Gilead Sciences, Inc. (GILD) Q3 2012 Earnings Report, Transcript and Summary

Patrick O'Brien Robin L. Washington - Chief Financial Officer, Principal Accounting Officer and Senior Vice President Kevin B. Young - Executive Vice President of Commercial Operations John C. Martin - Chairman of the Board and Chief Executive Officer Norbert W. Bischofberger - Chief Scientific Officer and Executive Vice President of Research & Development

Geoffrey C. Meacham - JP Morgan Chase & Co, Research Division Matthew Roden - UBS Investment Bank, Research Division Michael J. Yee - RBC Capital Markets, LLC, Research Division Rachel L. McMinn - BofA Merrill Lynch, Research Division Mark J. Schoenebaum - ISI Group Inc., Research Division Brian Corey Abrahams - Wells Fargo Securities, LLC, Research Division Geoffrey C. Porges - Sanford C. Bernstein & Co., LLC., Research Division Yaron Werber - Citigroup Inc, Research Division Robyn Karnauskas - Deutsche Bank AG, Research Division Marshall Urist - Morgan Stanley, Research Division M. Ian Somaiya - Piper Jaffray Companies, Research Division Brian Skorney - Brean Murray, Carret & Co., LLC, Research Division Thomas Wei - Jefferies & Company, Inc., Research Division Charles Anthony Butler - Barclays Capital, Research Division Alan Carr - Needham & Company, LLC, Research Division Joel D. Sendek - Stifel, Nicolaus & Co., Inc., Research Division

Ladies and gentlemen, thank you for standing by, and welcome to the Gilead Sciences Third Quarter 2012 Earnings Conference Call. My name is Ann, 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, Vice President of Investor Relations. Please go ahead.

Thank you, Ann. Good afternoon, everyone. We issued a press release this afternoon providing earnings results for the third quarter, which are available on our website. You can also find a detailed earnings slide deck that will support some of the topics we will cover on today's call. For our prepared remarks and Q&A, I'm joined by our Chairman and CEO, John Martin; our President and Chief Operating Officer, John Milligan; our Executive Vice President of Research and Development, Norbert Bischofberger; our Executive Vice President of Commercial Operations, Kevin Young; and our Chief Financial Officer, Robin Washington. Before we begin with our formal remarks, we want to remind you that we will be making forward-looking statements including plans and expectations with respect to our product candidates and financial projections, all of which involve certain assumptions, risks and uncertainties that are beyond our control and could cause our actual results to differ materially from these statements. A description of these risks can be found in our latest SEC disclosure documents and recent press releases. In addition, Gilead does not undertake any obligation to update any forward-looking statement 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 a press release, as well as our website. I will now turn the call over to Robin Washington.

Thank you, Patrick, and thank you, all, for joining us this afternoon. Gilead continues to deliver solid year-to-date and third quarter financial results. Product sales for the third quarter were $2.36 billion, an increase of 14% year-over-year. The U.S. contributed $1.4 billion to product sales, up 20% year-over-year, driven by strong retail demand and the expansion of our HIV franchise. Europe contributed $781 million to product sales, up 5% year-over-year and flat sequentially due to purchasing patterns related to the summer holiday. From an expense standpoint, non-GAAP R&D expenses increased to $384 million, up 42% year-over-year, primarily due to the continued advancement of our product pipeline and in particular, the progression and expansion of our Phase III studies in both liver disease and oncology. Additionally, non-GAAP SG&A expenses increased to $287 million, up 8% year-over-year due to the ongoing growth of our business and an increase in the U.S. pharmaceutical excise tax, partially offset by lower bad debt expense. Sequentially, SG&A expenses decreased 4% due to lower spending as a result of the European summer holidays, partially offset by investments in the launch of Stribild. Operating cash flows were $745 million. Recall the last quarter's operating cash flows of $1.29 billion included the receipt of approximately $460 million from collection of accounts receivable in Spain. Turning to financial guidance. Given the strong year-to-date commercial performance, we are increasing product sales guidance for full year 2012 to $9.1 billion to $9.2 billion, which represents a $200 million increase from the top end of our previous guidance range. The full year 2012 diluted EPS impact of acquisition-related restructuring and stock-based compensation expenses is now expected to be in the range of $0.62 to $0.65 per share, primarily due to changes in anticipated payment associated with the successful progression of our oncology and non-HCV liver pipeline. All other aspects of our non-GAAP guidance for 2012 remain unchanged. In closing, we are pleased with our financial performance during the quarter and are looking forward to closing out a strong fiscal year 2012. I'll now turn the call over to Kevin Young.

