Gilead Sciences, Inc. (GILD) Q2 2012 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by, and welcome to the Gilead Sciences Second Quarter 2012 Earnings Conference Call. My name is Chanel, and I'll be your conference operator today. [Operator Instructions] And as a reminder, this conference call is being recorded. I would now like to turn the call over to Patrick O'Brien, Senior Director of Investor Relations. Please go ahead.

Thank you, Chanel. Good afternoon, everyone. We issued a press release this afternoon providing earnings results for the quarter. The press release and earnings slides are available on our website. For our prepared remarks and Q&A, I am joined by the Chairman and CEO, John Martin; our President and COO, John Milligan; our EVP of Operations and CSO, Norbert Bischofberger; our EVP of Commercial Operations, Kevin Young; and our CFO, Robin Washington. Before we begin our formal remarks, we want to remind you that we will be making forward-looking statements that contains certain assumptions, risks and uncertainties that are beyond our control. These risks include the possibility that our actual financial results may differ materially from the revised guidance we are presenting today. The possibility of unfavorable results from our clinical studies, including those evaluating the GS 7977, and GS 5885 in various patient populations and the possibility that we may be unable to compete our clinical studies and regulatory filings or obtain regulatory approvals in currently anticipated timelines. A description of these risks can be found at our latest SEC disclosure documents in our recent press releases. Gilead does not undertake any obligation to update these forward-looking statements made during this call. We will also be using non-GAAP financial measures to help you understand our underlying business performance. The GAAP reconciliations are provided in our press release, as well as on our corporate website. I would now like to turn the call over to Robin Washington.

Thank you, Patrick, and thank you all for joining us today. We are very pleased with our strong operating results for the second quarter. Product sales were $2.3 billion, an increase of 14% year-over-year and 5% sequentially. The U.S. contributed $1.4 billion of product sales, up 21% year-over-year, driven by strong increased demand across all therapeutic areas. In the U.S., our early release of Ryan White Federal funds and the desire by states to reduce patient wait lists have contributed to strong year-to-date non-retail performance. This may impact ADAP purchasing patterns in the second half of 2012. The state's efforts to reduce ADAP wait lists are encouraging, resulting in a 60% reduction since the beginning of 2012. The current wait lists are at our level not seen since mid-2010. Our U.S. performance also includes a new milestone for Letairis, which exceeded sales of $100 million in the second quarter. Europe contributed $786 million of product sales, up 3% year-over-year, driven by increased demand. From an expense standpoint, operating expenses increased to $729 million, up 24% year-over-year due primarily to the continued advancement of our product development pipeline. Turning to other financial highlights. Operating cash flows continue to be strong. During the quarter, operating cash flows were $1.3 billion, which includes the collection of $460 million of past-due accounts receivables in Spain, contributing to a 15-day improvement in our DSO, highlighted on Slide 14. $350 million of bank debt was repaid this quarter and since January 2012, a total of $700 million has been repaid, which further deleveraged the balance sheet and will help achieve a targeted debt to EBITDA ratio of 1.5x by mid-2013. During the quarter, $241 million or 4.8 million shares of common stock were repurchased and retired. Finally, we are updating certain aspects of the full year 2012 guidance as outlined on Slide 17. Net product sales are now expected to be $8.8 billion to $9.0 billion, an increase of $200 million from the prior range. Non-GAAP R&D expenses are expected to be in the range of $1.45 billion to $1.525 billion due to accelerated clinical studies in liver disease and oncology. All other aspects of our guidance remain unchanged. I'll now turn the call over to John Martin.

Thank you, Robin. We are very pleased with the continued strong demand for our products and the maturation of the pipeline. During the quarter, 2 successful FDA Advisory Committee meetings were held. One on the use of Truvada for PrEP and one on Quad for the treatment of HIV infection. Just last week, FDA approved Truvada for prevention of sexual transmission of HIV in high risk individuals. This is the first approval of an anti-retro viral for pre-exposure prophylaxis, providing another option in the fight against the HIV pandemic. The U.S. government's commitment to HIV prevention, diagnosis and treatment was also demonstrated by the approval of the first over-the-counter HIV test. The next U.S. regulatory milestone is the upcoming August 27 PDUFA date for Quad and our scientific and commercial teams are well prepared for the launch. European Union approval of Quad is anticipated in the first half of 2013. This past quarter, NDAs and MAAs works for Cobicistat and Elvitegravir were submitted. And this week, the International AIDS Conference is taking place in Washington, D.C. Gilead presented 3 noteworthy abstracts at this conference. One was on 96-week data from Study 145 indicating that Elvitegravir was non-inferior to Raltegravir when added to optimized background therapy in treatment experienced patient. In addition, 48-week data from the Phase III study comparing Cobicistat to Ritonavir each in combination with Atazanavir and Truvada showed that Cobicistat was not inferior to Ritonavir and lastly, the SPIRIT study showed that switching completely suppressed patients from a PI containing regimen to the once daily Complera single tablet regimen was not inferior to patients maintaining their PI regimens. Finally, rapid progress has been made with the HCV Research Program and we now have a comprehensive Phase III development program. Norbert will cover that in his comments.

