Madrigal Pharmaceuticals Inc (MDGL) 2014 Q1 法說會逐字稿

Good day and welcome to the Synta Pharmaceuticals first-quarter 2014 earnings conference call. Today's conference call is being recorded and webcast.

At this time for opening remarks, I will turn the call over to Steve Bernitz, Senior Vice President of Corporate Development at Synta Pharmaceuticals. Please go ahead, sir.

Good morning and thank you all for taking the time to join us today. With me are Keith Gollust, Chairman of the Board of Directors and Executive Committee; Vojo Vukovic, Chief Medical Officer; and Keith Ehrlich, Chief Financial Officer.

This morning we issued a press release that reported results for the first-quarter 2014. We also issued a press release announcing the final results of the GALAXY-1 trial and have prepared a slide set summarizing these results. Both releases and the slide set can be found on our website.

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 at www.Syntapharma.com.

I will now turn the call over to Keith Gollust.

Thank you, Steve, and thank you all for joining us today. All of you who follow development-stage biotech companies know that it is a long and financially challenging road that must be followed before success can be attained. Our experience at Synta has been consistent with that.

However, with today's announcement of GALAXY-1 final results, we are one important step closer to attaining that success. The results we announce today demonstrate a significant benefit in a randomized population of chemosensitive patients with previously treated non-small cell lung adenocarcinoma. This is a very challenging cancer to treat, and the survival benefit observed in this final analysis of the data is very encouraging.

I want to point out that this benefit in a subpopulation of 177 patients out of 253, while a prespecified analysis, was not a prespecified endpoint. However, our review of early GALAXY-1 data in 2012 provided such a strong signal that we began planning a Phase 3 trial at that time, with overall survival in chemosensitive patients as the primary endpoint. As the data matured the signal was confirmed, and our decision to start the Phase 3 in 2013 was validated.

In light of these results, the operational experience gained from GALAXY-1, the incorporation of regulatory feedback into our Phase 3 protocol, the increase in study size announced in March, and the positive feedback we have gotten from key opinion leaders, we believe that the Phase 3 GALAXY-2 trial has excellent prospects for success.

Given the financial commitment required to conduct a large Phase 3 trial, we have undertaken a comprehensive review of our budget for 2014 and beyond, to make certain that our cash burn is manageable and is consistent with our available resources. Our philosophy is to treat every dollar like a best friend that we don't want to say goodbye to unless absolutely necessary.

An example of this philosophy in action is the decision which was announced this morning to close the ENCHANT-1 study. The promising data from that study provided the basis for our inclusion in the I-SPY 2 study, which will explore ganetespib in combination with standard-of-care therapy in a large, randomized proof-of-concept trial. Transition to the I-SPY 2 trial renders continued investment in ENCHANT unnecessary.

Controlling the burn rate is only one half of balancing the budget. The other half is attracting the capital we need.

So far this year we have added over $32 million through the issuance of equity. Our plan for the balance of 2014 is to combine an opportunistic use of the ATM that we have put in place with other forms of financing, in order to end the year with the cash resources necessary to take us through the final analysis of GALAXY-2 in the first half of 2016.

In addition, it is our intention to broaden our partnership discussions under the leadership of a new CEO, which, if successful, would further strengthen our balance sheet. I would now like to turn the call over to our Chief Medical Officer, Vojo Vukovic, who will discuss the results of the final analysis of GALAXY-1.

Thank you, Keith, and good morning. I am pleased to present today a summary of final analysis results from GALAXY-1, a large Phase 2b international randomized study of ganetespib and doxetaxel versus docetaxel alone. We locked the database last week, and the data is still hot off the press.

Today we will present a top-line summary of key efficacy and safety data. A more comprehensive set of results will be published later this year. The top-line summary is posted temporarily at the Company website as a slide set at www.SyntaPharma.com.

I will start here with slide 3. Ganetespib is a next-generation Hsp90 inhibitor that has shown synergy with taxanes and the ability to inhibit angiogenesis and metastasis both pre-clinically and clinically. Ganetespib has shown single-agent activity in several types of tumors, including non-small cell lung cancer and breast cancer, and has been well tolerated both as monotherapy and in combinations.