Thank you, Robin. I have been very pleased over the course of 2012 with the execution of the commercial organization. The third quarter was the strongest in Gilead's history, with solid performances across all therapeutic areas. The anti-viral franchise grew by 13% year-over-year and delivered over $2 billion in product sales, while cardiovascular products delivered quarterly sales of over $200 million for the first time. HIV sales in the U.S. were particularly robust and grew by 19% year-over-year. As we anticipated, following record non-retail purchases in quarter 1 and quarter 2 of 2012, quarter 3 non-retail sales were down sequentially. At this point, mid-way through the federal fiscal year, it would seem that budgets will allow for continued healthy ordering, assuming little or no rollover is allowed beyond March 2013. The latest update from [indiscernible] shows a total wait list of approximately 100 individuals across all states, down from a high of nearly 9,300 in September 2011. We were very pleased to be in a position to launch Stribild immediately after receiving U.S. FDA approval on August 27. Upon its launch, Stribild became our first priority detail. Uptick in the product in the first few weeks of launch have been extremely encouraging. Prescriptions are almost 2x that of Complera at a similar point in launch. New to brand patient starts on Stribild have surpassed Reyataz and are already approaching the level of Atripla and PREZISTA. Access to Stribild far exceeds that of where Complera was 2 months after launch. Most managed care organizations now list Stribild without restrictions. 47 state Medicaid plans and 37 AIDS drug assistance programs now have Stribild on formulary. Importantly, 4 of the big 5 ADAPs have full access to Stribild. We also continue to be pleased with the U.S. uptick of Complera, which sold $82…

Thank you, Kevin. During the third quarter, significant progress was made in our development programs. As Kevin mentioned, Stribild is the first integrase inhibitor containing single tablet regimen to receive U.S. FDA approval. FDA has accepted the MBAs for 2 of the components, cobicistat and elvitegravir, and confirmed PDUFA dates of 28th and 27th of April, 2013, respectively. For Complera, results from an 841-patient head-to-head study versus Atripla will be presented at the Eleventh International Congress on Drug Therapy in HIV Infection meeting in Glasgow next month. This is the first head-to-head study of Complera versus Atripla, as the previous Phase III studies supporting approval of Complera compared rilpivirine plus Truvada, to efavirenz plus Truvada. Also at the Glasgow meeting, 48-week data will be reported from a 482-patient study in which patients on a stable ritonavir boosted PI regimen were randomized to either switch to Complera or stay on the boosted PI regimen. Both these large studies were initiated with the intent to be included in the Complera label. The latter is particularly important because 40% of U.S. patients currently initiating Complera have switched from a PI containing regimen. GS-7340, a novel pro-drug of tenofovir, dosed at 1/10 the dose of Viread, was co-formulated with emtricitabine, cobicistat and elvitegravir into a single pill. This 7340 containing single tablet regimen is being directly compared to Stribild in a Phase II study. 24-week data from this study will become available in November. If the results are positive, we anticipate moving this 7340 containing single tablet regimen into Phase III in the first half of 2013. In HCV, there will be a number of presentations at the AASLD meeting in Boston next month, highlighting the safety, efficacy and resistance profile of GS-7977 in a variety of combinations in HCV infected individuals. Noteworthy, 7977…

[Operator Instructions] And our first question comes from the line of Geoff Meacham with JPMorgan. Geoffrey C. Meacham - JP Morgan Chase & Co, Research Division: Just got a couple for Kevin and then one for Norbert. For Kevin, I know it's early on Stribild, but are you guys seeing uptick at the expense of PIs? And how much of a driver do you think the switching studies will be with respect to share? And then for Norbert, you guys have talked about the 7977-5885 combo study in the last call and I wanted just to know if anything has changed and how you're thinking about the second pivotal study? What types of patients, what types of design do you think you will employ? Things like that.

Geoff, this is Kevin. Thanks for the question. We are pleased with the progress so far on Stribild. It is early days, almost 2 months of launch. The sales you saw in Q3 were very much the inventory fill for the wholesale as that's very similar amount to the amount we put out a year ago when we launched Complera. I think we're particularly pleased with the access that we've got. It's very much quicker than we have for Complera, the ADAP programs particularly have taken up Stribild and we're almost -- we almost have 5 of 5. We're just waiting for a signature to have the fifth of the big 5 ADAP programs have coverage of Stribild. Twice as many prescriptions for Stribild as Complera, so it's running ahead of Complera, it's running ahead of the uptick of PREZISTA. It's more looking like kind of an uptick of ISENTRESS. So I think overall, we would consider, I think, the launch of Stribild more Atripla-like and less Complera-like. Obviously, you don't have the bolus of conversions that Atripla had, but our sense is there's probably a little bit more being used in the naive setting than in the switch setting. But our other sense, and again it's early, is that the uptick is probably coming at the expense of the protease inhibitors as per your comment earlier. So seems to be more being used in the naive setting than switch. The switch data will help and having in the label will help, but the team, our expanded team is very much positioning this as the preferred regimen for upfront use in new patients as an alternative to Atripla, as an alternative to the boosted protease inhibitors.

