarter of 2012, the company's net loss was $15 million or $0.25 per basic and diluted share, as compared to a net loss of $12.7 million or $0.26 per basic and diluted share in the same period of 2011. As of September 30, 2012, the company had approximately $55.1 million of cash resources on hand. This includes net proceeds from the registered direct common stock offering to certain members of the Board of Directors in July 2012. The company expects to end 2012 with $38 million to $40 million in cash, cash equivalents and marketable securities. Based on our current operating levels, we expect these cash resources will be sufficient to fund operations into the second half of 2013. These estimates assume no additional funding from new partnership agreements or equity financing events, and that the timing and nature of activities contemplated for 2013 will be conducted subject to the availability of sufficient financial resources. I will now turn it back over to Safi for concluding remarks. Safi?

This concludes our prepared remarks. Operator, we will now open the call to questions.

[Operator Instructions] Our first question is coming from Thomas Wei from Jefferies & Company. Thomas Wei - Jefferies & Company, Inc., Research Division: Just wanted to understand a little bit more about the Phase III protocol. Can you share with us more details on exactly how you're defining rapidly-progressing patients for exclusion?

Sure, Thomas. Criteria is essentially patients who have experienced rapid disease progression on first-line therapy or shortly thereafter. And it's essentially the same as the times since diagnosis, less than 6 months. It's a very strong overlap in that criteria that came out of the Phase IIb. Thomas Wei - Jefferies & Company, Inc., Research Division: I see. So that's predominantly what's driving it, is that you have to have had a diagnosis of advanced disease greater than 6 months prior to study entry?

That will be the eligibility of the inclusion. And it's essentially that. It's not precisely that wording, because that's obviously a standard stratification factor. And then when you get into eligibility, moving from a stratification factor to an eligibility criteria, you want to get the language that's most acceptable to investigators and regulators. So we actually -- and it's interesting, we did a lot of work on this. We obviously have a ton of data that came out of the Phase IIb randomized data, which gives us enormous amount of insight into the patients that are really benefiting from the drug versus the patients that are not. And one of the findings that left out of that and it's not particularly unusual to ganestespib, is that patients who are rapidly progressing or just blowing through first-line therapy don't seem to be getting much benefit, either from the control on docetaxel or from the combination of docetaxel plus an Hsp90 inhibitor, and so that came very clearly out of the data. And then we spent quite a while talking to many, probably over 15 of the top lung cancer investigators in North America, Europe, Asia, who collectively have been involved in essentially every major pivotal lung cancer trial over the past 2 decades, and walked through the data with them and got their thoughts. And it was extremely broad support for the press [ph] that we're taking, which is to really understand which patients are benefiting, which of patients have the highest likelihood of benefit, number one. Number two, what is an objective, practical and reliable criteria to use in the pivotal trial that's going to be in a multinational setting and used for registration? And number three, what would be most acceptable to investigators, institutional review boards and regulators around the world? And that's just a general set of principles about how you think about any eligibility criteria in the pivotal program, but also in particular, this one in terms of when you have large randomized Phase IIb data set in the same setting and there's some clear signals. So when we look at, those are the 3 factors we're trying to optimize. We feel pretty comfortable that the specific choice of language that we're using which has a pretty high overlap with that, what people think about when they think about patients who progress very rapidly through first-line therapy, we feel pretty comfortable that we've had a very good balance between all 3 of those. Thomas Wei - Jefferies & Company, Inc., Research Division: And you had said before that you were going to share some of the statistical powering assumptions, both of the final overall survival analysis and at the interim when you announced the start of the study. Could you share those with us?

Sure. We'll obviously have more details when we have the start of enrollment press release, which should be early next year. But what I can tell you is that we powered for -- the powering gives you about an assumed hazard ratio of 0.7, an 85% power on the final analysis, and in the neighborhood of 40% and 70% on the first interim and the second interim if you assume a hazard ratio of 0.7. And if you assume a hazard ratio of 0.6 or below, then your final powering is over 99%, so your final power on the final analysis is over 99%. And then the interim analysis, it's extremely high as well in 80% to 90%. Thomas Wei - Jefferies & Company, Inc., Research Division: And just to tie it in to the first question, using the criteria as has been described on inclusion and exclusion and applying that to the Phase IIb population, what is the hazard ratio that you would see on that subgroup of patients from the Phase IIb portion?