The next slide, slide number 4, is discussing the concept of chemosensitivity in non-small cell lung cancer. Patients who respond to first-line treatment are more likely to respond to second-line treatment and generally have a better prognosis than refractory patients who fail first-line treatment or relapse quickly following first-line treatment. To ensure balance in clinical trials, patients are often stratified by chemosensitivity.

In GALAXY-1, for operational reasons we have used time since diagnosis of advanced disease as a surrogate for chemosensitivity. We could have used another surrogate, such as tumor response in first-line. But we decided on time since diagnosis since operationally it is easier to track and verify.

That brings us to slide 5, which lists the key populations for GALAXY-1. Elevated LDH and mutant KRAS populations have a strong scientific rationale supported by preclinical and/or clinical data. The chemosensitive population showed the strongest signal of activity and is supported by preclinical data showing further sensitization of cancer cells that are chemosensitive via down-regulation of key proteins in the mitochondrial apoptosis pathway.

The GALAXY-1 study design is shown on slide number 6. Patients with advanced non-small cell lung cancer with adenocarcinoma histology and one prior systemic therapy were randomized 1-to-1 to receive either docetaxel alone or docetaxel and combination with ganetespib. Patients were stratified by four prognostic factors: ECOG performance status; time since diagnosis of advanced disease; baseline serum LDH; and smoking status.

Co-primary endpoints of the study were PFS in elevated LDH, and PFS in mutated KRAS patients. Key secondary endpoints were PFS and OS in adenocarcinoma patients.

Slide 7 shows the baseline characteristics of the adenocarcinoma population. As you can see, the two study arms were well balanced with respect to the key demographic and prognostic factors listed here, including age, gender, ECOG status, smoking status, LDH, and others. A good balance of prognostic factors is strictly important for a valid trial. One thing to note is that approximately 40% of patients come from Western Europe or the USA and Canada.

Slide 8 shows that the chemosensitive patient population is also well balanced. This is a large group of patients totaling 177 patients, or approximately 70% of the total adenocarcinoma population.

As we see numerical differences in some parameters, we have utilized a standard approach to adjustments for imbalances, the Cox proportional hazards model. We will talk a bit more about adjustments when we discuss [askit's] results on the next couple of slides. As in the adenocarcinoma population, approximately 40% of chemosensitive patients are from Western Europe or North America.

Slide 9 shows one of the two efficacy summary tables. Here you see duration and comparisons of median OS and PFS for the four key populations: elevated LDH, mutant KRAS, chemosensitive patients, and adenocarcinoma patients.

Data maturity ranges from 75% to 83%, so we are looking at a mature, robust data set. Differences in median PFS and OS are more substantial in the adenocarcinoma and chemosensitive populations and are likely clinically meaningful.

For example, a 2-month increase in median PFS, along with a 3.6-month increase in median OS in the chemosensitive population is clearly clinically meaningful in second-line advanced non-small cell lung cancer. In the adenocarcinoma population, both median PFS and OS were improved in combination patients, but improvements were smaller: 1.3-month increase in PFS, 1.8-months increase in median OS.

Slide 10 shows OS and PFS hazard ratios for key patient populations: elevated LDH, mutant KRAS, chemosensitive, and adenocarcinoma. As in the previous table, the strongest signals of efficacy were seen in the chemosensitive and adenocarcinoma populations.

Unadjusted OS hazard ratios were 0.71 and 0.87 for chemosensitive and adenocarcinoma patients, respectively. Adjusted OS hazard ratios were 0.69 and 0.84.

PFS hazard ratios were similar, 0.75 and 0.85 unadjusted; and 0.74 and 0.82 after adjustment. Summarized, these results confirm results from preceding two interim analyses, indicating the strongest signal of efficacy in the chemosensitive population.

On slide 11, we can see the Kaplan-Meier plot for overall survival in the chemosensitive population. While some patients are still in follow-up, the data maturity is 75%. The OS curves are nicely separated. And the hazard ratio of 0.71, or 0.69 after adjustment, along with a p-value of 0.019 is clearly very encouraging.