Thank you, John. Before I get to HCV, I would like to state that multiple programs have advanced across all therapeutic areas. In HIV, 2 Phase II studies were progressed which evaluate the safety and efficacy of 2 single tablet regimens, containing a novel product of Tenofovir GS-7340. One trial, which is now fully enrolled compares a single tablet regimen of Elvitegravir, Cobicistat, Emtricitabine, GS-7340 to Quad and the other study, which should complete enrollment by the end of August compares a single tablet regimen of Darunavir, Cobicistat, Emtricitabine, GS-7340 to the single components of the Darunavir, Cobicistat and Truvada. In oncology, 2 Phase III clinical trials of GS-1101 for the treatment of chronic lymphocytic leukemia has been initiated. These studies in a placebo-controlled fashion compare GS-1101 added to either Rituximab or to Rituximab with Bendamustine in relapsed refractory CLL patient. My remaining comments will focus on hepatitis C, where progress has been rapid. At the EASL conference earlier this year, data were disclosed from 2 Phase II studies of a 12-week, 12-week course of GS-7977 and Ribavirin in genotype 1 infected treatment-naïve patients. In the ELECTRON study, the SVR4 rate in genotype 1 naïve patients was 88% or 22 out of 25 patients and in the QUANTUM study, the SVR4 rate in genotype 1 treatment-naïve patients was 53% or 10 out of 19 patients. I would now like to provide an update on new data from 2 Phase II studies evaluating 24 weeks of treatment with GS-7977 and Ribavirin, in treatment-naïve genotype 1 infected patients. In the QUANTUM study, 19 genotype 1 patients were randomized to receive 24 weeks of GS-977 and Ribavirin. Of those 19 patients, 10 were 53% achieved in SVR4. The second trial is conducted by the NIAID, in a cohort of genotype 1 infected predominantly…

[Operator Instructions] And our first question comes from the line of Geoffrey Porges, Bernstein. Geoffrey C. Porges - Sanford C. Bernstein & Co., LLC., Research Division: Norbert, I have to ask the obvious question, which is are the differences between the patients in the 2 Phase II trial arms that you reported, the 7977 Ribavirin, genotype 1 that would explain the gap between 53% and 100%. Anything that you see in either how studies were conducted, the compliance or anything else that would explain that or should we just split the difference?

No, Geoff, unfortunately I don't have a clear cut answer for you. Of course, we looked at all the obvious base line imbalances. There wasn't really anything there. It seems like at this point, really the QUANTUM study seems to be the outlier. We're seeing better results in ELECTRON and also better results in the NIAID study and I said in my script, that's a population where you actually would have expected lower response rates and yet we get 100%. So I do not understand it, the thing of small numbers and where the ultimate truth lies. I don't know, but either way, I think these numbers are actually a huge improvement with a very simplified, shortened all-oral regimen over where we are today. And that's also reflected by the way when we talk to regulatory authorities that they allowed us to do single arm studies with no control arm, that's a reflection that even regulatory authorities think that the standard of care control arms are not really worth including any more in Phase III clinical studies. So sorry, Geoff, I don't have a really good answer to your question. It's have something we have looked into, but we don't know.

Our next question comes from Geoff Meacham, JPMorgan. Geoffrey C. Meacham - JP Morgan Chase & Co, Research Division: Another one on the hep C franchise. So Norbert when I look at Slide 36 with the 7977, 7585 Phase III design, with the second confirmatory study that says TBD. So can we assume that you won't be able to use any 7977 Riba studies alone or do you plan to do that over any duration, 12 or 24-week as part of the regulatory strategy or is that sort of off the table at this point?