So Geoff, with regards to your question about the Phase III program, the big answer is no, nothing has changed. As John mentioned, the first Phase III study -- the first patient was dosed last week and just to underscore the repetitive enrollment we actually had to close the study for screening last week as well because we had enough patients in screening so that we can enroll the first 200. So we're going to take a look at this, the two 12-week arm sometime in the first quarter next year, probably March, and then we'll make a decision as to what to -- whether we're going to continue to randomize patients in this study. And then with regards to the other study, keep in mind, we have other data coming in. The 2 data sets will be presented at AASLD. One of them is the 12-week arms of the Bristol-Myers daclatasvir 7977 combination, another one is the 7977 5885 ribavirin in genotype 1 naive patients in the ELECTRON study and with them, we have a third arm in ELECTRON ongoing 7977 5885 ribavirin in genotype 1 null responders. And we will take all these data into consideration and based on that, design the second Phase III study, which should start pretty much at the latest around the same time as we're going to reopen this current study for re-randomization and re-enrollment. But just briefly, this could be a very simple, potentially 1-arm study with a 12-week arm, depending on the outcome. It could also include treatment, experienced patients, depending on what data looked like of the combination in the null responders, what it gain. So I hope that answers the question. It will all become clearer over the next 2 to 3 months what the nature and the design of that second Phase III is going to be. Geoffrey C. Meacham - JP Morgan Chase & Co, Research Division: That's really helpful. Just a quick follow-up, Norbert, is there anything...

Geoff, that's 3 questions already. We're going to have to move on to somebody else.

And our next question comes from the line of Matt Roden with UBS.

Question for Norbert on GS-7340. Is it reasonable for us to think about the potential to lower the incidence of renal adverse events with 7340 versus the standard of care? How many patients do you think would need to be treated and for how long to really start to tease out whether or not there's a difference here?

Yes. So Matt, that's exactly what we're hoping to be able to show. So essentially, 7340 is the same active ingredient as is in Viread, but it's dosed at 1/10 the dose. And they always said unless you believe in homeopathy you would -- it's reasonable to assume that this compound has better, say, tolerability and safety. And the 2 things in particular are BMD and creatinine clearance changes. Now I think -- I personally think that BMD changes are easier to show because simply we saw about a 2% net BMD loss with Viread across a whole number of studies and if this compound is better then it should be easier to show statistically. With regards to creatinine clearance, it really depends on the effect size of the variability. But we think, in a Phase III study, you should actually be able to show that. And this would become clearer by the way, as John Martin said in his script, we should actually get the Phase II -- the final 24-week Phase II data from the first study in a month or so and then it should become clearer what the powering will be for the Phase III study.

And our next question comes from the line of Michael Yee with RBC Capital Markets.

My question is for Norbert. Obviously, we've seen a lot of the data come out recently from late breakers 12 and 24 weeks with and without ribavirin from some other oral compounds. Maybe, since that's always been a question as to whether duration and with and without ribavirin matter, maybe you could hypothesize or speculate how you're thinking about that data and whether that changes your thinking since you've gotten some answers from other drugs recently.

Michael, I don't think I have a good answer for you. So as you may have seen also from our abstracts, one of them on the resistance of 7977 that's going to be at AASLD, we have never seen a breakthrough on 7977. And with one single exception, we have never seen resistance. And the one single exception was when 7977 was used by itself without anything else. So given that, then I would -- you would expect that treatment duration should matter. If you don't develop resistance, you don't have breakthrough, and you don't achieve 100% response with 12 weeks. If you think about it, going to 24-week should increase that response rate. Now that's theoretical. In the few instances where we have actually tested that, we didn't see that. So that unfortunately is the only thing I can tell you. I can't quite make sense out of the treatment duration yet. I think it should become clearer once we see the Bristol-Myers data and also the -- our Phase III results where we're directly comparing in a large sample size 12 versus 24 weeks.