Slide 12 shows the Kaplan-Meier plot for progression-free survival for the chemosensitive population. The unadjusted hazard ratio is 0.75, or 0.74 after adjustment, consistent with the OS result, and confirms that there is an efficacy signal in chemosensitive patients.

Slide 13 shows the Forest plot of subgroup analysis for the chemosensitive population. The first visual impression is that virtually all subgroups benefit from combination treatments. Rather than going through the entire set on this slide I would like to highlight two important points.

First, at the top of the first plot, you can see OS results in patients with and without KRAS and EGFR mutations. While the patient numbers are small, this data suggest the combination works in chemosensitive patients, irrespective of KRAS and EGFR mutation status.

Secondly, at the bottom of the slide you can see OS results broken down by region. Consistent with previous analysis, the OS results in Western patients are very encouraging.

Slide 14 shows the table with key adverse events. In summary, the safety profile in GALAXY-1 is consistent with earlier observations and reports.

Most frequent adverse events are in the gastrointestinal tract and include diarrhea, nausea, vomiting, and stomach pain. You see a slight increase in number of AEs, or adverse events, in patients treated with the combination. In most cases, those adverse events are mild to moderate and can be effectively managed or prevented with standard supportive care.

At the bottom of the table we have listed medically significant adverse events for this patient population, which include mucositis, neutropenia, hemoptysis, pulmonary embolism, and febrile neutropenia. As with general adverse events, we see a slightly higher frequency of adverse events in patients treated with the combination.

You should also keep in mind that combination patients are on treatment approximately 50% longer than control patients. Considering the higher drug exposure and treatment duration, some increase in adverse events is expected.

Finally, consistent with prior reports, visual impairment frequency is very low. Only one case of mild and transient visual impairment was reported thus far.

Moving on to the next slide, slide number 15, these are the conclusions of the GALAXY-1 study. The combination of ganetespib and docetaxel improved overall survival and progression-free survival in the chemosensitive patient population.

We did not see evidence of activity in KRAS mutation patients. There was a positive trend in overall survival in patients with elevated LDH.

Overall, ganetespib in combination with docetaxel was well tolerated. And finally, final results from GALAXY-1 validate the selection of chemosensitive patient population for the Phase 3 GALAXY-2 study.

Going forward our emphasis is on execution of the Phase 3 GALAXY-2 trial, which is shown on slide number 16. The GALAXY-2 study incorporates all lessons learned in GALAXY-1.

Final results from GALAXY-1 confirm the key elements of GALAXY-2 design, such as the choice of patient population, the choice of optimal certification factors, and the choice of primary endpoint. GALAXY-2 is active and enrolling, and we remain on track regarding the interim analysis in the second half of 2015 and final analysis in the first half of 2016.

This concludes the top-line summary presentation of the final GALAXY-1 results. I will now turn the call over to Keith Ehrlich for financials.

Thank you, Vojo, and good morning, everyone. There were no revenues recognized in the first quarters of 2014 and 2013. Research and development expenses were $17.6 million for the first quarter of 2014 compared to $16.4 million for the same period in 2013. General and administrative expenses were $5.3 million for the first quarter of 2014 compared to $3.9 million for the same period in 2013, the difference being substantially accounted for by the separation agreement with our former CEO.

The Company reported a net loss of $23.6 million or $0.28 per basic and diluted share in the first quarter of 2014, compared to a net loss of $20.7 million or $0.30 per basic and diluted share for the same period in 2013. As of March 31, 2014, the Company had $78.8 million in cash, cash equivalents, and marketable securities, compared to $91.5 million in cash, cash equivalents, and marketable securities as of December 31, 2013.

During March and April 2014, the Company sold approximately 6.6 million shares of common stock under its at-the-market sales agreement for $27.3 million in net proceeds. In April 2014, the Company sold approximately 1.25 million shares of its common stock for $5 million in net proceeds and a registered direct offering to an affiliate of a Director who is our largest shareholder.

In May 2014 the Company entered into a new at-the-market issue and sales agreement with MLV under which the Company, at its option, may issue and sell shares of common stock having an aggregate offering price of up to $40 million from time to time through MLV as its sale agent. The shares would be sold pursuant to the Company's effective shelf registration statement on Form S-3.