So, Geoff, our current working assumption is that the data that Bristol-Myers generated their study will be replicated in our studies. We have made this argument before that from a virological point of view, the Daclatasvir and GS-5885 are very similar and there's no reason to expect a different result. So if that's the case, then obviously the second study might be a single arm study that looks at the fixed dose combination either with or without Ribavirin for either 12 or 24 weeks depending on how those data come out. So that's what our current thinking is. We do not currently think that GS-7977 Ribavirin will be included because GS-7977 with 5885 has -- its virologically more potent and moreover has the advantage of being much safer in 1 pill once daily, so it's the ultimate frontier In HCV therapy. That's what we want to pursue. Geoffrey C. Meacham - JP Morgan Chase & Co, Research Division: And just as a follow-up to that. When you run the interim analysis or you get that data, could you start the second confirmatory study to save some time. What sort of shortcuts do you feel like you can make or what sort of amendments can you make to accelerate an all oral filing, do you think can still, maybe to the end of '13, not mid-'14?

So, Jeff, the interim analysis, as I said is on the first -- on the 12-week treatment arms, so if the 12-week treatment arms show high response rates, then, of course, we would immediately be in a position to initiate the second confirmatory study as a very simple, perhaps single arm 12-week duration, fixed dose combination plus or minus Ribavirin and keep in mind, as I said in my script, we also have data from 2 cohorts in the ELECTRON study and be able [ph] to at that time that look at 12-week of treatment with GS-7977 and 5885 and Ribavirin in both null responders and naïve patients. I think with the totality of those data, it should be fairly straightforward to design the second Phase III study.

Our next question comes from Rachel McMinn, Bank of America Merrill Lynch.

I think you've answered a lot of the hep C questions, but I just want to be really clear. So regulatory agencies just don't want to see any control versus an interferon arm, so we should expect to your competitors as well to have similar designs? And then also to clarify the population here, this is just treatment-naïve, there is no category around interferon intolerance. And then just a totally separate question on 7340 can you just give us a sense of what antiviral efficacy or excuse me -- what kind of differential you need to see -- are you looking for primarily around perimetral [ph] density. What would be a good outcome there?

So the first question you asked was about NEUTRINO, of course, I can't speak for our competitors, but I would assume that the same comp set that regulators communicated to us is also applicable to them. So whatever in the future Phase III studies are going to be done might be much simpler studies that do not include a current standard of care control arm, that's the clear message we got and that's why they allowed us to do that single arm Phase III study that by the way, is powered against historical controls. The second question you asked was about 7340. So Rachel, as you may know, the monotherapy data were presented at the last CROI meeting and we showed that 25 milligrams of 7340 actually has higher potency in terms of HIV RNA reductions compared to Tenofovir DF 300 milligrams. Now in a triple combination regimen, higher potency in my mind is almost impossible to show, because all these triple combination therapy are very, very potent. They suppress the virus. So what we're hoping for is to show some benefit in terms of safety and you pointed it out already, one thing that we should be able to show is less of a BMD effect and the other thing we're exploring in terms [ph] of 7340 in advanced -- in renal impairment and maybe we can show some benefit there, but it's not quite clear. But you know, the big thing I think about 7340 will be that physicians will just feel much more comfortable. They do not have to worry about whatever potential, theoretical or hypothetical or real side-effects there are. Don't worry about bone, don't worry about renal, this is the same or higher potency at 1/10 the dose. So I think it would be a great addition to what we have today.

Our next question comes from Mr. Brian Abraham, Wells Fargo Security.

Related to the 7977, 5885 study, just wondering if you have any sense from the FDA or EMA in terms of what they're looking for, for what would be an approvable SVR12 bar just given the lack of standard of care. And also any sense of the bar for your futility there? And then related on NEUTRINO, do you have any sense from payers about what kind of flexibility docs may have to prescribe 7977 if it's approved in genotype 1, but the flexibility they would have to use it in interferon-free regimens before you get the approval for the 5885 combo?

So Brian, I'm happy to talk about the powering. By the way, the powering for all our Phase III studies is approximately a similar line of thinking and so first of all, keep in mind that we will enroll in our study at least 20% cirrhotic patients. There has been some criticism by the medical community of previous development programs that have studied relatively healthy patients and then the ultimate experience that physicians made was worse than what was observed in Phase III studies because the more advanced cirrhotic patients were not studied. We're not going to do that. So we have 20% target. And by the way, in our FUSION study, we have actually have 30% cirrhotic patients. So that's number one. Number two, we calculated historically what would be an expected response rates in genotype 1 in a population that is 20% cirrhotics and the answer to that is if you look across a lot Tenofovir Resapravir [ph] studies, the answer is 65% is the expected response rate. Now regulators gave us a bonus. They say you don't have to even meet 65%, if you meet 60% that will be perfectly fine because you get the bonus for shorter treatment duration and better safety. So what we have to beat in other words is 60% and let me point out, that's a very low hurdle. If you look at the ATOMIC study, we have 93% -- 92% response rates and so that's how the interim analysis is going to be conducted. If we see in the futility analysis, less than 60% response rate, that will be considered futile. But if we 60% or higher, that will be acceptable. And as I said, my expectation, based on previous Phase II studies and also Bristol-Myers' experienced that we will see much higher numbers than 60%. I hope that answers your question.