And of ribavirin?

You know what, Michael, what overall it looks like is that 7977 is a very potent compound. But by itself, it's not quite enough. You have to add something else. Now if you add ribavirin it will bring the response rates up to a very respectable 80% to 90% to 100%, but not in all populations, particularly you may remember in genotype 1, it's not 100% and then in null responders, it's not 100%. But ribavirin, from a biological point of view, adds, actually, very minimal efficacy. So I think by adding something a little bit more potent to it, like for instance an NS5A, as has been shown by the Bristol-Myers data, you get 100% response rate. So that's how I look at 7977. It's a very potent compound, not quite enough by itself, so you have have to add at least one other -- at least semi-potent antiviral to it, like a PI or a NS5A or a NS5B.

And our next question comes from the line of Rachel McMinn with Bank of America Merrill Lynch.

I just wanted to better understand strategy in HCV. When we think about dual therapy versus triple therapy and, Norbert, I just -- I did listen to your comment on ribavirin, but I was curious how you're thinking about 9669. So you've talked about that as being a third agent, but you don't seem to be aggressively developing a safety database around that and then just separate from all that, curious on when you might consider starting an 8-week study with 7977 in either genotype 1 or genotype 2, 3.

Yes. So Rachel, with regards to 9669, so it's one of the most potent non-nuc NS5B inhibitors that has been described, very well behaved compound, once daily, seems to be safe. The reason why we're not as aggressively pursuing it is simple, because we're not quite sure what the place is in hep C therapy. We have 5885 with a PI. We have in the -- in preclinical in the works a pan-genotypic NS5A inhibitor, so we're just not sure what the role of 9669 is. So we are doing, as I'm sure you know, 2 cohorts in ELECTRON that look at 9669 with 7977 in ribavirin, both in genotype 1 naive and null responders, and I'm sure those data will look very good. But again, we're struggling with this issue where 9669 fit. I want to mention something. One role that I think triple combination therapy could have. If you shorten the treatment -- so let's say you get really good results with 2 agents and you shorten the treatment duration and at one point reach a duration where the response is suboptimal, then it's conceivable by adding a third agent then you will get, again get good responses. So that's a line of experiments that we're pursuing. And with regards to 7977 for shorter treatment duration than 12 weeks, there's actually 1 presentation at AASLD. It's going to be an Ed Gane's [ph] oral where he's looking at 7977 ribavirin for 8 weeks in genotype 2, 3. So that will give you some indication of we're thinking about, depending on the outcome of the other cohorts in ELECTRON, about looking at 5885-7977 ribavirin combination with a shorter than 12-week treatment duration. But there is nothing definitive yet.

And our next question comes from the line of Mark Schoenebaum with ISI Group.

Since it's one question, I'll respect that. I just want to clarify, are you guys going to give us the interim data from the Phase III in the 1Q. Because it just seems to me like that is of monumental importance. Obviously, everyone's going to go to AASLD and look at your 5885, 7977 riba 12-week data, but the real -- the really important data set's that 200-patient interim. Are you committed to giving that SVR4 data to Wall Street in the 1Q when you have it?

Mark, this is something we are still debating internally. On the one hand, we've realized the importance of the data for the financial community. On the other hand, we also -- you have to be cognizant of the fact that this is an ongoing randomized study and we can't really -- we would be influencing the conduct of the study potentially if we announced to the world the interim analysis. So something we're thinking about I think there might be other cohorts available at that time that would answer the question. So we would make those available publicly and not make available the actual interim analysis.

So, sorry, what are the other cohorts, Norbert? I just...

It's still in the planning stage, but we're thinking about initiating 1 or 2 other cohorts that would answer that same question that you have.

Cohorts in the Phase III or in the Phase II?

This would be Phase II.

And our next question comes from the line of Brian Abrahams with Wells Fargo Securities.

Continuing on the hep C front, what do you guys see as the bar for SVRs, for both naives and nulls, that would optimize the commercial opportunity just given some of the competitive data we've seen unfold? And how much weight should we place on sort of the predictive value of what we see come out of ELECTRON? And then just lastly, do you have any updates on the disclosure plan for the ELECTRON nulls arm?