It's less than 0.4.

Our next question is coming from Brian Klein from Stifel, Nicolaus. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: Digging in a little deeper on the Phase III. First, can you tell us if the Phase IIb was stratified equally between the 2 arms in terms of patients who are rapidly progressing?

I think what you're asking is, was there a stratification for rapidly-progressing to the patients who entered the study with rapidly progressing disease, which if there is a stratification in the trial, it's balanced automatically. I think that's what you're asking. And the answer is yes. So one of the--there are 4 prespecified stratification variables, and one of them is this time since diagnosis of less than 6 months, which is a very good proxy for rapid disease. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: And did you see that approximately 30% to 40% of the patients enrolled in the Phase IIb were rapidly progressing? Is that consistent?

Yes. In the Phase IIb ESMO data, I think it was about 36% of the patients were diagnosed with less than 6 months or rapid disease progression. And 64% were normal tumor velocity patients or normal progressives. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: So what do you then estimate will be the median survival for the Taxotere alone-arm who are not rapidly progressing?

That's a good question. You can look at both historical data and the results in our trial. And I think if you look at the Kaplan-Meier graphs in our trial, it's right there. And to what extent that median survival is reflected of? Historically, it's something that we've looked into, and I think it's generally within the range that you might expect. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: So if we take the data from the Phase IIb, and there, the Taxotere alone-arm survival was about 7 months or so?

That's right. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: Can we assume that in the Phase III, the -- now that we're excluding rapidly progressors, the median survival should be higher than 7 months?

It's hard to -- there a lot of different factors. In general, everything else being equal, that would be the case. And the one thing I would caution is that medians are a one point in time, and people tend to get -- especially media tends to get very focused on median. But that really is one point in time. And obviously, your final approval is based on the entire shape of the Kaplan-Meier graph of this survival graph. And that's really what is captured in the hazard ratio, and what is captured in the log rank statistical test, the P value. So the most important thing to focus on is certainly when we or our biostatisticians' look at the data and make-plan is really hazard ratios. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: Got it. I think the purpose of my question is I'm really to trying to drill down in terms of when we might get that final analysis. Is it safe to assume that the enrollment of 500 patients will take about a year from start?

Yes. Our projections is that we would -- and we've done these analyses, and we've done these predictions, and that was behind our estimates of having interim analyses in the first half of -- on the Phase III. Interim analyses in the first half of 2014 and final analyses -- final analysis in the second half of 2014. And that assumes enrollment gets completed in 2013. Brian Klein - Stifel, Nicolaus & Co., Inc., Research Division: Okay. But then in terms -- is your ability to have final top line data will require at least 75% to 80% of events to occur, and given a median of about at least 7 months in the control arm, do you still think it's feasible to really have that data by the end of 2014?

We've done a whole series of simulations, analyses and projections. And with those assumptions that I mentioned, that you complete enrollment and fairly standard assumptions, we do estimate interim analysis would take place in the first half of 2014, and final analysis will be in the second half of 2014.

Our next question is coming from Jim Birchenough from BMO Capital Markets.

It's Nick standing in for Jim this morning. And I apologize if I missed the beginning of the call. So perhaps, you've addressed this. But when you had your End-of-Phase II meeting with FDA, can you summarize for us, perhaps what some key messages from FDA were that you've included into the Phase III protocol? And also, did you discuss with them the possibility of filing based on the GALAXY IIb data, and what perhaps additional data that you would need to support an early filing?