Based on current operating levels, the Company expects its cash resources of approximately $78.8 million at March 31, 2014, plus the $23 million in net proceeds from common stock sales during April 2014, will be sufficient to fund operations at least into the first half of 2015. This estimate assumes no additional funding from new partnership agreements or equity financing events, and that the timing and nature of certain activities contemplated for the remainder of 2014 and 2015 will be conducted subject to the availability of sufficient financial resources.

I will now turn the call back over to Steve.

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

(Operator Instructions) Thomas Wei, Jefferies & Company.

Thanks. Just a couple of follow-ups on some cuts of the data that you have shown us before in the past. First, just on -- if you took out the Eastern European countries that you spoke about at World Lung, what would the final numbers look like on over analysis -- on overall survival there?

Then also, I know it would be small patient numbers, but if you looked at just the not-chemosensitive subgroup, so those with a shorter than 6-month window from diagnosis, can you share with us what the median overall survival was in each arm of that subgroup?

Okay. I will take that question. Hi, Thomas; this is Vojo. The data we presented today is the data that we have currently in hand. As we mentioned, the data base lock occurred last week, and we will complete additional analysis in the coming days.

So we can't really comment on your question about exact hazard ratios in Western patients versus Eastern patients. Also more detailed analyses in the other subgroups are not completed yet. The one thing that I know and I can share with you is the adjusted hazard ratio for survival in the less than 6 months population is now exactly 1.

That's very helpful. Maybe just a couple more then. If you had completed the HER2+ cohort in ENCHANT-1, how much would that have cost?

And then just when you made the comment about 50% longer treatment duration in the ganetespib arm in GALAXY-1, did you mean 50% higher docetaxel intensity? Or is that really the same docetaxel intensity in both arms with some extra ganetespib in the treatment arm? Thanks.

Thomas, this is Keith, Keith Gollust. I will answer the first question and Vojo will answer your second question.

Our budget for this year for ENCHANT study was $2.4 million. Some of that has been incurred, but the majority of that amount will be eliminated from our budget this year. Vojo?

Okay, so when we refer to 50% longer duration of treatment, what we meant by that is the median number of cycles for the control arm is four; and the median number of docetaxel-containing cycles for the combination is six. In addition to that, a proportion of the combination patients has also received ganetespib monotherapy maintenance.

Great, thanks.

Mike King, JMP Securities.

Good morning. Thanks for taking my questions and appreciate all the detail that you have provided on GALAXY-1. First question I wanted to ask would be for Vojo, if you could remind us what the powering assumptions are again for GALAXY-2.

And if you feel that, given the final data at GALAXY-1 with hazard ratios in the final stats on survival and PFS, that you've got enough headroom for the inevitable statistical deterioration that you typically get when you go to larger study sizes.

Okay, so I am going to try to answer the question the best we can. The powering assumptions for GALAXY-2 assume an approximately 87% power to detect a hazard ratio of 0.75. That means that we are considering the current final GALAXY-1 data, which points to an approximate hazard ratio of 0.70 range, we feel that we have some margin there.

In addition to that, I will just remind you, Mike, that a power of 87% detect hazard ratio 0.75 also means that we have a considerable amount of power to detect a higher hazard ratio. And then finally, as you will probably remember from our last call, the total number of patients for GALAXY-2 is estimated to be approximately 850.

Right.

Giving us the minimum of 700 patients that meet the EGFR/ALK double-negativity criteria. And that is a minimum; we actually expect to see more than 700 patients. And the more patients we have over 700, the higher the power is of the study.

So these are the factors that give us some good comfort level about the powering assumptions for GALAXY-2.

Remind me, Vojo. The stipulation of exclusion of Eastern European centers was made before GALAXY-2 started?

It was made actually after the GALAXY-2 started, so we do have a small proportion of patients. We currently estimate this to be less than 10% of the overall patient number at the end of the study from Eastern Europe.

Okay. But those will be part of the final database analysis?

That's right. And when I say Eastern Europe, I just want to clarify, it is really Russia and Ukraine.

Okay.

Because in GALAXY-1 it was in Russia and Ukraine that we have detected some differences between those patient populations and other patients in the study.