I guess, the second question was about the flexibility about using combinations. And so as you know, once 7977 is approved. Doctors have the flexibility to prescribe it in the way that they see fit. And I guess, we could imagine that there may be other antivirals approved that could be useful in combination with 7977 during an interim period before we have a fixed dose combination out. My own feeling is that certainly, treatment-naïve patients, and there's a bunch of things we don't understand here including what the cost of the other therapy would be. That in the treatment-naïve patient there would probably be more control over going to a peg-interferon 12-week regimen, but in patients who can't take interferon for various reasons or have failed interferon, I can imagine a scenario where payers will allow those patients to be dosed by combination therapies, but we really have to see how the data play out in those areas too. If there's anything to support those combinations or not.

Our next question comes from Mr. Matt Roden from UBS.

So your phase III with the combination 7977 and 5885 is contingent on this Phase I bioequivalent study. Norbert, can you give us a sense for whether or not there's any risk or if this is the slam dunk sort of study where you would expect to get adequate exposures a priori. And then secondly, on duration of therapy, it looks like at least in QUANTUM you didn't see any change in the efficacy, whether it's dosed at 12 or 24 weeks. Do you think that there's any read-through to other regimens such as nucleokinesis [ph] [NS5A] combination and whether or not you expect to see any meaningful difference between 12 and 24 weeks with the nucleokinesis [ph] [NS5A].

So the first question had to do -- by the way I want to state it's not a bioequivalent study, but a bioavailability study. Bioequivalent is very strictly defined by being equivalent to a test regimen. We don't have that. We just want to make sure that roughly the levels of 7977 and 5885 are in the ballpark of what we have seen in the previous studies and while it's always difficult to say what the risk is, I think both compounds 7977 and 5885 are pharmaceutically well-behaved compounds. So we have the first formulation, but needless to say if this first formulation doesn't work out, of course, we have backup formulations that we will then put into development, that would delay the program by 2 months or so. The other question you asked was -- duration of therapy. So again I don't have an answer. I honestly -- I do believe that there should be a difference in duration of therapy. Why we are not seeing that in QUANTUM I do not know. And needless to say, we have looked at everything possible that we could look at. The compliance was not quite what it should have been in QUANTUM, although that's a difficult thing to measure. And the reason why I'm saying there should be a difference with regards to duration of therapy is simply because we have never observed breakthroughs. All our patients that are nonresponders are rebounders. So after you eliminate, after you take 7977 away -- your 7977 and Ribavirin, then the virus comes back. During treatment, we have actually never seen breakthroughs. That's why I think 7977, by itself or with Ribavirin is a very potent, very effective agent. And if you can't cure somebody with 12 weeks by going longer, you should actually get the better outcomes, but unfortunately, I only have the data that we have today and I can't really speculate anymore.

[Operator Instructions] Our next question comes from the line of Marshall Urist, Morgan Stanley.

So another, predictably, another hep C question. So just to be clear on the interim analysis for the 5885 combination study for 7977, is the -- Norbert, is the hurdle rate there 60%, number one. Number two, how will the -- will this passing of the interim analysis be something that will be explicitly communicated to everyone when it happens. And then maybe finally on just if you could talk a little bit more about your FDA discussions and was it principally the Bristol combination data that got the FDA comfortable going forward with this design or are there other data that they saw as part of your discussions that kind of tied it all together?

Yes, Marshall, the interim analysis, let me say it again, so the -- we want to beat a response rate of 60%. So if 65% is the expected, the response rate with the current standard of care, Tenofovir Resapravir [ph], we get a 5% bonus, so we want to beat 60%. That means the lower bound of our 95% confidence in total [ph] has to be 60% or higher. And if we see a point estimate that's less than 60%, so let's say we see 59%, that would mean we only have a 5% chance of achieving our goal. I hope I was clear in what I said. So 59% or lower would be futile to show superiority, 60% or higher would allow us to go forward. Then you had another question about the FDA. No, so the FDA they were comfortable with the fact that we initiated the Phase III study, but to an interim analysis after 200 patients. And so really the totality of the data both the Bristol-Myers data, our own virology data and the previous -- current experience with 7977 even by itself or in combination with Ribavirin that has made the FDA and 3 European regulatory agencies comfortable with this approach.