Sure. The first part of your question, as you know, we and most companies can't get into minute details of regulatory discussions. I think one of the things, and in this particular case, with this particular protocol, as you know, it's -- we're using the gold standard survival endpoint, so that's about as noncontroversial as it gets. We are using standard-of-care patient -- in both arms, you see standard-of-care patients, and the experimental arm get our drugs. So that's a very plain vanilla and precedented trial. So there's not a lot of surprises that the trial with a lot of precedent in prestandard trial. And it's really an extension of the IIb. It's the -- for which we've -- now have 1.5 years, 2 years of experience and filing in many different countries, works on many different regulatory agencies. So they are in a lot of surprises. I think, one thing I can say is that a fair amount of the discussion are actually we're just making sure we converged and agreed on the statistical plan, and we felt very good. We're pleased with where we ended up in the statistical plan and we accepted entirely the agency's suggestion on the statistical plan. And that was, I think, the majority of the discussion or the focus. I think your second question was on early filing. And I want to caveat here that our -- the absolute #1 focus of the company is on winning in 2014 on having a positive outcome in the Phase III trial in 2014. We and everyone involved with this drug, whether internally at Synta or the many investigators and many centers we work with in U.S., Europe, Asia, understand how important it is to get this right. We understand that we have an agent that has shown pretty remarkable clinical activity in the patient group that is -- got extraordinarily high and urgent unmet medical need, and we take that responsibility extremely seriously. And that's been our discussion with investigators and the investigators' discussion with us. So we are not in the business of trying to -- it's not the highest priority in the world to us to try to see if we can move quicker by a quarter or 2 quarters. We're thinking about the 5 years and the 10 years, and what this drug can do for lung cancer, whether it's second line, whether it's first line, whether it's maintenance, whether it's a combination with this one agent or other agents. We think it has very broad potential, and we do not want to take shortcuts. Number two -- and by the way, not just lung cancer, but breast cancer. We think the drug has so much potential. It'll be a mistake to focus on trying to see if we can shave off a quarter here and there. That being said, this question has come up quite a few times. Since, obviously, when we have 300 patients of randomized multinational monthly center data, it's pretty substantial data. And what we've said consistently is when we get there, we will have those discussions. Right now, those discussions are too premature. If there is -- especially if there is a patient subpopulation or patient subpopulations that have extremely high unmet need, where there are little or no options, or they're dying very quickly, and there is some particularly-impressive demonstration of benefit, that would be the kind of context that maybe a year from now, one would have some interesting discussions with the agencies about possible options. But again, that won't change the focus.

And so how did you come up with an additional 60 patients? Is that -- does that give you, you think, enough patients worth of data in those 2 particular populations to have a meaningful discussion about how we're really changing the natural history of the disease?

Yes, exactly. So that's -- there was prespecified in the protocol, how many patients you would need to answer a particular statistical question, and so that calls for some particular numbers. And based on what we're seeing in enrollment rates, which is that both groups are in the approximately 25% to 30%, you can roughly back out that you need approximately -- I mean, in the neighborhood of 60 patients to get that number of patients you need to answer that question.

And that's 60 patients each? Not -- I'm sorry, total, not each?

Correct. It's 60 patients total. So that the final total cumulative number of adeno patients would be approximately 300.

Our next question is coming from George Zavoico from MLV & Co. George B. Zavoico - McNicoll, Lewis & Vlak LLC, Research Division: Safi, Iman and Keith, congratulations on beginning the Phase III trial. Something, obviously, we and everyone has been anticipating with greatest sentiments. The question about the rapid progressors. This is a fairly easy stream, it's basically patient history. You don't have to do any sort of other analysis, so this is not going to delay patient enrollment. Is that correct?

That's correct. And that came up in discussions with many investigators. And I have to say, it's a pretty quick discussion with investigators. They looked at the Phase IIb data, they say, "You're obviously seeing tremendous benefit in this group, these patients who are blowing through first-line therapy. They're not benefiting from your drug. In our experience with other trials, that's pretty standard, they actually don't benefit much from second-line therapy anyway." So focusing on the narrower group where you see the bigger advantage is the right thing to do. And in fact, they think it's something that should be done in the future for more second-line trials or pivotal trials. And as you say, it's a fairly easy, objective, practical, reliable screen. George B. Zavoico - McNicoll, Lewis & Vlak LLC, Research Division: And on the other patient subgroup that didn't do so well in the Phase IIb portion was the smokers versus the current smokers versus the never smoked or recently stopped. How are you handling that number in the Phase III?

That is a good question, George. We thought a lot about that. And what we saw in the data is that, as you can imagine, there's a very high overlap between patients who exhibit rapidly progressing disease. Meaning, their disease is growing and they didn't respond much to first-line therapy and it's -- there's a very high tumor velocity, and some of these other negative characteristics such as current smokers or liver bone mets, and so forth. So by -- with the one eligibility criteria, you actually reduce the number of these patients who are not very responsive pretty broadly across the board. And that's reflected in the fact that in the Phase IIb data, the overall hazard ratio is south of 0.4, just the one eligibility criteria. George B. Zavoico - McNicoll, Lewis & Vlak LLC, Research Division: So in other words, most of the rapid progressors were also smokers?