Right. Then just on the interim analysis at the end of 2015, that will be after what proportion of events?

So, the time, let me just remind the timing for the two interim analyses. They are projected for the second half of 2015, not necessarily the end of 2015.

Actually, the first one will happen closer to midyear -- second half, but closer to midyear. And the other will happen later in the second half.

I don't believe that we have previously guided on the exact number of events that are needed. We have actually -- what I can say here is that we have discussed this in detail with the regulatory agencies, including the FDA, and we have actually exceeded what the FDA requires as a minimum number of events.

Typically the FDA requires [50%](technical difficulty). We have exceeded that number.

So the power for this first interim analysis will be quite good, more than what the FDA number requires. And consequently, the power for the second interim will be even higher than that.

Okay, thanks. Then just further on that, the filing, an NDA filing would be done when? Or will that depend upon the outcome of these interim analyses? Or would you prefer to file on the final analysis?

Well, the interim analysis has been designed and agreed upon with the FDA to be efficacy interim analysis. They are fully powered.

And should they exceed the prespecified threshold of activity, we can immediately stop the study, based on the DMC recommendation, consult with the agency, and file immediately. And we will get ready to file immediately, you can trust me.

Right. All right. Then switch quickly to breast cancer, and then I will get off. Obviously, there is a -- it is commendable that you are going to be saving money by discontinuing ENCHANT to focus on I-SPY. However, the market opportunity and potential for what you might have achieved in ENCHANT is pretty different than that of I-SPY.

So I just wonder if you might comment about the strategic implications of that pivot, in addition to the financial savings. Thank you.

Hi, Mike. This is Iman. I think it's -- we are really all very excited about what we have achieved so far in the ENCHANT trial, and the fact that we have achieved the objective of this study, and it was an enabler to start the I-SPY 2.

I think this is our immediate strategic focus in breast cancer. Clearly there are very interested investigators who are keen to explore other avenues for ganetespib in different investigative initiatives and trials.

So we will continue to keep open-minded, but certainly I totally agree with you that the evidence that we have generated in the ENCHANT trial so far is an opportunity for ganetespib for subsequent clinical development.

Yes, I was going to add to that, Mike, that as a practical matter we think that the greatest benefit of the use of ganetespib is in combination therapy, and ENCHANT was a monotherapy study. So while the market opportunity that you are referring to is still there, I think at some point in the future we should be able to explore that through a combination study.

Maybe just another addition to that regarding market opportunity. The I-SPY 2 is set in the neoadjuvant treatment setting, which is a very substantial market segment in breast cancer.

Yes. No, I realize that. And pathologic complete response could be a quick route to market; but it is relatively small compared to metastatic disease. That is all I am saying.

Daniel Brims, Brean Capital.

Hi, thanks for taking my question. So, winding down ENCHANT-1, when do you see that being completed? And when would we probably be seeing the complete data from that study?

Yes, we are in the process of clearly winding down, and we would expect to present final data towards end of this year.

Okay. Will GALAXY-1 be at ASCO this year? Or is it going to be at a later meeting?

GALAXY-1 data, the final data will not be presented at ASCO. Our plan is to publish this data very quickly in a scientific journal, and then we will decide at a later point about presentation at scientific meetings, perhaps later this year.

Okay, thank you.

Graig Suvannavejh, MLV.

Great. Thank you for taking my questions; and good morning, everyone. Just wanted to get a sense just on the current CEO search and any comments you could make as an update on how that process goes.

(technical difficulty) search, what I can say about that is that the search is proceeding as planned, and as soon as we have a new CEO there will be an announcement. Other than that we are not going to attempt to characterize actual discussions.

Then if I could ask another question, just -- I think many of us left the AACR conference very enthusiastic about the HDC platform that you have. And I was wondering if you could give us -- if there were any updates with respect to progress you are making with your HDCs.

The HDC platform is certainly scientifically very interesting, and we continue to maintain a strong interest, and we have plans for developing HDC. But as Keith has mentioned in his opening remarks, we are focusing a lot of effort, a lot of attention within the Company making sure that we do everything we can to lead GALAXY-2 to success.