Brian, it's Kevin. I wouldn't give you a pinpoint target of a kind of a threshold for SVR. Let's not forget, I think you should be SVR 12 because those are the endpoints of the study. But clearly, you've got to be very high these days. I think there are going to be a number of regimens from a number of companies that are going to have high SVRs and I hope -- that also includes Gilead in that comment. My personal feeling is that as we get towards launch, that other parameters, in addition to SVR, will come to play. The simplicity and the convenience of the regimens and the tolerability. And that's why approaches like 5885 plus 7977, plus ribavirin, or no ribavirin, becomes very attractive, 12 weeks, well-tolerated. It almost becomes sort of a Complera or Stribild approach, but over a 3-month period. So I think we are going to see high SVRs from a number of different approaches, different molecules, but I actually think the addition of the other parameters are going to be important in the real world, particularly beyond the specialist hepatology clinic when you start to go into the general gastro treater in the community, potentially other physicians beyond where the PIs are used today. And our expectation would be that the new regimens can go much further afield than the current prescribing of telaprevir and boceprevir.

And you asked 2 other sub question. You said what's the predictability of ELECTRON cohorts. I think it's entirely predictable. Just keep in mind you have to take it with a grain of salt because of the sample size. Some of the cohorts are 10, an n of 10, others are an n of 25, neither one is particularly large. So just keep in mind there is always variability. And secondly, you asked about a null responder disclosure. So when I talked to Ed Gane [ph], he let me know that probably what would be available at the time of the AASLD presentation will be a total of 3 patient SVR4. That's 3 out of 10. And we're a little bit debating internally, is that worth showing? Because we don't want to mislead anybody one way or another if the result is a certain way, but it's an n of 3. Now that is a really small sample size.

And our next question comes from the line of Geoffrey Porges with Bernstein. Geoffrey C. Porges - Sanford C. Bernstein & Co., LLC., Research Division: And Norbert, 3 patients has been enough to build whole companies in hep C. I just wanted to follow up with some of the questions about disclosure. Will you tell us about the results for 7340 in Q4? Or will we just learn from the decision to start the pivotal trials? It seems important. And just wondering, John, if you could give us an update on the prosecution of the patents for the combination products, where there's any news there?

Yes, we were actually planning on letting the financial community know about the results from the Phase II study because, Geoff, as you correctly pointed out, it has a direct consequence if the results look promising then we would, of course, move into Phase III and that's something that would become public one way or another at one point.

And Geoff, the second part of your question is we don't have any update on the prosecution of the combination patents, but if that procedure changed, of course, we will update you.

And our next question comes from the line of Yaron Werber with Citigroup.

So, a question, Norbert, just for you, I just want to follow up and to really understand a little bit. You're going to have data -- it sounds like imminently from the Bristol combo, which is going to give you a sense of whether you need riba or not and whether 12 weeks is enough. And it sounds like from what you're saying, it sounds like you guys are feeling pretty optimistic 12 weeks is enough and if I'm reading correctly, maybe even riba is not adding much. So I'm trying to understand. Then you are going to have data from the triple for 12 weeks from ELECTRON, but that's not a huge sample. Where I'm going with this is, is this enough for you and then have, by March, you're going to have preliminary, it sounds like, SVR4 at least from your current Phase III. Is that enough at that point to make a determination that the Phase III is only going to be 12 weeks? And if so, when do you think you can file for an all-oral for genotype 1?

Yes, Yaron, absolutely, that will be enough. So the argument that we always made and that we firmly believe in that daclatasvir and 5885 are very similar compound at least in genotype 1, they're very potent, similar potent against 1A, 1B, behaves similarly against polymorphs with maybe one exception, but in that polymorph, the Y93H [ph] it's still way above -- the IQ was still a factor of 6. So very potent. The only thing we're missing right now is a clinical proof of that and that will come from ELECTRON. So what ELECTRON would do for us if the data look promising, that would absolutely prove that -- it would establish the link between daclatasvir and 5885. That, together with, I said with the null responder data from ELECTRON and the Bristol-Myers data will give us enough confidence to design the second Phase III study. And then if we get the interim analysis, by the way the sample size is again not huge, it's 100, of the two 12-week arm, so 50 per arm. And again, it would -- I mean this all has to hang together and be confirmative of each other. But each study ideally confirms the other and we would have then enough pieces of evidence that we would feel very comfortable designing the second Phase III study. It's essentially ELECTRON, Bristol-Myers and our own Phase III study. Those are 3 studies, if you add up the number of patients, about 130 total and that's enough to give you confidence to go into Phase III.

And at that point, when do you think you can file for an all-oral, if you start the study, let's say, early Q2?

Well, I think, Yaron, you can do your own calculation. It's just assumed fairly rapid enrollment 12-week of treatment if that's what the treatment duration is, 12 weeks of SVR, 12 follow-up and add to that 3 to 4 months for filing. And then another 8 month for approval if it's priority review or 12 months if it's not.