And just to remind everybody please keep your questions to one question only, without 3 parts or we'll just answer the first part of your question and then move on so everybody has an opportunity to speak.

Our next question comes from Michael Yee, RBC Capital Markets.

On the QUANTUM 24-week data that you reported, which are the 53, do you know the IL 20B breakdown? I recall in the 12-week, there was a lot of differences, they were harder to treat patients. Maybe that was the reason?

So generally, in the QUANTUM study, the CC genotype was not as high as in the ELECTRON study. If I remember it correctly, it was 25% or 30% and there was no difference between the 12-week and the 24-week. That is still true. But it was a little bit higher in the ELECTRON, but that doesn't account for the whole explanation, I would think. In my mind, if you ask me to speculate, it has to do with study conduct and remember, the QUANTUM study was a study where patients actually were blinded on treatment assignment, they were taking 938 or placebo. They were informed that there was a liver toxicity issue with 938 that maybe that could have influenced somewhat the study conduct at compliance. So that if you asked me to speculate, that's probably the best I can come up with although we can't substantiate that currently from our data.

Our next question comes from Mark Schoenebaum of ISI Group.

You may have answered this. I'm sorry, I was sort of on and off. I really apologize. So I may end up wasting my question, but I'm just confused, when do you think you will be able to file for the fixed dose combination file for approval? Is the single trial enough or are you going to need both trials? And I guess, what kind of time line are you expecting for that first fixed dose combination trial, I'm obviously talking about the NS5A plus 7977?

Mark, since this is a new chemical entity. It contains a new product, we need 2 Phase III studies, 2 confirmatory -- 1 confirmatory Phase III study and one other Phase III. So the current time lines if everything works out and everything goes without a glitch, roughly the filing for the fixed dose combination would follow one year behind the 7977 filing, which would then be middle of 2014. That's our best guess currently.

Our next question comes from Yaron Werber, Citi.

Just a quick, maybe Robin, in case you're still in the room, because no one cares about financials anymore. The quick question on, so you beat by roughly about $100 million, but you're raising by 200, what's the -- and I'm going to try to make it one sentence, so it's one question. What's the difference? How much of inventory stocking did you see and were the strengths unusually strong in Europe relative to your expectation?

Overall, we saw a strong non-retail demand in the whole second half. Remember, if you recall, Kevin talked last quarter about the fact that we expected some reduction in Q2. While we saw some of that, it wasn't as much as what we had anticipated. So we do potentially see some inventory run off that's [ph] happening in the second half. I mean, all that being said, we're balancing that with the fact that we clearly know that the states are trying to reduce overall wait list.

Yes, I'll just add to that, Yaron. Q2 was pretty strong. It was not as high as Q1, but it was as good as Q4 of last year. So that was pretty strong and certainly far stronger than Q2 of 2011. So it was -- it quite surprised us. Puerto Rico, Illinois were particularly big in their purchases. I think at the moment, Yaron, there is quite a lot confidence in HIV right now. The budgets went out early. You saw just about week ago that the remaining $75 million was allocated. So pretty much all of the budget bar $6 million has gone out now. So the states know what they're working with, vis-à-vis the Federal funds, it's quite remarkable to see the wait list now down 1,800, that's over 80% from its peak at 9,200. So I think there's a sense of greater confidence, desire to get the wait list down, the mantra of test, treat, retain is very strong right now. You saw the comments coming out of the World AIDS meeting. So I think people recognize that treatment is the right way to go for the individual and it's right way to go for prevention and -- there is a sense of momentum there that I don't think was there this time last year.

Our next question comes from Robin Karnauskas, Deutsche Bank.

Just a quick question for you, I guess, on HIV since most of the hep C has been covered. Since all the macro good news coming out the guidelines, do you think the new stock market will be growing. I think before you'd said around 50,000 a year. Would you expect that to go up?

I think it's going to continue to be on its run rate, Robyn. I think the remarkable thing about HIV is it's been so consistent and that's through guideline changes and it's through more data, single tablet regimens and it just got this remarkable robustness and consistency about it. So there might be some small tick up, but I think the reassuring thing is it's just been remarkably consistent and I think that's the way we see it going forward.