Great. My last question. I realize that you have -- you feel comfortable with the timing around the interim analyses in GALAXY-2. But is there anything you can comment about the speed of enrollment and whether there are any -- anything that could potentially make that enrollment go faster or slower?

We have traditionally not given specific guidance in terms of operational activities. I think all I can say right now is that, as we confirmed our guidance regarding the data readouts, what that also means is that we remain on track operationally to recruit the trial.

We want to make sure that we, first and foremost, do a really good job in working with the right sites and selecting the right patients, so that we really maximize the probability of technical success for the study. That is our number-one priority, and I think we are very well on track on that.

We are also on track with patient recruitment; doing very well on that. And so that is really all I can say about the operational activities right now.

Great. Thank you so much for taking the questions.

George Zavoico, H.C. Wainwright.

Hi, good morning, and thank you for the detailed analysis of GALAXY-1. I have a question about perhaps the -- about the financing. Taken together, you have about a little over $100 million at the end of the quarter; and the R&D expense presumably reflects mostly GALAXY-2.

You project it to somewhere between a year to a year and a quarter to go from now. What portion of that is actually the GALAXY-2 budget?

Well, it is hard to break it down because the GALAXY-2 budget is extended over a period of years. As a rough estimate, the total cost of GALAXY-1 is in the neighborhood of $100,000 per patient; but that expenditure is spread out over a period of almost three years.

So as we project our cash needs between now and final analysis of GALAXY-2, which would be expected to occur in the second -- in the first half of 2016, you can take a look at our current burn rate and project it out over that period of time to get a pretty rough idea of how the numbers will work out.

Okay. You didn't talk at all in this presentation about the rationale for excluding the ALK patients. Could you go over the data you have in support of that decision, Vojo, please?

Sure. The rationale to exclude ALK and EGFR mutation-positive patients has really not so much to do with the activity of the drug -- or the combination, I should say. Because we believe that activity of the combination is comparable in this patient population relative to other patients.

The reason why we decided to exclude these patients really occurred after discussion with regulatory agencies who, for one, insisted that such patients are identified and have received standard-of-care treatments prior to entering our study. So for example in, ALK inhibitors or EGFR inhibitors.

Now, in global settings that is obviously a complication because these standards of care are not universally applicable. So what's standard of care in US may not be standard of care in the Czech Republic. So in order to get to the common denominator, we decided to exclude these two patient populations.

Another additional benefit of that decision is that we will increase the homogeneity of the patient population, because patients with ALK and EGFR mutations have different prognosis compared to other lung cancer patients. And so for that reason I think increased homogeneity is definitely a good thing for GALAXY-2 and gives us actually a better shot at getting a positive result.

Okay. In that regard, you are talking about the differences in different countries, is the lead-in, the first-line therapy the same across the whole -- all the countries that are involved in GALAXY, then?

Yes. So for GALAXY-2, we have an inclusion criteria that the patient must have received platinum-based chemotherapy in first line. So we really want to make sure that the patient population for GALAXY-2 is very homogeneous.

But the -- well, the platinum-based, often that is given also in combination, isn't it? So is the combiner with the platinum-based going to be consistent, or is there some variability there as well? Or is someone going to get monotherapy?

Yes, so platinum is a backbone; and then to platinum you add different drugs. And there is a seminal paper by Joan Schiller in New England Journal of Medicine which has actually shown equivalency of four different commonly used platinum regimens. So as long as they have platinum in the combo that is given to patients, they achieve all very comparable outcomes.

Okay. Thank you, Vojo. Last question, about AML-18 and AML-19. What is the difference?

AML-18 I see compares ganetespib with two other agents in combination. I know most of the funding for this trial is going to be external to Synta, but what is the design of AML-18 and AML-19 and the rationale there?

So the series of the AML trial -- actually there is three of them. One is called AML-LI-1, and then there is 18, and then the third one is 19.

The design is very similar. These are all trials in patients with AML in the first-line treatment setting. What's the difference is the patient population.

In the LI-1 study, that's approaching a go/no-go decision midyear, the patient population are elderly patients unfit for induction chemotherapy. In the AML-18, the patient population are elderly patients fit for induction chemotherapy. And then in AML-19, it's younger patients that are fit for induction chemotherapy.