And our next question comes from the line of Robyn Karnauskas with Deutsche Bank.

I guess, first, you showed that Atripla trends are sort of flat. I wonder if you can comment a little bit about what you're seeing in Atripla and with Atripla use? And second, just a follow-up to Mark's question on the arms that you might be adding in Phase II that might give us a hint about Phase III. Are you talking about maybe adding 24-week arms or maybe non-riba arms to the ELECTRON trial? Is that what you're sort of implying, sort of additional arms in perhaps ELECTRON that would give us a read by March as far as how the Phase III might be shaping up?

Robyn, I'll answer your second question quickly before Kevin answers the first. I don't want to get into that because it would give away -- it would preempt some of the data that you're going to see at AASLD. I think once you see the data at AASLD you will more -- it will be more evident what we're thinking about.

Yes, Robyn. Back to the question. Atripla is coming down. Let me try and give you just a couple of benchmarks. The Q2 Ipsos or Synovate data has Atripla at 34% of the naive market. Truvada is 31% and Complera is at 25%. So that gives you your kind of total of about 90% for Gilead that you've seen before. But clearly the mix is changing. So Atripla has progressively, over the last sort of 5 quarters, come off its high. And another marker in terms of change is that Complera gets 25% of its prescriptions from Atripla switch. So you can see that there is Atripla switching more than likely because of tolerability, the CNS side effects of efavirenz, coming across to Complera. The only other thing I'd add, Robyn, is the shift that you saw in the actual product sales in Q3 versus Q2. That was largely because of the non-retail, that was largely because of ADAP ordering. ADAP ordered quite large quantities in Q1 and Q2 of this year of Atripla, and they were relatively quiet on Atripla in Q3. So the actual product sale changes that you saw there were largely the result of the phasing of ADAP purchase.

And our next question comes from the line of Marshall Urist with Morgan Stanley.

So just one, Kevin. If you could give us a sense of how you think the dynamics between Complera and Stribild will play out, seeing as I think Complera is obviously, outperformed I think a lot of people's expectations. So what's the impact between those 2? And Norbert, maybe just a slightly different question, which is, has the data that we've seen from Abbott and others sort of changed your perspective or your thinking? Or at least what is your current thinking on the importance of getting rid of riba and dual therapy from the initial -- for the initial Phase III regimen? Or given the competitors, is it -- is there a little bit less pressure, you think, competitively, to drop riba at this point?

Complera, we're very pleased with. I think we've proven that if you put rilpivirine with a good backbone in a single tablet, it can do very well. And we particularly hear anecdotal feedback from physicians on the tolerability. How well patients like Complera. So it has done very nicely in real sort of clinical practice. However, Stribild is our first detail. We are positioning Stribild as first-rank therapy in the vast majority of patients. So you would expect that Stribild would quickly become the lead product in the naive setting, our label is naive. So you would expect it to take over. One just small piece of information that wasn't included in the script, but I think is nice to add for Stribild is that we have had a late breaker accepted at the Glasgow HIV meeting on our 96-week data. So that will be presented there and then hopefully, we'll get that into publication quite quickly for Stribild and then we hope going into 2013 that, that will be helpful eventually for moving Stribild into a preferred setting and guideline. So the profile of Stribild is likely to become stronger. So I think we've got a very good #1 product in Stribild. We're getting the data on switching for Complera and that could well become, I think, a very strong second switch product. The good news in all of this is that I think we've got 2 options in Stribild and Complera that are superior to Atripla. So I think it's a very good position that we're in.

So Marshall, with regards to the competition, that's an important question that I would like to comment on. So first of all, Abbott has a viable regimen. They obviously came to the conclusion that they needed all 5 agents, differing only in terms of treatment duration, 12 versus 24 weeks. But I will mention that what we do and the urgency that we feel is not really determined by what the competition does, but rather by where we ultimately want to go with the therapy, and our goal has always been to have 1 pill, once daily, without ribavirin, and I think that's entirely possible and that's what we're pursuing. So we, at one point, ribavirin will disappear. People won't use it anymore because of the obvious issues, it's BID, causes anemia and you don't have to use it, you wouldn't use it. And that, I think, we are going to achieve. The question is when and with what regimen.

And our next question comes from the line of Ian Somaiya with Piper Jaffray.

Just wanted to follow up with Norbert on sort of your longer-term vision on hep C. I think you mentioned the goal of moving to a once daily, 1 pill regimen. Just can you help us think about -- is the next step addition of a third pill, a third drug that you had co-formulated into a single pill? Or would it be a different combination, different -- let's say switching out your NS5A with something else?