Our next question comes from Tom Russo, Baird. Thomas J. Russo - Robert W. Baird & Co. Incorporated, Research Division: The 12-week arms in the BRISTOL 7977 study started enrolling really earlier this year and I think its status recently updated to no longer recruiting and that would suggest a decent chunk of patients might be out, some number of weeks post-treatment in 12-week arms, so I was just wondering if you're comfortable saying are you making any inferences from that on relapse or lack thereof that make it worthwhile to include 12-week arms in the Phase III study that you're announcing that. In other words, is it informed on that at all yet?

No, Tom, we're not using any of those ongoing studies. That's a study, as you know, conducted Bristol-Myers and if I am not mistaken, they said that they could announce some data by AASLD. So the -- but that's entirely up to BMS. It's not our study.

Our next question comes from Ian Somaiya, Piper Jaffray.

I had a question on the HIV front, specifically 7340. Given how close you were in the 102 trials to achieving superiority, I'm just curious why you're not pursuing a trial of potentially adequate ties as part of the Phase III program to get a superiority claim. I guess, if we learned anything, you guys are close in 102 and maybe another 100 patients and I guess, applying the Glaxo strategy could get you over that threshold?

Ian, we would of course try and be superior, but as you know, most of these noninferiority, superiority trials, so the nonresponders in HIV studies, a large part of those are either discontinuation due to adverse events or lost [ph] to follow-up and there aren't that many virological failures, the few that there are, you always wonder they're probably due to noncompliance more than due to inherent limitations of the potency of the regimen. So I just think they've lost [ph] to follow-up and it's conditional due to adverse event and noncompliance. Those are kind of generic things that haunt you in any study and I think by having 7340 in the regimen, I'm not sure whether you would -- whether that would address -- would not address those issue and I'm not sure if you would see superiority. But needless to say, of course, a good idea that's what we're, of course, attempting to do.

Our next question comes from Ravi Mehrotra, Credit Suisse. Ravi Mehrotra - Crédit Suisse AG, Research Division: Question for Norbert. When do you expect to get the results for your SVR12 from ARM 13 from the ELECTRON study 7977, 5885 and will you announce those as you get them?

Which study? What is it now Ravi?

ELECTRON, the 5885, 7977 ELECTRON arm.

We may have something -- Ravi I'm actually not sure. I have to look it up, honestly. I don't know where Ed Gain [ph] is in the recruitment of the patients. I can get back to you. I don't know.

Our next question comes from Jason Kolbert, Maxim Securities.

Just wanted to ask a quick question. With approval of the Quad hopefully coming up, how does that patient population differ versus the dynamics associated with Complera?

So I think we've always said that the opportunity that we have with Quad is to go after the majority of new starts, so we'll have data both -- well, we do have data from 102 and 103 versus Atripla and versus the protease inhibitors. So essentially we're able to position our Quad against the current, almost 9 out of 10 patients who are starting on those 2 third agents, either with Truvada or as part of Complera, part of Atripla. So that's the biggest slice of the naïve pie currently, Complera operates in. I think we've seen some very, very high response rates at 88% and 90% with the Quad, 2 very well controlled studies. We'll have to see, but I think we are hopeful that Quad would go into guidelines into the U.S. guidelines fairly quickly so I think that would give more opportunity for the Quad than we have today with Complera. Having said that, Complera is starting to do very well. So I think our expectation is that the first single tablet regimen becomes the Quad and then the second single tablet regimen becomes Complera.

Our next question comes from the line of Brian Skorney, Brean Murray. Brian Skorney - Brean Murray, Carret & Co., LLC, Research Division: I just want to kind of get some color on eventual pan-genotypic strategy. I know you've said in the past that 5885 won't be developed in GT3 due to the potency shift. When I look at Slide 36 and the Phase III fixed dose combo study, especially in the arms including Ribavirin. It just seems to me like you already have a pan-genotypic regimen with those 3 drugs right now and why wouldn't you think about just expanding the study to include all comers and potentially moving Ribavirin to a once daily dose. It would seem like the activity Riba has in GT3 would make up for any miss in 5885, and 5885 would make up for any weakness in Riba in the genotype 1 population.

Brian, it's a hypothetical possibility. You may remember the Bristol-Myers data that were presented at EASL where they saw 100% response rates in genotype 1, they saw lower response rates, in other genotypes. So clearly, 7977 Daclatasvir is not as good in non-genotype 1 patients. That's also, by the way, completely consistent with the virology data and our own virology data on 5885 would actually indicate that 5885 would really not be suitable for genotype 2. We see a lot of polymers there where GS-5885 activity [ph]. So -- I know what you're saying. You're saying, well you have 7977 and Ribavirin for genotype 2, but then you have a drug in there 5885, which doesn't really do much in genotype 2. So to make a long story short, we have other strategies that would come as kind of the third wave. And we're working on other nucleoside and we have some very interesting pan-genotypic compound in 2 other classes in preclinical development that would soon go into the clinic. That's what we're thinking about then the ultimate regimen will be 2 drug. It could be NS5A or PI -- or a nuc [ph] combined with 7977 that would truly be a good pan-genotypic regimen. That's our goal and plan.