In fact, it's about -- if I remember right, about 1/2 of the smokers who are exhibiting rapidly progressive disease. So about 1/2 of those are gone when you do that one criteria. George B. Zavoico - McNicoll, Lewis & Vlak LLC, Research Division: That helps your exclusion criteria immensely here. Okay, and finally, could you estimate at this time what the cost of the Phase III portion of GALAXY may be for patients [indiscernible]?

I don't think we've gotten into that level of specific details in any of our guidance. I think what we've said in previous discussions is that the cost of our program is in line with the cost of other randomized Oncology trials in this setting, which tends to be anywhere from $40,000 to $70,000 or $80,000 all included per patient.

Our next question is coming from Robin Davison from Edison Investment Research.

Just some quick ones, really, on the Phase III design. First of all, are you specifically excluding patients that have metastasis or -- obviously, they are likely to be patients who are rapidly progressing anyway. But there's other specific exclusion criteria?

No. There's no specific exclusion criteria, for example, on the liver bone mets or such. We, like most trials, have eligibility criteria around CNS mets. You'll find that in almost every lung cancer in its trial, and in fact, many other tumor types as well.

Right. Okay. On the KRAS and mutant and elevated LDH, are you going to look at that at all in the Phase III, or do we see those patients would be excluded by the -- again, by the poor prognosis?

No, they're not particularly excluded by the rapid progression criteria, and we are. We do have prespecified secondary endpoints around those populations, because they remain of very high interest to our investigators and to us. As you know, patients with mutant KRAS have very limited treatment options. There's quite a few drugs that are -- have either been shown to be ineffective there or not indicated there. Patients with high LVH die very quickly. Median survival is in the 4-month range. So there's very high need there. And so it remains a particularly high interest.

Right, okay. And just slightly, I notice reading currently the announcements that you're planning to allow 4 to 6 cycles of docetaxel, and then patients switch to or who kind of elect to switch to maintenance therapy. Now I'm just sort of wondering, with the -- whether you have any observations, really, I mean, obviously, the point where you expect to see separation after, sort of a bit, was certainly in the interim data of about 100 days. Patients would actually be on ganestespib monotherapy. Is there something we can read into that?

Yes. First, a bit of a caution -- first, the 4 to 6 cycle is general practice. I mean, there are centers that use 4 cycles of docetaxel, there's centers who use 6, and there's a few centers that use more. But it's correct. As in the Phase III will repeat the same exact regimen and schedule as the Phase IIb. And yes, when these patients complete the chemo, and patients just can't tolerate chemo in general, which is why it's usually terminated after the 4 to 6th cycle. They do go on maintenance, single-agent ganetespib. I think there are a number of the lung cancer docs on the panel at ESMO, which you can listen to, which talked about their experience with long-term maintenance. I mean, a number of them had mentioned their patients have 2- 10 or 12 cycles with ganestespib in maintenance and tolerating it well, and in their perception, deriving benefit from extended usage. So -- and then in terms of your question about separation or not separation, I think that'll become apparent as the data evolves and we have some more maturity. But again, we're talking about medians. And therefore, 4 to 6 cycles is 3, 4 months. And while the median PFS time typically is about 3 months, that does mean that 1/2 the patients are living longer than 3 months. So there's quite a few patients out there who are getting ganestespib maintenance as a single agent. And just as a reminder, there's no crossover in this study.

Our next question is coming from Ryan Martins from Lazard Capital Markets.

Can you talk about what may predispose a first-line patient to be a rapidly progressing patient? What are the key variables and factors that lead to patients eventually becoming a rapidly progressing patient?

This is something that's observed in, empirically in many trials, whether it's lung cancer or other tumor types. And sometimes, it's characterized in the literature as outright progressors. These are patients who show up, receive first-line therapy, then with chemotherapy. And on their first scan, 6 weeks later, the tumors have continued to increase or increased very significantly. And that just happens in a pretty substantial number of patients. They're just totally unresponsive. And exactly what is the underlying biology nature of that is something that quite a people -- few people are trying to figure out. But I think the way people think about it is that their disease is extremely advanced. There's probably not one driving oncogene mutation, but probably multiple mutations, multiple pathways have been activated, multiple resistance mechanisms have already been activated. So essentially, you have tumors that have evolved and have become very complex with many pathways turned on, both in terms of driving oncogenes and in terms of the pathways that drive drug resistance. Does that answer...