So we have basically with these three trials the entirety of the AML indication set.

For LI-1, you are in combination with low-dose Ara-C. Is that the same for 18 and 19?

No, no. Induction chemotherapy is different. They will be using doxorubicin. They are also using other drugs as well. Ganetespib is added to that regimen.

Okay, all right. All these are run by -- it is Alan Burnett, right?

Alan Burnett is the most senior person there. He works with a team; and for the individual studies, different people are the PIs. But Alan Burnett is really the person who is orchestrating this whole thing, and his institution acts as the sponsor.

I should just mention that these trials are entirely funded by third parties, and all Synta is doing is really just providing the drug.

Okay. Thank you very much, Vojo and Keith. Thank you.

Brian Klein, Stifel.

Great, thank you for taking my questions. I wanted to dig into the data a little bit further.

On slide 9, I noticed that between the entire adenocarcinoma population versus the chemosensitive population, you have a significant decline in overall survival; but the PFS seems to be consistent. I was just wondering if you could give a little bit more clarity on what you suspect is driving that drop when you move from the all-comer population to the chemosensitive population.

Well, I think the simplest way I think how to explain the difference between adenocarcinoma and chemosensitive is the fact that when we look at the two subpopulations for adenocarcinoma -- one is chemosensitive, the other is refractory, defined as less than six months since diagnosis -- all of the effect that we see in adenocarcinoma is really contained in the subpopulation of chemosensitive, which represent 70% of the patients.

In the refractory patients, as I mentioned before, the OS hazard ratio is exactly 1. That means in the 30% of that adenocarcinoma population, the refractory patients, there is basically no effect of the combination.

So in chemosensitive, you have a pure signal; and in adenocarcinoma you have a signal mixed with 30% of patients which experience no benefit.

Right. I am referring to the placebo arm -- or not placebo, but the docetaxel-alone arm. If what you are saying is true --

Oh, you mean difference in median? 7.4 and 8.4?

Correct. Yes. Why do you see that drop?

When you study in detail the Kaplan-Meier plots, you see that is just basically by chance. The 7.4 number, if you analyze it, that 55% instead of 50% or 60%, it would be a different number. It is just the random fluctuation.

When you look at the -- well, actually you can't see here. But when we look at the confidence intervals we see that the confidence intervals for the control, for the control in the chemosensitive, for the control in adenocarcinoma, almost completely overlap.

Okay. I guess one interpretation would be that if you use the 8.4 months for the Taxotere-alone arm, instead of 7.4 you might not have reached statistical significance. And if your claim that there was no benefit from the rapid progressors, it just seems a little strange to me that you would see that decline.

But I guess my second question is in regards to the adjusted versus the unadjusted. Obviously it looks like there is a benefit in the chemosensitive arm when you adjust the numbers.

And if I look back at the baseline characteristics on slide 8, it looks like the docetaxel-alone arm was actually a more advanced or more diseased population, in terms of both stage at initial diagnosis -- if you look at the percentage of patients that received prior chemotherapy -- and if you look at elevated LDH. All of those numbers are skewed towards the Taxotere-alone arm.

So I am wondering. If you adjust the patient population, I would assume at least that the Taxotere-alone arm would actually improve and your adjusted number would actually decline; but here you are showing a benefit. So I am just wondering if you could explain that as well.

Yes. So what we show on slides 7 and 8 are based on some of the baseline characteristics of the adenocarcinoma and the chemosensitive populations, not all. When you look at slide number 10, you see in the footnote, we show you the details which parameters are used to do the adjustments using the Cox model.

I think in a simplified version, the way how it works, we're testing all these parameters individually; determine which one carry prognostic or predictive value; and those remain in the multi-variant analysis.

So this analysis has been in use for 40 years. It was standard in academic and industry trials, and it is used exactly for this purpose that you have mentioned. When there are some numerical differences between the treatment arms, it is a very precise, widely adopted method to adjust for such imbalances.