Our next question comes from Thomas Wei, Jefferies. Thomas Wei - Jefferies & Company, Inc., Research Division: Just a question on the QUANTUM data. You had mentioned I think in response to a question that you had done some assessments of compliance in QUANTUM and that there are lower rates there may be relative to the other studies that have been done with 7977 and Ribavirin. Can you give us a sense of how poorly compliant those patients were in QUANTUM and what it looks like in the other studies as well?

Yes, so, Tom, the other thing you have to remember in ELECTRON and in the NIAID study, these were single-side studies where -- like Ed Gain [ph] and the ELECTRON study, he knows every patient by name, he is in very close contact with them. The same at NIAID, they are very close contact with the patients, where as in QUANTUM is more of a large center study. So we have looked at the measure of compliance in QUANTUM was simply a self-recorded, did you take your drug? And as you know, that's somewhat less than reliable and we have seen low report rate, it's just -- the question is how much of that do you believe? It would be nice to have a somewhat more -- like plasma levels are the ultimate things that don't lie, that you can't hide behind and we just don't have those data. But, Tom, we would move forward with the Phase III study and then the Phase III studies we would see what the 2 response rates are, so I'm not too worried about what is it with QUANTUM that we see lower rates than in the other studies. And I'm convinced in the Phase III study we would see higher than what we saw in QUANTUM. That's my own feeling.

Our next question comes from Jim Birchenough, BM Capital (sic) [BMO].

I wanted to just push a bit on the concept of confidence intervals because we don't get much data bounded by confidence intervals. So on QUANTUM and the NIAID study, do you have confidence intervals on those response rates. And I'm trying to understand in the interim when you've got 50 patients per arm, what's the difference between a 59% point estimate that's futile and a 61% when you think about the confidence intervals there?

Jim, we actually have done conference interval in QUANTUM and ELECTRON and one thing I can tell you is they are fairly large because the number is so small. So typically, if you have like 10 or 20 patients and you see a 50% response rate, the 95% confidence interval will go from 20 to 80, so plus 30, minus 30, that's how large the confidence interval is there [ph], so it's not very informative. So with regards to the interim analysis, I want to say it again. So our goal is to be superior to 60%. So in other words, that lower bound of the 95% confidence interval at the end of the study has to be 60% or higher. So if we see 59% in the interim analysis as the point estimate, that means we will have only a 5% chance of ever making that and that will then be a futile attempt, so 59% or lower we would say this is futile, to reach this point at 60% or higher we would think it's possible.

Can I push on that a bit? I guess, what I'm trying to understand is that if you're trying to hit a lower bound that's above 60%, shouldn't your point estimate be way higher than that. And I guess, where I'm coming from is 59% suggest that less than 5% chance to hit your end point. What's your chance of hitting your end point if the point estimate is 61%, because presumably, there's still a pretty wide confidence interval that's going to well below 60%, that's what I'm trying to understand.

So Jim, in order to reach the lower bound of 60% or higher, the point estimate calculation is 66%. So if we have -- if at the end of the study in 800 patients, the point estimate is 66% then the lower bound of the 95% confidence interval is 60%, so we got from 60% to 72%, the 95% confidence interval. With regards to your second point, I completely agree, but keep in mind, this is such a low bar that I do not believe this will ever be an issue. I mean this shows you how long the bar is. Like in ATOMIC, we had the same power calculation, we want to be at least 67% point estimate, but we were at 92% and I am convinced in this 5885, 7977 study we will see a similar result.

Our next question comes from Phil Nadeau of Cowen and Company.

You've done a good job identifying a Phase III program for genotype 1 treatment-naïve patients. I think the last thing we're waiting for is a Phase II program for genotype 1 treatment experience patients. Can you give us some sense of where your primary thinking is there and when we could get a more definitive Phase III program structure for that population?