Yes. And just a decision to increase enrollment in the Phase IIb, was this post the FDA meeting or was it something that was planned, anyways, by the company as part of the protocol?

It was just enormously high demand from investigators. I mean, we have sites and investigators that are e-mailing us almost on a daily basis. As I mentioned, enrollment almost doubled in last month since ESMO, number one. Number two, a lot of the feedback from investigators, as you have a drug that's working, you want to get more data and really understand that very well. And in particular, these are patients that these 2 prespecified primary populations are the ones where have some underlying biology, where there could be an important mechanistic connection between the disease and the mechanism of the drug. And there's really nothing for patients for either of those 2. So there was just very high demand from investigators. It does create some options for patients with very limited alternatives. And it builds a much rich -- it just builds that much richer data package for us in 2013. We'll have 300 or more patients in a large, multinational, randomized setting for a drug that's showing very promising activity. And when you're there, having more data, it's an incremental relatively low cost, and it creates some very attractive options and some very useful information.

Okay, and maybe one question just to follow-up on the Phase II data that's already been presented at ESMO. We saw 2 cuts of the data, the hazard ratio. There was a slight deterioration in the hazard ratio from, I think, about 0.7 to about 5.57. Do you have an update on how that hazard ratio's looking? Has that bit -- obviously, you've had more events since then. Has that improved since then? Or can you talk about how that's been trending?

We haven't looked at the data since then, since ESMO. And I think I would be careful about positioning that as a -- you have one analysis, which has more maturity or had more follow-up time, and you have one analysis which was very fresh, where a lot of the patients have been recently enrolled. So that's the kind of thing you would typically see in the middle of a trial, and those kinds of things will disappear once patients have been enrolled and you have some good follow-up.

Okay. Are you going to have another follow-up of this data before you get the final data?

Currently, we don't have any plans to do another analysis before the first half of next year, which would be the final -- triggered by the final PFS endpoint.

And Safi, just one final one. Based on your cash levels and the impending, I guess, initiation of the trial. How are you thinking about, going forward, the cash levels, et cetera?

As you know, we have a board and large shareholders that are following our data extremely closely that understand that we are in the middle of what can be a very rapid and sizable inflection, given the amount of data that's coming up. And they're obviously very sensitive to dilution, being large shareholders who invested over many years. And because of that, we have no near-term financing plans that would be -- cause major dilution. We're looking at a number of other options. That being said, our board regularly evaluates all options, and we'll continue to do so.

Our next question is coming from Thomas Wei from Jefferies & Company. Thomas Wei - Jefferies & Company, Inc., Research Division: Just some more details on the Phase III trial, what exactly are you using in stratifying patients for randomization?

In general, we're using -- in general, what one wants to use is prognostic variables that are well known and well established and confirmed in marked data. And then we're using ECOG performance status, LDH, and I don't remember off the top my head exactly what else, you'll see more details, and I think in the press release on startup enrollment. Thomas Wei - Jefferies & Company, Inc., Research Division: And what about in terms of limits on, say, the number of patients from a particular geography?

No, we don't. That's not very common to place quotas. We don't have any planned quotas. We anticipate the geographical mix will be comparable to the Phase IIb. That has been working very well. One thing I would emphasize is, we've spent a lot of time, as I mentioned, with many of the top KOLs who've been involved, in essentially every pivotal lung cancer program. And the folks internally here have many, many, many years of experience in many pivotal trials at many companies for many different agents. And one of the main lessons is that one of the biggest Phase II to Phase III transition risks that has probably not been appreciated very well, in addition to when sponsors change dose or schedule a regimen or don't have a lot of evidence or insufficiently power going in. But one of the main ones that's probably not as appreciated very well is the operational risk. When Phase IIs are conducted in one country, it might be 10 or 20 academic centers. And then the Phase III is in 150 centers in 15 countries. And one of the things that we are -- we built into our planning 2 years ago was that before we ever get to the phase, we've already generated the experience in working in these 15 countries. So we're going to be in the same 15 countries, in many of the same centers. And the countries and the centers -- the investigators in these different countries and centers have already gotten the practice and the experiences in the real world, using our drug with patients. So there's no -- there won't be any practicing in the Phase III. A lot of that will have gotten -- already happened. So we feel quite comfortable in continuing the same countries, the same centers and the same geographic mix, essentially, seamlessly, between the Phase IIb and the Phase III. Thomas Wei - Jefferies & Company, Inc., Research Division: And so with the more than doubling of the number of centers, you still think that geographic mix that we've seen so far in the Phase IIb study, it will be reflected in the Phase III population as well?