Right. I guess I am just wondering why, if the docetaxel-alone arm seems to have worse patient characteristics, when you adjust them, when you try to balance those two arms, at least I would expect that the adjustment would lead to a closer -- or a hazard ratio that approaches 1 as opposed to going the opposite direction, once you make those adjustments. That is what I am trying to understand.

Okay, so I will turn this question over to Ilker. But before I do that, you see here four or five parameters in the table; but there is additional parameters that we analyze that are part of the model. So Ilker, do you want to maybe just walk us through this question?

Yes, it is hard to explain a difficult concept over the phone, but I will try my best. First, the adjustment in the Cox regression uses multiple variables. So you cannot think of like very simply one factor.

Assume that treatment has more normal LDH patients. Those patients have better prognosis, we think, right?

But what if all those patients are smokers? So you cannot really look at one factor and say this is imbalanced in the one group.

The second thing is that it is not only the number of patients, but also what the prognosis for those patients are. Let me give you an example.

Let's say the treatment group ends up with more normal-LDH patients than control group. Look, LDH patients, normal-LDH patients have better prognosis, right? So since it has more normal-LDH patients, you may think the treatment group has an unfair advantage over control group because of those extra normal-LDH patients.

But what if the normal-LDH patient treatment group live much shorter, meaning have worse prognosis than the ones in control? The imbalance in this case will cancel out, right, because you have more normal-LDH patients?

But like I said before, if they are all smokers, it is going to cancel out. So, the Cox regression looks at all these things and adjusts accordingly.

Okay, thank you. Then just the last question, please. The values that you show for the hazard ratios and the p-values are all one-sided. So I just want to confirm that in the GALAXY-2 Phase 3 study they are going to be two-sided analyses.

Yes, absolutely. We should not forget, GALAXY-1 is a large, randomized Phase 2 study and the stats around that is designed accordingly.

GALAXY-2, on the other hand, is a fully powered, very large, Phase 2 trial and of course it will use two-sided analysis. And again, the statistical details of the planned analysis have been discussed and agreed upon with the agencies.

Okay, thank you.

Thomas Wei, Jefferies.

Thanks for taking the follow-up. I guess I just wanted to see if you were willing to provide any more granularity on when you might complete enrollment in GALAXY-2.

And then also if, based on what you have seen so far from enrollment, anything that you would point out that looks different from the GALAXY-2 patients relative to GALAXY-1, things like the geographic breakdown, whatever you think might be notable.

Hey, Thomas. Again, unfortunately, we are not providing details of the operational activities. All I can reiterate is that we feel highly confident that we will complete the enrollment in time to generate the data for interim analyses; and we gave you some guidance when those are.

When it comes to enrollment patterns, we are very happy and very confident in what we have seen so far operationally in terms of opening up the number of sites. And then also, of course, the enrollment of patients themselves is actually going slightly ahead of our plans.

We are trying to really use the lessons learned from GALAXY-1 in terms of patient selection and all the study procedures to optimize the process, and I think it is working. We see, I think, pretty good recruitment in Western countries; but it is quite similar to what we have seen in GALAXY-1. In GALAXY-1, just to remind you, we had approximately 40% of patients from the West.

So overall, I think we are very happy with the way -- how the operational aspect of GALAXY-2 is progressing.

Thanks. That's very helpful.

Mike King, JMP Securities.

Thanks for taking the follow-up. I just had a quick question on the AML trials. Is there any possibility that any part of that data could be presented at this year's ASH? Or is that a too-aggressive timeline?

Yes, that's a very good question, Mike. We have guided to the AML-LI-1 go/no-go decision point for midyear. And the timeline for ASH submissions is traditionally the second week of August.

So there is a possibility, obviously, if we have the data set. But because this is a trial that is not run by us -- it is run by the academic consortium -- the decision about publication rests with them. If we can, we will encourage them, but we obviously don't know the outcome of that.

Okay, thanks.

Thank you. That concludes the question-and-answer session. I would now like to turn the call back over to Mr. Gollust for closing remarks.

I'd like to thank everyone on the call today for your time. If you do have follow-up questions, we are all here, and we are available to take your calls. And I look forward to speaking with many of you in the future. Thank you.

Ladies and gentlemen, this concludes today's call. You may now disconnect your lines.