Ian, again, I think the -- it's clear, if you just look at the Bristol-Myers data that were presented at EASL that you can have 2 drugs, in that case daclatasvir and 7977, dosed for 24 weeks gives you 100% response rates in genotype 1. It didn't give you that number in genotype 2, 3 for obvious reasons because daclatasvir and also our own NS5A inhibitor is really not optimized for genotype 2, 3. We have, very early in development, a pan-genotypic NS5A that at least biologically looks just absolutely wonderful. It's a compound that works against all genotypes, against most polymorphs and we think that will be the -- a simple answer to a 1 pill, once daily for all genotypes. The only thing I would like to say, Ian, you said about 3 drugs. It's -- the only role I can see for 3 drugs is if you wanted to further shorten the treatment duration, so it's possible, even though the concept hasn't been shown, that you could, for instance, get away with 10 weeks of treatment with 3 drugs, whereas if you have 2 drugs it needed 12 or 16 weeks. That's where I'm seeing a role of a third drug, but if you have 2 good drugs, you can actually -- that's enough. That will suffice to cure most hep C genotypes.

And what -- do you have a sense of timelines for the NS5A that you have in preclinical development, when that could move into clinic and when ultimately we could see that in the marketplace?

Ian, it's too early, really, for me to speculate on that. I would -- we're thinking about disclosing that sometime at the future conference. We haven't really decided when yet.

And our next question comes from the line of Brian Skorney with Brean Murray. Brian Skorney - Brean Murray, Carret & Co., LLC, Research Division: I'll jump a little off page and actually ask a question on GS-1101 and just the potential plans for a fileable single-arm study. I just think the response rate, as I recall, in Phase I in indolent NHL was about 57%, but I think that included a number of the higher than 150 mig BID doses. I just wonder, have you guys broken out the responses to 150 migs specifically for these patients and what should we think of as a potentially fileable hurdle for this study? Is it reasonable to think of relapsed refractory indolent NHL as equivalent to other relapsed refractory hematological malignancies? Or given that it's indolent should it -- would we anticipate a closer look at kind of a PFS?

Yes, Brian. We have not, actually -- I don't think we have broken it up by dose, but as you noted, the current 120 patient Phase II study that John talked about in iNHL is fully enrolled. We will look at response rates very soon and the only thing that what we don't know yet is -- and what is equally important is the duration of response and the reason is we just haven't followed enough patients for a long period of time. That will sometime happen next spring, but I also would like to caution you, we think it's unlikely that we will get approval or we will even file under single-arm study. It's more likely that we will get approval based on the Phase III studies that are ongoing. One last comment about response rates and what's required, you may want to look at carfilzomib that got approved on a single-arm study. It was a different disease in myeloma, but the response rate overall was 22% and the duration of response was 8 months. And I think it's probably an exception, but if we have something north of that, then we would definitely approach the FDA and see what they think. Brian Skorney - Brean Murray, Carret & Co., LLC, Research Division: And I guess, I mean -- so that kind of goes back to my question, I mean, do you -- and I was thinking about carfilzomib specifically, but do you think that the relapsed refractory indolent NHL population sort of meets the same hurdles as the relapsed refractory myeloma population, in terms of not really having exhausting all resources out there, especially with maybe increased use of bendamustine and rituximab now than previous years?

Yes, Brian, they do with one caveat. So they do in the sense that they are relapsed refractory to CD -- anti-CD 20s and cytotoxics, so bendamustine, rituximab. One potential argument that you could use to say they are not unmet need is because these people have not been treated with radioimmunotherapy. And of course the counter argument is nobody uses that. That was something that was brought up by the Advisory Committee meeting by Anderson. Again, this is something we don't know what the FDA would think. I mean, would they say these people have not exhausted all their options yet because radioimmunotherapy hasn't been used, or would they accept the previous relapsed response to cytotoxics and anti-CD 20s. That's kind of the question.

And our next question comes from the line of Thomas Wei with Jefferies. Thomas Wei - Jefferies & Company, Inc., Research Division: I wanted to ask -- have 3 questions. Just an update on how you're thinking about the genotype 1 cirrhotic population? How competitive do you think you're going to be there relative to some of the other companies that seem to be expediting development in that population? And how much more work do you need to define the right duration and regimen for cirrhotics?