Phil, so I didn't get into this, but we decided with the first filing for GS-7977 by middle of next year, we would not include genotype 1 treatment experienced patients. And the reason is as you may remember, we presented ELECTRON data, I think they were somewhere before EASL, where we saw about a 10% response rate in the null responders, that's just too low. We will, however, a clue -- of course include treatment experienced patients in our GS-5885, 7977 program. I didn't talk about that today, but that would of course be another Phase III study that we would do because this compound -- the combination of the 2 should actually be a great compound to look at -- experienced patients. And as I said in the ELECTRON, there is 1 cohort of genotype 1 null responders. So that will give us the proof or the confidence that this is worth doing and then we would include that in the Phase III program. There is, finally, a lot of other studies that I have not talked about that will be included -- that includes special population transplant, HIV co-infection, it goes without seeing that this will be a comprehensive program. I've just touched on the main Phase III studies today to give you some sense as to what the main body of information is.

Would you be filing for treatment experienced patients in mid-2014?

Yes, it could be under same filing.

Our next question comes from Howard Liang from Leerink Swann.

When do we see the data of 7977 in combination with protease inhibitor -- the Tibotec [ph] compound?

The study is now enrolling and honestly, I am not at liberty to talk about the time lines, that's a J&J Tibotec [ph] managed study. We have regular meetings with them and I know their study is enrolling, but I do not currently know what the time lines are for that study.

Our next question comes from Tommy Butler, Barclays Capital.

Can you think about what you would like to present or what you're anticipating presenting at AASLD in November?

I, Ken and Paul [ph], we have talked internally about that -- large number of potential presentation and I'm really not prepared to mention them right now. We haven't finalized them yet, but you will get updates on many of these studies, QUANTUM, ELECTRON, we'll present whatever is available to -- for them to be presented.

Our next question comes from Alan Carr, Needham & Company. Alan Carr - Needham & Company, LLC, Research Division: Can you speak to what will be going on like behind 5885 with some of your other antivirals like 9669 in combination with 7977?

So we are currently doing a number of again small cohort within ELECTRON, looking at 9669 with 7977. We're looking at our protease inhibitor with [ph] 7977, but as I said, the real goal is something we haven't really disclosed or talked about is pan-genotypic classes and one of them, we have -- other nucleosides that we're working on that could be combined with 7977. We have NS5A inhibitors that could -- are potentially pan-genotypic and we have protease inhibitors. And that's just something now that we have -- we're clear on what the first NDA, MDA [ph] filing looks like. We have now decided on the second filing. Now we can seriously start thinking about or formalizing a plan to move ahead with our third phase and the third phase will clearly be a one pill, once daily, maybe 12-week course for all genotypes. That's our goal still, and we're very close.

Our last question comes from Joel Sendek, Stifel Nicolaus. Joel D. Sendek - Stifel, Nicolaus & Co., Inc., Research Division: Another HIV question actually. Just to follow up on the earlier comment that you made about Dolutegravir. I'm wondering, given what you said about noncompliance and those sorts of things, do you view that drug as a threat, obviously, it showed superiority but is that meaningful at all when you look at the market down the road once QUAD's on that market?

Just a quick comment on the superiority. No, actually, if you compare the data -- response rates on the Dolutegravir arm, they were very similar to the QUAD. We actually had 90%, they had 88% and interestingly enough, the superiority was entirely driven by fairly high discontinuation rates due to averse events on the Atripla arm. Those discontinuation rates are twice as high as we have seen in our own study and they're also twice as high as other competitors like Merck has seen in their study. So that is just a common -- it's a peculiar result, why in their study, the discontinuation rates on rates on Atripla were so high. They were 10%, in our studies they are 5%. I'll let answer Kevin answer the other part of the question.

Joel, I think we take all forms of competition very seriously. One thing to point to out is that, with an expectation that the agent would be co-formulated with EPZICOM. EPZICOM only gets right now 4% of naïve patients. So it really is hardly used in the naïve treatment setting. And even though it was elevated in the recent IX guidelines, it's not associated with all the third agents as is Truvada. So we do think that the preferred regimen for an integrate inhibitor is the QUAD. The results are very good. The results are published. I'll go back to the point I made about -- hopefully early entry into treatment guidelines and we don't exactly know the timelines of the competition, but we probably have about a year head start on the market, and with an expanded field [ph]force, which is trained and ready for the PDUFA date. We're going to try to make most of that. So we do think we have an excellent opportunity coming out first with the single tablet regimen of an integrate inhibitor.

Mr. O'Brien, at this point, we have run out of time for additional questions.

Thank you, Chanel, and thanks for you all joining us today. We appreciate your continued interest in Gilead and the team and I are here to follow up with any follow-on questions that we can help with. Thanks.

Ladies and gentlemen, that concludes the presentation. Thank you for your participation you may now disconnect. Have a good day.