Yes, yes. And there is a -- one of the other advantages of continuing the sort of tail-end enrollment, past the 240 patients, is that there are many centers that have approached us eager to participate over the past several months. And this gives them the opportunity to participate. So many of the centers that we are -- will eventually end up with the Phase III, will be signing up over -- you have either recently signed up or are signing up now, and will get some experience operationally before the Phase III. Thomas Wei - Jefferies & Company, Inc., Research Division: And what about blinding and independent central review of responses and progression?

This is not a blinded trial, it's survival endpoint. And in fact, that actually never even came up at the Head of Regulatory as a new crowd.[ph] That never came up in the FDA meeting. Independent central review for scans, if we would we would need to use PFS or response, we would certainly consider an independent review. But we haven't, that's not the primary endpoint. Thomas Wei - Jefferies & Company, Inc., Research Division: And then just lastly, can you tell us what the triggers are for the 2 interims and the final overall survival analysis?

That level of detail, I expect, would be the kind of thing that will be in the press release. I don't actually have the numbers with me.

Our next question is a follow-up from Jim Birchenough from BMO Capital Markets.

My first question is on the comment you made regarding the responding brain metastasis. And I know this is an end of one, but were you surprised by that? And do you think this is sort of an autocrine or paracrine function, i.e., as ganetespib across the blood-bearing barrier, assuming that it's been breached by a lesion? Or is the systemic effect of Hsp90 inhibition sufficient to have a paracrine effect in what is considered a protected area within the brain? And then my second question is, are you planning to validate any other biomarkers than you've discussed in the Phase III trial?

For both those questions, I'll turn it over to Iman.

Okay, for the first question around the brain metastasis, we actually, what you've seen is that these pieces of brain metastasis, as you know, they will have breached an intact blood-bearing barrier which facilitates the cross of ganestespib into these lesions and really exerting their -- its effect. Some other lesions, even if one mostly have not been really successful in, in really -- successfully affecting brain metastasis. So I think, for us, this is extremely exciting. I really mean to see that you have a clear effect in terms of tumor shrinkage and in terms of reaching to a little of the disease stabilization in the brain by ganestespib. And it could be a function, clearly, of the ability to cross the blood-bearing barrier, as well as a specific activity of Hsp90 degradation. In terms of the second question, regarding the biomarkers, of course, we have extreme interest to understand in more details the patient population that we have enrolled in our Phase II trial. And we'd be actually doing a series of biomarker analysis to be able to understand the characteristics and the level of activity, correlation with the clinical outcome with the ganestespib combination treatment.

But at this stage, you don't envisage that data making it onto the label?

Not for this program. I think if you look at where we are in Phase IIb and the types of -- admittedly, it's interim data. But if you look at the historical context of hazard ratios that have been seen in lung cancer, what we're seeing already is pretty spectacular and it's very encouraging. And if we can -- that's a very good reason and rationale to go forward with the program as it is, with the exclusion criteria and the focus on the patient group that's most likely to benefit as we've defined. If something is simple and works and it's easy to apply and is objective and reliable. That's terrific, like a clinical characteristic. As you know, biomarkers require biopsies, they require vendors, they require -- there's some level of analysis that raises-- that can be difficult. There's the tumor heterogeneity questions. So when you have something that's a clinical characteristic or a CRM marker, that has many advantages.

That concludes the question-and-answer session. I will now turn the conference over to Dr. Bahcall for closing remarks.

Thank you everybody for your time and attention today. That ends this call.

Ladies and gentlemen, that concludes today's conference. You may now disconnect, and have a wonderful day. We thank you for your participation today.