Thanks, Thomas, for your question. Thomas, we're actually pursuing that in all our Phase III studies. So all our studies mandate at least 20% cirrhotics and actually in 1 study, the POSITRON study that you will hear about in this year, the results, that study enrolled 30% cirrhotics. So we are actively addressing that question. And we always add as a goal to have a representative sample of patients in our Phase III study that represents the patients that are actually going to be treated and a lot of those are cirrhotics. So those are included in all our studies in our Phase IV, Phase III studies that are currently ending or coming to an end. And they will also be included in the 7977-5885 fixed dose combination.

And our next question comes from line of Tony Butler with Barclays Capital.

I'll be brief, Norbert. Just again on cirrhotics. You did make reference to the notion that the only way you would think or you would hypothesize that a 3-drug regimen would be necessary, would be due to duration. Does that imply that you think that a fixed-dose 2-drug combination would be sufficient for cirrhotics, are you still leaving that door open? And then very quickly to Kevin. I'm respectful of the Stribild label for another naive setting, but why would it not erode or dampen the Complera demand, especially when Stribild is in first line on the -- at least, on those Atripla switches?

Tony, you are right. In terms of special populations, we may well realize that certain populations you can't -- 2 drugs are not sufficient, you may need 3 drugs. But again, I want to point back to our observation that we have, to this date, never seen a biological breakthrough on 7977, not one. All our patients that were nonresponders are rebounders. Again, that makes me -- convinces me that it's a function of duration of therapy rather than the number of drugs that are needed.

Yes, it's a very good point you make around sort of dampening the uptick of Complera, Eviplera. I think that's certainly the case in naive setting. It might have an effect in the switch setting. Of course, if people are on Atripla, then they've already been taking an NNRTI. And if it's a tolerability issue, it's a simple switch from 1 NNRTI to another simply based on the CNS. One area that we haven't talked about and we'll be waiting for data and of course again it's a label situation is that we have done the switching study from Truvada plus ISENTRESS, so raltegravir and there are over 50,000 patients on that regimen today. So I think that's more likely a potential switch in the situation that the doctor takes that decision individually to do that. So I think for Stribild, it's more likely a PI switch or a Truvada ISENTRESS switch.

And our next question comes from the line of Alan Carr with Needham & Company. Alan Carr - Needham & Company, LLC, Research Division: Just a quick one. Could you comment on the share buyback activity and strategy as well as managing the debt as well?

Sure, Alan. We did repurchase shares, it's actually outlined in the deck that we provided. Our focus, in terms of our share repurchase program, is secondary to paying down the debt. So we actually slightly underachieved offsetting dilution this quarter. I think it was slightly under $200 million or so that we repurchased. In terms of the Pharmasset debt paydown, we are focused on paying down the short-term debt and we paid about $350 million down this quarter to bring our debt to EBITDA down to about 2.2. Our goal is by mid-2013 to get it to 1.5. That's our target.

And our next question comes from the line of Joel Sendek with Stifel, Nicolaus. Joel D. Sendek - Stifel, Nicolaus & Co., Inc., Research Division: So HIV question. I'm wondering on the Stribild launch whether you encountered any pushback on reimbursement given the price? And then also regarding Europe, I'm wondering why the franchise is growing a little bit slower? I understand this particular quarter's issue, but if you look at each of the last couple of quarters, it's not been growing that much. I'm wondering if there are any other reasons why.

Joel, it's Kevin. I couldn't be happier with actually the access to Stribild. To have 37 ADAP programs, and 4 of the big 5 already in a little under 2 months is where we were really with Atripla. It's very much quicker than what we had with Complera. So we're really having very, very good acceptance. Now just as a reminder, the pricing of Stribild is lower than the boosted protease inhibitor. So that's quite appealing to managed care. Not by a large amount, but it is lower. So managed care likes that. So we priced it very competitively from a regimen cost point of view. When it comes to Europe, Joel, it isn't easy out there. I think we do well because we've got specialist products, HIV requires treatment. We have a great deal of support from physicians. We have good payor support and we've got good advocacy, but some of the markets are tough. Any pharmaceutical company will tell you that, particularly Southern Europe. But our business in the likes of Spain and Portugal, Italy are holding up, I think, pretty well. So it might look a little bit quieter in terms of growth in the U.S., but on a relative basis, I actually think we're doing remarkably well. So Europe, there are still some stresses and there are still some challenges there.

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

Thank you, Ann, and thank you, all, for joining us today. We appreciate your continued interest in Gilead and the team here looks forward to providing you with updates on future progress. Thanks, again.

Ladies and gentlemen, we thank you for your participation in today's conference. This concludes the presentation and you may now disconnect. Have a good day.